Land Reclam. 37.indb - Annals of Warsaw University of Life
Transkrypt
Land Reclam. 37.indb - Annals of Warsaw University of Life
Annals of Warsaw Agricultural University Land Reclamation No 37 Warsaw 2006 Contents ROMANOWICZ R.J. Data Based Mechanistic rainfall-flow models for climate change simulations 3 RAVAZZANI G., MANCINI M., MERONI C. Design hydrological event and routing scheme for flood mapping in urban area 15 DYSARZ T., WICHER-DYSARZ J. Assessment of hydrologic regime changes induced by the Jeziorsko dam performance and morphodynamic processes in the Warta river 43 BANASIK K., BYCZKOWSKI A. Estimation of T-year flood discharge for a small lowland river using statistical method 27 MAJEWSKI G., PRZEWOŹNICZUK W. Characteristics of the particulate matter PM10 concentration field and an attempt to determine the sources of air pollution in the living district of Ursynów 55 WOODWARD D.E., SCHEER C.C., HAWKINS R.H. Curve number update used for Runoff Calculation 33 BARYŁA A. Runoff volume and slope gradient relationship – laboratory investigations 69 PAŃKA D., ROLBIECKI R., RZEKANOWSKI CZ. Influence of sprinkling irrigation and nitrogen fertilization on health status of potato grown on a sandy soil 75 MUSIAŁ E., BUBNOWSKA J., GĄSIOREK E. Variation of climatic water balance and heat balance for various ecosystems in Wrocław in the years 1964–2000 101 MUSIAŁ E.,, BUBNOWSKA J., GĄSIOREK E., ŁABĘDZKI L. Heat balance and climatic water balance in vegetation period of spring wheat 83 VABOLIENÉ G. Investigation for biological nitrogen removal from wastewater using simultaneous nitrification/denitrification technology 111 KASPERSKA-WOŁOWICZ W., ŁABĘDZKI L. Climatic and agricultural water balance for grasslands in Poland using the penman-monteith method 93 LITWIN U., JANUS J., ZYGMUNT M. Development of technologies used in agricultural engineering work on an example of selected stages of land consolidation process 123 Series Editorial Board Elżbieta Biernacka, Warsaw Agricultural University, Chair Wojciech Bartnik, Cracow Agricultural University Marian Granops, Warsaw Agricultural University, Waldemar Mioduszewski, Institute for Land Reclamation and Grassland Farming, Poland Marek Lechowicz, Warsaw Agricultural University, Józef Mosiej, Warsaw Agricultural University Gunno Renman, Royal Institute of Technology, Stockholm, Sweden Michael Hirschi, University of Illinois at Urbana-Champaign, USA Lubos Jurik, Slovak Agriculture University, Nitra, Slovakia SERIES EDITORS Józef Mosiej – Chairman Gunno Renman Janusz Kubrak EDITORIAL STAFF Jadwiga Rydzewska Krystyna Piotrowska WARSAW AGRICULTURAL UNIVERSITY PRESS e-mail: [email protected] ISSN 0208-5771 Edition 500 copies PRINT: Agencja Reklamowo-Wydawnicza A. Grzegorczyk Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 3–13 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Data Based Mechanistic rainfall-flow models for climate change simulations RENATA J. ROMANOWICZ Lancaster University, United Kingdom. Abstract: Data Based Mechanistic rainfall-flow models for climate change simulations. Analysis of climate change indicates an increased probability of high rainfalls and related flooding. Climate change may also lead to an increase in rainfall variability and the probability of extreme events. This forces researchers to focus on both high and low flow models. This paper presents the application of a Data Based Mechanistic (DBM) approach and Stochastic Transfer Function (STF) methods to rainfall-flow modelling, with special emphasis on low flows. We present two different models, one with nonlinearity applied to the input, and the second with nonlinearity on the output. In both cases, the application of stochastic methods of identification and estimation of model parameters allows for the evaluation of predictive uncertainty of the estimated flow. The first model introduces input nonlinearity in the form of effective rainfall. The second model applies the STF approach to log – transformed flow acting as a surrogate of water storage in the catchment. Each of the models has two modules: a groundwater storage module and a surface water module. Logarithmic transformation of the output introduces a bias towards low flows. In order to model also high flows, the fast flow component is transformed using the linear STF model. Both models are applied to a karstic catchment, the River Thet in the United Kingdom. Key words: DBM and STF models, low flow, karstic catchment. INTRODUCTION The increased variability of rainfall magnitude and the resulting increased variability of river flows observed in the past two decades may have been caused by climate changes related to NAO intensification, NSO cycles and/ or global warming. This time period is not long enough to provide the statistical evidence supporting any particular thesis (Robson et al. 1998, 2002). However, globalisation of the world economy requires global thinking about the possible water crisis in order to avoid global disaster. Ongoing developments in modelling future climate and its influence on water resources requires incorporating models able to simulate rainfall-flow processes under changing climatic conditions. From the point of view of water resources management and planning the prediction of extreme (both high and low) flows is the most critical. Low-flow hydrology deals with river flows during the annual dry season. River flow is a result of complex processes of rainfall-water transport operating on a catchment scale. Usually we distinguish fast and slow components of river flow. The slow component, with a large time constant, is attributed to base flow, because the majority of the streamflow during the low flow period originates from groundwater storage. Groundwater storage may be related to the drainage of the saturated top soil zone or a groundwater aquifer. Another source of 4 R.J. Romanowicz low flow may be relatively slow moving groundwater drainage in fracture zones with significant lateral components. Yet another source may be permanently wetted channel bank soils, the bottoms of alluvial valleys or wetlands. The existence of these different sources of low water depends on catchment geology. Smakhtin (2001) lists several different processes which he attributes to low flows, such as karst formations, lakes and glaciers. Additionally there are also anthropogenic impacts on low flows, among others, deforestation, afforestation, industrial and agricultural direct water abstraction, irrigation, return flows and dams and river regulation. The first statistical approach towards the separation of quick and slow flow components was introduced by Young and Beven (1994), who developed a bilinear model for the rainfall-runoff process with input nonlinearity representing the effective rainfall. The main underlying concept was the development of a Data Based Mechanistic (DBM) approach to modelling, combining the data-based model with a mechanistic interpretation of its components (Young 1998). This approach applies Stochastic Transfer Function (STF) methods and uses flow as a surrogate of information on the soil water content in the catchment. Young (2003) extended this methodology to model rainfall-runoff processes “offline”, i.e. in a simulation mode applying a simple first order linear model to describe the soil moisture storage in the catchment used for the nonlinear transformation of rainfall. In this paper we propose an extension of this approach, consisting of using a Transfer Function model representing the flow which is subsequently used as a surrogate of the catchment soil moisture storage. We then compare this model to a novel methodology for modelling low flows. It applies the STF approach to a logarithm of flow, representing water storage in the catchment (Romanowicz 2006). The logarithmic transformation enhances the model performance for low flow values. It introduces a bias to the model solution, as the mean of the logarithm is not equal to the logarithm of its mean value. We apply a separation of the fast and slow components of the estimated STF model and attribute the base-flow to the slow component only. The fast component, representing catchment quick response related to surface or near surface waters, may be modelled using the ordinary STF model without a logarithmic transformation of flow. The model structure has to vary depending on the flow values. We introduce potential evaporation as additional information on flow levels in the catchment to distinguish between high and low flow periods. It is clear that the base-flow module may be calibrated on information-rich data, i.e. in the catchment with a noticeable base flow component in the catchment outflow. The model is suitable for climate change simulations as it uses rainfall and evaporation as input variables, without the necessity of explicit modelling of soil moisture content in the catchment. Moreover, the application of stochastic methods of identification and estimation of model parameters allows for the evaluation of predictive uncertainty of the estimated flow. It is interesting to note that water resource engineers use the logarithm of flow to test model performance for low flow, but this Data Based Mechanistic rainfall-flow models for climate change simulations transformation has never been explicitly applied in a rainfall-runoff model with the aim of low flow modelling. Section 2 presents the structure of both DBM STF models. In section 3 we present the application of a standard DBM rainfall-flow model developed for climate scenario simulations to the River Thet in south-east England. In section 4 we describe the application of the logtransformed low flow model to the same catchment. Section 5 is a discussion of the results. DATA BASED MECHANISTIC (DBM) RAINFALL-FLOW MODEL In this section we present two DBM STF based models developed for the purpose of climate change simulations. The first model follows the DBM approach to rainfall-flow modelling introduced by Young (1998, 2003) and Young and Beven (1994). This model applies the nonlinear transformation of the rainfall. The second model, described in Romanowicz (2006), applies a nonlinear transformation of the flow. Thus the first model has an input nonlinearity, while the second uses the output nonlinearity. This nonlinear transformation is introduced in order to represent process nonlinear behaviour of the process thus assuming that its dynamics is linear. DBM model with nonlinear transformation of the rainfall The methodology follows the approach described in detail by Young (2003) with the difference of applying a Multiple Input Single Output (MISO) 5 Stochastic Transfer Function (STF) model to obtain an estimate of the flow variable, which issubsequently used as a surrogate of soil moisture content in the catchment. It is assumed that the model structure has the first order: b1 st = rt −δ1 + 1 − a1 z −1 + b2 1 − a2 z −1 evt −δ2 + ζ t (1) where: st – is the observed flow used as a surrogate for soil moisture at the end of sample time t; rt, evt – denote, respectively, rainfall and evaporation at the same sample time t; δ1, δ2 – denote any pure, ‘advective’ time delay for each input; ζt – represents the noise (which in some cases can be considered as zero mean, serially uncorrelated Gaussian white noise); a1, a2, b1, b2 – model parameters and the operator denotes a backward shift in time, i.e. z–1ut = ut–1 The parameters are estimated using the Simplified Refined Instrumental Variable (SRIV) algorithm in the CAPTAIN Matlab toolbox (e.g. Young 1989, 2000, Young et al. 1999). Later on, the rainfall-runoff process is described by a second order model, which is decomposed into the fast and slow components (Young and Beven, 1994). Looking for a physical interpretation, the slow model component represents ground water storage and the fast component is related to quick response of the catchment. In the case of daily time periods the third component, representing 6 R.J. Romanowicz direct runoff, may also appear. The model has the following form: Fast component: Z1,t = β1,1 1 + α1 z −1 uefft −δ Slow component: Z 2,t = β2 1 + α 2 z −1 uefft −δ Direct component: Z3,t = β3uefft uefft = rt ⋅ stp yt = Z1,t + Z 2,t + Z3,t + ξt where: α1, α2, β1, β2 – parameters derived from the identified second order STF model; uefft – denotes the effective rainfall obtained from the rainfall multiplied by a nonlinear gain expressed by the power of the flow estimate obtained from (1). All the parameters of both equations and the power p are estimated simultaneously using optimisation methods from MATLAB® optimisation toolbox. Due to the application of stochastic methods of estimation, this methodology gives estimates of the variance of the predictions, as well as the variance and the correlation structure of the model parameters. DBM model with logarithmic transformation of flow In a second approach we apply STF method to the logarithm of flow. The physical explanation of this procedure is described in detail in Romanowicz (2006). In brief, we look at the rate of change of the flow rather than the water balance in the catchment. As mentioned in the introduction, the resulting flow estimates are based towards low flows. The model structure is similar to that shown in equation (2) with second order model dynamics, thus allowing for the decomposition of the model into fast and slow components: Fast component: Z1,t = + β1,1 1 + α1 z −1 β1,2 1 + α1 z −1 rt −δ1 + evt −δ2 Slow component: Z 2,t = + β2 1 + α 2 z −1 β2,2 1 + α 2 z −1 rt −δ1 + evt −δ2 (3) yt = exp(Z1,t + Z2,t + ηt) where: α1, α2, β1,1, β1,2, β2,1, β2,2 – parameters derived from the identified second order STF model. Due to the introduced logarithmic transformation of flow, the variance of model predictions is heteroscedastic and has the form: μ f = exp[(2μ s + σ2s ) / 2] σ2f = exp(2μ s + 2σ2s ) − exp(2μ s + σ 2s ) (4) Data Based Mechanistic rainfall-flow models for climate change simulations where: μs – denotes the mean value of the state variable Zt = Z1,t + Z2,t, σ2s – denotes the estimate of its prediction variance. In order to model also high flows we should include a high flow module in our model. The high flow module may consist of a linear STF model based on either the errors between the low flow component and the observed flow or directly on the observations (without logarithmic transformation) for the periods with high flows. Yet another option is a nonlinear transformation of a fast flow component. This module would require the use of additional information on the conditions in the catchment provided by temperature or potential evaporation. The low flow module can have the form of an STF model for log transformed flows or we can use the slow component as a representation of a base flow in the catchment. In other words, the difference lies in the way we construct the base flow module, either as a full module obtained from log-transformed flows or its slow component only. The choice will depend on the catchment characteristics. For dry catchments, dominated by base flow, the first approach may be suitable. However, for catchments with predominantly high flows, the second approach will be more suitable. APPLICATION TO RAINFALL-FLOW PROCESS IN THE RIVER THET, UK: DBM MODEL WITH EFFECTIVE RAINFALL The Thet catchment, situated in the SE of England, is characterised by predominant base flows, due to its chalk- 7 based geology. The catchment has an area about 307 km2 with mainly arable land and very mixed geology. There are reservoirs in the catchment affecting runoff as well as industrial and agricultural abstractions and effluent returns. Additionally, runoff is affected by groundwater abstraction and recharge. In this section we apply the nonlinear rainfall-flow model (1–2) for climate change simulations with the nonlinearity on the input to the Thet catchment. In order to represent the nonlinear relation between flow and rainfall we apply a State Dependent Parameter (SDP) approach (Young 2000) to the daily rainfall and flow data. Figure 1 shows the resulting nonlinear gain for the rainfall as a function of flow. This nonlinearity is subsequently approximated using the power law, as shown in eq. 2. However, instead of the observed flow we use the estimate of flow obtained from the first order MISO model (1) identified from the data. FIGURE 1. SDP identified nonparametric gain for Thet The first order model for soil moisture has the form: 8 R.J. Romanowicz st = 0.0278 rt + 1 − 0.9844 z −1 0.0131 − evt −1 + ξt 1 − 0.9844 z −1 (5) The second order rainfall-flow model with nonlinearly transformed rainfall has the form: Fast component: Z1,t = 0.0532 1 + 0.7953 z −1 uefft −δ FIGURE 2. DBM model with effective rainfall, validation, Thet; 55% of data variation explained Slow component: Direct component: purpose of comparison with the DBM model with effective rainfall presented in the previous section, we use the same as before datasets for calibration and validation stages. Z3,t = 0.0017 ueft Low flow model Z 2,t = 0.0107 1 + 0.979 z −1 uefft −δ (6) uefft = rt ⋅ st0.21 yt = Z1,t + Z 2,t + Z3,t + ξt This model has an additional direct component representing catchment response shorter than a day. The years 1970–1972 were used for the calibration and the model explained 88% of the observed flow variance. Validation was performed on the years 1990–1993 and only 55% of the flow variance was explained (Figure 2). The best identified MISO STF model has a structure described by [2 2 2 1 2 0], which can be decomposed into the fast and slow components for both rainfall and evaporation. The model has the following form: log(Q1 ) = 0.011rt −1 + 1 − 0.9931z −1 0.027 rt −1 0.0061evt −2 + + + −1 1 − 0.7350 z 1 − 0.9931z −1 0.0078evt −2 − + ξt 1 − 0.7350 z −1 (7) DBM MODEL WITH LOG-TRANSFORMED FLOW In this section we present an application of the DBM model with log-transformed flow to the Thet catchment. For the where rt denotes rainfall and evt denotes evaporation. Data Based Mechanistic rainfall-flow models for climate change simulations The variance of the transformed model estimates (one step-ahead prediction) is 0.038. The results of the calibration give a 92% explanation of the data variation for water storage and 89% of flow after back transformation. The decrease of efficiency is related to the bias towards lower flows, which is introduced by the logarithmic transformation. The decomposition of the model into slow and fast component is shown in Figure 3. The slow component has a time constant of 144 days and describes the base flow in the catchment, while the fast component has a time constant of about 3 days. Due to the logarithmic transformation of flow, after back transformation both components are multiplied by each other. ( ) The efficiency RT2 of the model during the validation period (all 27 years) is 77.6%. The results of the validation on the years 1990–1993 are shown in Figure 9 4. These three years were characterised by a very low flow, and the model gives worst fit to the observations out of the whole validation period. Full DBM model with log-transformed flow For high flows the DBM low flow model output has to be augmented by the surface flow from saturated areas in the catchment. We use evaporation levels to distinguish between the low and high flow periods. For evaporation higher than a certain level, the model output is equal to the exponent of groundwater storage in the catchment. When the evaporation is lower than this threshold value, the flow is increased by the value obtained from the STF model identified from the errors between the base flow and observed flow. The threshold value of evaporation is optimised together with the parameters of the linear in flow FIGURE 3. Decomposition of the model into fast (lower panel) and slow component (upper panel); the observations are shown by a dotted line; calibration stage, years 1970–1973 10 R.J. Romanowicz –1972). The validation was performed on the years (1990–1993) and is shown together with 95% confidence bands in Figure 5. FIGURE 4. The DBM log-flow model: validation stage, years 1990–1993; 57.6% of the data variation explained; observed daily flows are shown by dots, continuous line denotes model simulations, shaded area represent 95% confidence bands model. This model gives nonzero output only for the evaporation smaller than a threshold value. The form of the STF model, identified using Captain toolbox IV (instrumental variable) methods is [1 1 1 2 5], it is a first order model as follows: FIGURE 5. Full DBM model – validation on the years 1990–1993, 65.1% of data variation explained; observed daily flows are presented by dots, continuous line denotes model simulations, shaded area represent 95% confidence bands ⎧ ⎫ 0.08 0.1233 ⎪ ⎪ Qt ,sim − Qt ,obs = ⎨ r − ev −1 t − 2 −1 t −5 evt < 0.4 ⎬ 1 − 0.6886 z ⎪1 − 0.6886 z ⎪ evt > 0.4 ⎭ ⎩ 0 where Qt,sim denotes the flow estimated using (7) and Qt,obs denotes observations. The full DBM rainfall-flow model is the sum of (7) and (8). The threshold value for the evaporation was estimated as 0.4, using an optimisation routine with simultaneous estimation of the model parameters. This model gave a 93% explanation of the flow variation for the calibration period (years 1970– (8) Figure 6 shows the log-log plot of the full DBM model for all the available years 1970–1997. As might be expected, it differs from the groundwater storage model only for high flows. Compared to that model, the gain is 2% over the whole validation period and it considerably improves the high flow predictions. Further work should be done on the use two different time constants for evaporation and rainfall and on the use Data Based Mechanistic rainfall-flow models for climate change simulations 11 FIGURE 6. Log-log results for the full DBM model, River Thet, validation on all the data from the years 1970–1997. of temperature as a threshold value. In the model presented here the choice of the threshold, although based on data, is site specific. It is related to the seasonal temperature variation through evaporation and thus explains high flows which usually occur in winter in the UK. It might not be general enough for other countries with different climatic conditions. CONCLUSIONS We have presented two DBM models suitable for off-line climate change simulations. The first rainfall-flow model applies nonlinearity on the input, following the methodology of Young (2003). The second DBM model for low flows applies log-transformed flows in order to obtain better estimates of the base flow. The model uses evaporation as additional information on soil moisture conditions in the catchment. The full DBM model consists of baseflow and high-flow components. Baseflow is determined by the separation of a low flow module into slow and quick catchment responses. The logarithm of flow is treated here as a substitute for groundwater storage in the catchment. In hydrology the assumption of lognormal distribution for the flow is well justified (Yevjevich 1972). In the result of logarithmic transformation of flow variable, the variance of the predictions of flow is heteroscedastic, depending on the predicted estimates of flow. The models have been tested on the River Thet catchment, UK. The low flow model performs better for this karstic catchment than the model with effective rainfall. In this application the simulated base flow matches well the observed low flow, thus showing that the approach developed is suitable for the climate change simulations and low flow predictions. Further research is underway on generalising the low flow model for catchments with larger differences between high and low flows. Experience gained so far points towards the transformation of the quick flow component of the low flow model 12 R.J. Romanowicz rather than to building a model based on the error between the observed flow and simulated base-flow. The simple structure of a surface store module may be extended by including a representation of off-line effective rainfall, as in the first model, or can be made more general by introducing a complex expression for surface runoff. The important difference from all previous approaches lies in the way in which we condition our model using the available observations. Namely, we use the logarithmic transformation of flow to enhance information on low flows in order to better identify the base flow and we divide the dataset into periods characterised by high flows in order to identify the runoff component of the model. REFERENCES ROBSON A.J., JONES T.K., REED D.W., BAYLISS A.C. 1988: A study of national trend and variation in UK floods, “International Journal of Climatology”, 165–182, 18. ROBSON A.J. 2002: Evidence for trends in UK flooding, “Philosphical Transactions of the Royal Society London A”, 1327– –1343, 360. ROMANOWICZ R.J., 2006: Data Based Mechanistic model for low flows: implications for the effects of climate change, submitted to Journal of Hydrology. SMAKHTIN V.U.: 2001 Low flow hydrology: A review, “J. Hydrol.” 147–186, 240. YEVJEVICH V.M. 1972 Stochastic Processes in Hydrology, Water Resources Publications, Fort Collins, Colorado. YOUNG P.C. 1989: Recursive estimation, forecasting and adaptive control, In: Leondes C. T. (Ed.), Control and Dynamic Systems: Advances in Algorithms and Computation Techniques in Dynamic Systems Control, 30, Academic Press, San Diego pp. 119–166. YOUNG P.C. 2000: Stochastic, dynamic modelling and signal processing: Time variable and state dependent parameter estimation. In: W. J. Fitzgerald, A. Walden, R. Smith, & P.C. Young (eds.), Nonstationary and Nonlinear Signal Processing, Cambridge University Press: Cambridge, 74 –114. YOUNG P.C. 1998: Data-based mechanistic modelling of environmental, ecological, economic and engineering systems, “Environmental Modelling and Software”, 105–122, 13. YOUNG P.C. 1999: Data-based mechanistic modelling, generalised sensitivity and dominant mode analysis, “Computer Phys. Communications”, 113–129, 117. YOUNG P.C. 2003: Top-down and data-based mechanistic modelling of rainfallflow dynamics at the catchment scale, “Hydrological Processes”, 2195–2217, 17. YOUNG P.C., BEVEN K.J. 1994: Data-based mechanistic modelling of rainfallflow nonlinearity, “Environmetrics”, 335–363, 5. YOUNG, P.C., PEDREGAL D., TYCH W. 1999: Dynamic harmonic regression, “J. Forecasting”, 369–394, 18. Streszczenie: Modelowanie niskich przepływów do analizy wpływu zmian klimatu na zasoby wodne. Racjonalna gospodarka zasobami wodnymi, jak również symulacja scenariuszy zmian klimatu nie są możliwe bez lepszego zrozumienia bilansu wody w zlewni, jak również lepszych metod do modelowania niżówek. Niniejszy artykuł przedstawia zastosowanie Mechanistycznego Bazującego na Danych (DBM) podejścia do modelowania procesu opad-odpływ. Przedstawione są dwa modele, pierwszy stosuje nieliniową transformację wejść (opadu), drugi, stosuje logarytmiczną transformację wyjść (przepływu). Obydwa modele wykorzystują liniową Stochastyczną Funkcję Przejścia (STF) do opisu dynamiki procesu. Wyprowadzone modele są zdekomponowane na dwa moduły: jeden reprezentujący szybką dyna- Data Based Mechanistic rainfall-flow models for climate change simulations mikę przepływu (wody powierzchniowe) i drugi, reprezentujący wolną dynamikę zwiazaną z wodami gruntowymi. Zastosowanie stochastycznych metod identyfikacji i estymacji parametrów modeli pozwala na ocenę niepewności predykcji. W związku z logarytmiczną transformacją wyjścia, uzyskana wariancja predykcji modelu z nieliniowością na wyjściu jest zmienna w czasie. W artykule przedstawiamy zastosowanie obydwu modeli do modelowania zlewni o podłożu karstycznym w Wielkiej Brytanii (rzeka Thet). MS. received November 2006 Author’s address Renata J. Romanowicz Environmental Centre Lancaster University Lancaster, LA1 4YQ United Kingdom 13 Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 15–26 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Design hydrological event and routing scheme for flood mapping in urban area GIOVANNI RAVAZZANI1, MARCO MANCINI1 and CLAUDIO MERONI2 1 Politecnico 2 di Milano, MMI s.r.l., Milan (Italy) Abstract: Design hydrological event and routing scheme for flood mapping in urban area. Definition of flood risk maps is a task to which modern surface hydrology addresses a substantial research effort. Their impact on the government of the flood prone areas have increased the need for better investigation of the inundation dynamics [Fema 2002]. This identifies open research problems such as: the definition of the design hydrograph, the identification of the surface boundary conditions for the flood routing over the inundation plan, the choice of the hydraulic model that is the most close to the physical behaviour of the flood routing in the specific environment, such as urban areas or river valley. Most of academic and commercial mathematical models resolving the De Saint Venant equations in mono or bidimensional approach, fail on complex topography. Steep slopes, geometric discontinuities, mixed flow regimes, initially dry areas are just the main problems an hydraulic model should solve. In this study, we address two points: the definition of the critical event for an inundation area and a flood routing modelling technique for a highly urbanized flat area. For this latter we show that, in urban areas, a modelling scheme of a network of connected channels and storages, gives a better representation of surface boundary conditions such as aggregation of buildings and road network and sufficient accuracy for flood risk mapping purpose respect to a real 2-D hydraulic routing model. Key words: design hydrograph, distributed model, flood hazard maps, quasi-2D model, urban flood. INTRODUCTION Flooding is one of the most common environmental hazard, due to the widespread geographical distribution of river valleys and the attraction of human settlements to these areas. Floods can be generally considered in two categories [Castelli. 1994]: flash floods, the product of heavy localized precipitation in a short time period over a given location, and general floods, caused by precipitation over a longer time period and over a given river basin. Although flash flooding occurs often along mountain streams, it is also common in urban areas where much of the ground is covered by impervious surfaces. Fixed drainage channels in urban areas may be unable to contain the runoff that is generated by relatively small, but intense, rainfall events. Many modelling approaches exist to simulate floods [Leopardi et al. 2002]. The choice of the simulation approach depends on the questions to be answered [Horrit and Bates 2001, Ferrante et al. 2000]. In this work we focus on determination of flood hazard map for 16 G. Ravazzani, M. Mancini, C. Meroni an urban area located in the Liguria province in Italy. Traditional approach is based on determination of flood extent for a given frequency event, using steady flow 1-dimensional model. A new rule has been recently introduced for highly urbanized area; it identifies the hazard according to both flow depth and flow velocity, claiming the necessity to use more accurate unsteady flow numerical models. The aim of the work is to assess the accuracy of a quasi-2D model versus pure 2D models for flood prediction in urban area. THE CASE STUDY The study area is located in the western Italy, in the province of Liguria (Fig. 1). The area approaches the sea and has an extension of about 1.8 km2 in which 6 rivers are encountered: Gorleri, Varcavello, S. Pietro, Pineta, Rodine and Madonna. The drainage basin ranges from 0.32 km2 of the river Rodine to 18.05 km2 of the river S. Pietro. The maximum elevation is reached in the Evigno Peak at 988.5 m. a.s.l. The morphology is characterized by steep slopes in the upper part of the basins, decreasing while approaching the sea. The restricted coastal plain has attracted tourism activity with the effect of an exponential growth of urban density. As a consequence the rivers often have been channelized into artificial drainage. In this situation, we can frequently observe, during storm event, water overtopping the levees and flowing into main roads and districts, dragging people and cars. The target area is crossed by a railway that divides it in two parts: the upper with a slope of about 1.5% and the lower with an average slope of about 0.8%. THE DIRECTIVES FOR BASIN PLANNING For the purpose of land use planning and management, flood risk maps are required. The Italian legislation leave the task of the assessment of flood risk maps to the River Basin Authority with FIGURE 1. Aerial photo of the study area. The six rivers and the railway are visible Design hydrological event and routing scheme for flood mapping... the Basin Plan study. The traditional approach simulates the inundation processes with a steady 1-dimensional model. The analysis is iterated for 50 and 200 year return period peak discharge. The 50 year flood extent is classified as class-A hazard and 200 year flood extent is classified as class-B hazard (Fig. 2a). Class-A areas are subject to more restrictive rules than class-B areas. A new scheme recently proposes [FEMA, 2002; Rosso 2003] to determine the hazard map on the basis of both hydraulic depth and flow velocity. According to this scheme, if an area is frequently flooded but with low depth and low velocity, the level of hazard is reduced (Fig. 2b). Moreover a new hazard class is introduced: according to the new scheme, an area can be classified as class-A, class-B and class-B0 hazard. The class-B0 area is the less restrictive one. The necessity to employ flow velocity data, requires the use of more sophisticated hydraulic model, with a 17 the ability to simulate unsteady flow in complex urban terrain. THE DESIGN STORM HYDROGRAPHS FOR INUDATION MAPS There are many types of design hydrographs that have been developed over the years [Maidment 1993; Chow et al. 1988], and the debate is open on the frequencies of the flood map due to the difference between the frequencies of the peak discharge and the hydrograph volume [De Michele et. al. 2005; Salvadori and De Michele 2004]. Moreover, if a rainfall runoff model is used for hydrograph computation, the hypothesis of the equivalence among the return period of the peak discharge and the rainfall is false too. Inundation maps are generally computed using peak discharge for given return period as input variable and simple steady 1-dimensional analysis b FIGURE 2. Evaluation of hazard maps according to the actual rule (a) based on the flooding extent for given return period events and according to the new rule (b) which considers both hydraulic depth and flow velocity 18 G. Ravazzani, M. Mancini, C. Meroni for the river channel, that means that the hydrograph volume and the routing on the inundated areas are not considered. As well known, especially for plain urban area, the storaging volume can significantly affects the routing and the extension of the inundated areas. According to this we think that it is more correct, in the assessing of inundation map, to take into account the inundation volume and its statistics as the key variable for map characterization. For ungaged basin, when the hydrograph is determined from rainfallrunoff transformation, we propose a methodology for the inundation volume definition based on research of the critical rainfall event for an inundation area. This is defined as the one that gives the maximum value of inundation volume for a given return period of the rainfall in the hypothesis that the intensity duration frequency (IDF) curve represents the rainfall behaviour. The maximum inundation volume is defined as the maximum value of the integral of the difference between the incoming hydrograph and the bankfull discharge. For this purpose, rainfall runoff distributed model, FEST [Mancini 1990; Montaldo et al. 2002; Rulli and Rosso 2002; Montaldo et al. 2003; Montaldo et al. 2004; Salandin et al. 2004], was employed. FEST is a distributed hydrologic model especially developed at the Politecnico di Milano focusing on flash flood event simulation. As a distributed model, FEST can manage heterogeneity in hillslope and drainage network morphology (slope, roughness, etc..) and land use [Rosso 1994]. From a family of IDF curves it is possible to obtain an hydrograph for any duration at a given frequency and so a series of hydrographs for every return period (Fig. 3). Given the bankfull discharge for the examined river branch, the hydrograph that presents the maximum inundation volume identifies the critical rainfall event (Fig. 4) for the FIGURE 3. Procedure for the search of the critical event defined as that event which is characterized by the maximum potential inundation volume. The results refer to river Varcavello for the 50 years return period Design hydrological event and routing scheme for flood mapping... 19 FIGURE 4. Series of potential inundation volume for different return period. The design hydrograph for a given return period is the one characterized by the maximum value of the potential flooding volume inundation map. In the end, according to retention pool analysis [Artina et al.1997], the critical event for a flood mapping is different from the critical event for the maximum peak discharge (Fig. 3) and the correlated hydrograph peak presents different return period for a given rainfall frequency (Tab. 1) . DESCRIPTION OF THE HYDRAULIC MODEL Urban areas are usually characterized by streets and aggregation of buildings (in the following termed blocks), that can be schematized, from a flood routing point of view, as network of channels where the flow velocity is greater than zero, and storages where the velocity is about zero. This latter hypothesis derives from the consideration on the friction induced by the macro roughness of the manmade obstacles present in a block. So that the implemented network model presents three main unit: the main rivers, the channels along the main streets and the storages for the aggregation of buildings. De Saint Venant equations are then integrated along the river branches and street channels using the Preissman implicit numerical scheme [Wallingford Software 2005]. Energy and continuity equation is verified at nodes given by the channel intersection. Water discharge can flow in every direction in the channel network according to the hydraulic gradient. Reservoir equation is used to model diffusion in the blocks. The connection between storages and channels are discretized in the specific nodes and the weir equation controls the water flux to and from the reservoir according to the difference of water level. The river branches feed the channel network where the cross sections are insufficient respect to the flood discharge. When the riverbanks are overtopped, water enters the street channels. The channel network model is a very good representation of urban flood routing when the hydraulic depth 20 G. Ravazzani, M. Mancini, C. Meroni TABLE 1. Comparison between maximum peak discharge and peak discharge of the hydrograph with the maximum inundation volume computed for a given rainfall frequency. In the last column the return period of the peak discharge of the hydrograph with the maximum inundation volume is reported Rainfall return period (years) 50 Return period of the peak Peak discharge of the Maximum peak discharge of the hydrograph hydrograph with the maximum discharge (m3/s) with the maximum inundation inundation volume (m3/s) volume (years) 94.04 79.82 25.5 100 111.75 86.37 38 200 132.82 95.61 54 500 161.24 99.7 64.5 and energy is lower than the height of the surrounding buildings which act as impervious boundary elements. The described channel network model was implemented (Fig. 5) in the Infoworks-CS software [Wallingford Software 2005]. Rivers and main roads are represented by conduits. Manholes allow the exchange of water between river and roads and are used to represent crossroads. Conduits are linked with storages through weirs. The main model parameters are the roughness coefficient which controls flow velocity in the channels and the weir discharge coefficient which controls the amount of water exchanged between channels and storages. The adopted set of values are reported in Table 2. FIGURE 5. Network quasi-2D hydraulic model representing urban area: detail of the river Varcavello Design hydrological event and routing scheme for flood mapping... 21 TABLE 2. Parameters describing components of the network quasi-2D model Parameter Value Natural river channel Strickler roughness 30 m1/3s–1 Concrete river channel Strickler roughness 50 m1/3s–1 Not asphalted street Strickler roughness 30 m1/3s–1 Asphalted street Strickler roughness 50 m1/3s–1 Weir discharge coefficient for districts upstream the railway (a) 0.1 Weir discharge coefficient for districts downstream the railway (a) 0.28 (a) 50 m Weir length districts downstream the railway (a) 20 m Weir length for districts upstream the railway (a) Weir formula adopted for discharge computation in the model: Q = Cd BDu g ( Du − Dd ) , where Q is the discharge [m3/s], Cd is the discharge coefficient, B is the width of the weir [m], Du is the upstream depth with respect to the crest [m], Dd is the downstream depth with respect to the crest [m] and g is the acceleration due to gravity [m/s2]. THESIS VERIFICATION USING 2-DIMENSIONAL MODEL The basic assumption of the network model is that velocity of flood flow over the urban area is greater than zero in the main streets while, in the blocks, velocity can be neglected. To verify this assumption, we analysed flood dynamic in the blocks, by means of a high resolution 2-D model. A subset of urban land in proximity to river Varcavello, has been extracted from the main domain. It is composed by two main blocks: one upstream the railway and the other downstream. The upper one is characterized by an average slope of about 1.5%, the lower by a slope of about 0.8%. These subsets are considered to be representative of the whole study area. A full 2-D model was implemented using the SMS software [www.bossintl. com] (Fig. 6). A steady analysis has been performed. Hydraulic depths deriving from channel network simulation have been taken as boundary condition of the blocks. Buildings have been modelled in two different ways: as impervious area (buildings are excluded from the model domain) and as a high roughness surface (Strickler coefficient equal to 0.01 m1/3s–1). The latter way means that water is free to move in the whole domain, but the flow through buildings is made difficult because of an high value of roughness parameter. Two cases for the Strickler coefficient for gardens surrounding buildings were considered: in the first case its value was fixed to 5 m1/3s–1and in the second case to 10 m1/3s–1. Velocity field deriving from simulations has been mapped to a regular grid and the cumulative frequency of the velocity values has been evaluated (Fig. 7). In the upstream block (Fig. 7a), we can note that most of the cells (67%) has a value of velocity less than 0.6 m/ s even for the simulation with Strickler coefficient for gardens equal to 10 m1/3s–1 and impervious buildings. In the other simulations, the percentage of cells with velocity under 0.6 m/s increases nearly to 100%. Both 22 G. Ravazzani, M. Mancini, C. Meroni a b FIGURE 6. Subset of the study area (a) near river Varcavello upstream the railway and (b) representation in the 2-D model a b FIGURE 7. Cumulative frequency of velocity magnitude simulated by means of full 2D model in (a) upstream district and (b) downstream district; ks_g and ks_b denote, respectively, the Strickler roughness coefficient for gardens and buildings considering buildings impervious or as high roughness surface, the peak flow velocity reaches, anyway, a maximum value of 1.3 m/s in just a few cells. The observations are valid for the downstream block (Fig. 7b) too. As a consequence of a milder slope, flow velocity is lower indeed. We can conclude that motion of water in the aggregation of buildings is characterized by low values of velocity. The assumption to simulate blocks as storages in the channel network model seems reasonable. Design hydrological event and routing scheme for flood mapping... THE FLOOD HAZARD MAPS Flow velocity and hydraulic depth values, computed respectively in the channels and in the nodes of the network model, have been interpolated over the entire study area to obtain a continuous map. For this purpose the borders of districts have been considered as barriers. The overlay of the velocity with the depth map, produced the hazard map (Fig. 8), according to the new directive of the River Basin Authority (§ 3). A comparison is made with the previous study performed by Basin Authority [Regione Liguria] which made use of steady flow computation model for delimitation of flood extent, not regarding flow velocity. Total amount of flooded area has increased in this new study (Fig. 9) from 0.97 to 1.27 km2. The class-A areas and class-B areas have decreased, respectively from 0.56 to 0.24 km2 and from 0.42 to 0.16 km2. The complementary area is included in the class-B0 area which was not defined in the previous analysis. The differences in the framework of the present study respect to the previous 23 one, have counterpart in the resulting flooding map (Fig. 8). In the network quasi-2D model, overtopping water is routed along main roads as far as distant blocks. This is good explanation for those areas that, in the previous study, are not even water logged by 200 year flood event: the pure 1D steady model can’t predict flood processes in urban area. The new framework can also predict water logging due to manmade obstacles orthogonal to the flow direction. The railway divides the city on north-south direction and behaves as an impervious levee to water flow. This does not seem to be completely represented by the 1D model. CONCLUSIONS A procedure is presented for estimation of design hydrographs. It is based on the search for the critical event which maximizes the potential flooding volume. The iterative process makes use of the FEST model for the rainfallrunoff transformation. The distributed model permits to well represent basins characterized by heterogeneous FIGURE 8. (a) Flood hazard map resulting in this study, compared to (b) actual hazard maps published by Basin Authority 24 G. Ravazzani, M. Mancini, C. Meroni FIGURE 9. Extension of class-hazard area as resulting from this study, compared to previous study performed by Basin Authority morphology and land use, as the ones in this work. The critical event guarantees that the hydrograph with the maximum effect on territory is assumed, even if peak discharge is lower than the maximum peak discharge for the given rainfall frequency. Classical one dimensional models are poor tools for flood analysis in urban area. They can be used as long as the main stream is not overtopped, as they fail to simulate flow component other than along river direction. Two dimensional models, on the other hand, are time consuming and, for unsteady flow simulations on complex topography, they fail on steep slopes, geometric discontinuities, mixed flow and initially dry areas. This work proposes an hybrid approach. The urban and drainage system is modelled by means of a network in which both rivers and roads are modelled as channels linked by nodes. The basic assumption is that high density urban blocks can be modelled as storages in which flow velocity is null. The assumption is verified by a 2D model results which confirm that flow velocity in the blocks is negligible. The channel network model seems to well represents flood routing in urban area, it is not computationally expensive as 2D model and, above all, is much more stable on complex topography. On the other hand it requires a deep knowledge of the territory and a good skills of the modeller to set up the hydraulic sketch. REFERENCES ARTINA S., CALENDA G., CALOMINO F., LA LOGGIA, G., MODICA C., PAOLETTI A., PAPIRI S., RASULO G., VELTRI P. 1997: Sistemi di Fognatura. Manuale di Progettazione. Hoepli, chap. X,XVIII, Milan. (In Italian). Boss International. http://www.bossintl.com CASTELLI F. 1994: Spatial scales of frontal precipitation. In Advances in Distributed Hydrology, R. Rosso, A. Peano, I. Becchi and G. A. Bemporad Editors, Water Resources Publications, 87–114. CHOW V.T., MAIDMENT D.R., MAYS L.V. 1988: Applied Hydrology. McGrawHill. Design hydrological event and routing scheme for flood mapping... De MICHELE C., SALVATORI G., CANOSSI M., SETACCIA A., ROSSO R. 2005: Bivariate statistical approach to check adequacy of dam spillway. Journal of Hydrologic Engineering, 10 (1): 50–57. FEMA, Federal Emergency Management Agency, 2002: Guidelines and specifications for flood hazard mapping partners. FEMA Publications. FERRANTE M., NAPOLETANO F., UMBERTINI L. 2000: Optimization of transportation networks during urban flooding. Journal of the American Water Resources Association, 36 (5): 1115– –1120. HORRIT M.S., BATES P.D. 2001: Predicting floodplain inundation: raster-based modelling versus the finite element approach. Hydrological Processes, 15, 825–842. LEOPARDI A., OLIVERI E., GRECO M. 2002: Two-dimensional modelling of floods to map risk-prone areas. Journal of Water Resources Planning and Management, 128(3),168–178. MAIDMENT D.R. 1993: Handbook of hydrology. McGraw-Hill. MANCINI M. 1990: La modellazione distribuita della risposta idrologica: effetti della variabilitŕ spaziale e della scala di rappresentazione del fenomeno dell’assorbimento. PhD thesis, Politecnico di Milano. (In Italian). MONTALDO N., MANCINI M., ROSSO R. 2004: Flood hydrograph attenuation induced by a reservoir system: analysis with a distributed rainfall-runoff model. Hydrological Processes, 18(3), 545–563. MONTALDO N., RAVAZZANI G., MANCINI M. 2003: The role of the antecedent soil moisture condition on the distributed hydrologic modelling of the Toce alpine basin floods. International Conference on Alpine Meteorology and MAP-Meeting 2003, http://www.map2. ethz.ch/icam2003/ICAM-MAP2003.htm MONTALDO N., TONINELLI V., MANCINI M., ROSSO R. 2002: Coupling Limited Area Models with Distributed Hydrologic 25 Models for Flood Forecasting: the Toce Basin Study Case. IAHS, 274, 229–236. Regione Liguria, Autoritŕ di Bacino di regionale. Piano di bacino stralcio per la difesa idraulica ed idrogeologica, (Ambito di Bacino no 7 – DIANESE). (In Italian). ROSSO R. 1994: An introduction to spatially distributed modelling of basin response. In Advances in Distributed Hydrology, R. Rosso, A. Peano, I. Becchi and G.A. Bemporad Editors, Water Resources Publications, 3–30. ROSSO R. 2003: Consulenza tecnico scientifica per la definizione degli ambiti normativi relativi alle fasce di inondabilitŕ in funzione di tiranti idrici e velocitŕ di scorrimento. Abstract on http://www. regione.liguria.it (In Italian). RULLI M.C., ROSSO R. 2002: An integrated simulation method for flash-flood risk assessment: 1. Frequency predictions in the Bisagno River by combining stochastic and deterministic methods. Hydrology and Earth system Sciences, 6(2), 267–283. SALANDIN A., RABUFFETTI D., BARBERO S., CORDOLA M., VOLONTČ G., MANCINI M. 2004: Monitoraggio e simulazione numerica del fenomeno finalizzata alla previsione e gestione dell’emergenza. Neve e Valanghe, 51. (In Italian). SALVADORI G., De MICHELE C. 2004: Analytical calculation of storm volume statistics involving Pareto-like intensityduration marginals. Geophysical Research Letters, 31 (4). Wallingford Software. http://www. wallingfordsoftware.com/products/ infoworks_cs/. Site visited on 2005-07-08. Streszczenie: Konstruowanie hydrogramów i schematów obliczeniowych do odwzorowania obszarów zagrożonych powodziami na terenach zurbanizowanych. Wyznaczanie map ryzyka powodziowego jest jednym z ważniejszych zagadnień we współczesnej hydrologii. Ich wpływ na zarządzanie terenami zalewowymi zyskał duże 26 G. Ravazzani, M. Mancini, C. Meroni znaczenie, co spowodowało potrzebę badań nad dynamiką powodzi. Było to również przyczyną intensyfikacji badań nad następującymi zagadnieniami: zdefiniowanie hydrogramu projektowego, identyfikacja powierzchniowych warunków brzegowych do obliczeń zasięgu powodzi, wybór modeli jak najlepiej opisujących zasięg powodzi w specyficznym środowisku, jakim jest obszar zurbanizowany lub doliny rzeczne. Większość modeli matematycznych zarówno szkoleniowych, jak i komercyjnych, rozwiązuje równania De Sait Venant’a w jednym lub dwu wymiarach, nie sprawdza się jednak dla skomplikowanych topograficznie powierzchni. Duże spadki, nieciągłości geometryczne, różne zmieniające się reżimy przepływu, początkowo suche obszary są głównymi problemami, które powinny być rozwiązywane przez modele hydrauliczne. W tych badaniach skupiono się na dwóch zagadnieniach: zdefiniowane zdarzenia krytycznego dla terenów zalewowych i technice modelowania terenów zalewowych dla silnie zurbanizowanej płaskiej powierzchni. W tym przypadku schemat połączonych kanałów i zbiorników retencyjnych dał lepsze odwzorowanie powierzchniowych warunków brzegowych, takich jak kompleksy budynków, sieć ulic i wystarczająca dokładność w określaniu obszarów zagrożonych powodziami w porównaniu do dwuwymiarowego modelu hydraulicznego. MS received November 2006 Authors’ addresses: Giovanni Ravazzani Marco Mancini Politecnico di Milano Plazza L. da Vinci 32, 20133 Milano Italy Claudio Meroni MMI s.r.l, Via Aselli 24, 20133 Milano Italy Corresponding author: e-mail: [email protected] Fax +39 02 2399 6207 Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 27–31 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Estimation of T-year flood discharge for a small lowland river using statistical method KAZIMIERZ BANASIK, ANDRZEJ BYCZKOWSKI Warsaw Agricultural University – SGGW, Department of Water Engineering and Environmental Restoration Abstract: Estimation of T-year flood discharge for a small lowland river using statistical method. The 34-year series of daily discharges from a small agricultural river basin of the Zagożdżonka river at the gauging station of Płachty Stare (A = = 82.4 km2), located in the center of Poland, were used in the investigation. The Pearson type 3 (i.e. 3-parameter gamma) distribution and the log-normal distribution were considered to find the best fit with the empirical data. The IMGW (Institute of Meteorology and Water Management) computer program, which applies the method of maximum likelihood, was used to estimate the parameters of the above distributions. Also commonly applied in engineering practice, the Pearson type 3 distribution with the method of quantile for estimating the distribution parameters was additionally used. Log-normal distribution gives better fit with the measured annual flood flows, applying Akaike criterion, and the best fit applying subjective visual criterion. Key words: flood flow, small river basin, peak flow, frequency curve. INTRODUCTION Estimation of flood discharges are needed in designing of hydraulic or road structures, as well as in flood protection activity. It is often assumed that in case of existence of long term hydrometric records, the procedure for estimating T-year flood discharge could be carried out without larger difficulties, which however does not always take place, as selection of various probability function may produce significant differences in the results (Maidment 1993, Pilgrim and Doran 1993). In case of the studied river basin, which is research river basin of Department of Water Engineering and Environmental Restoration of Warsaw Agricultural University, the results are also important as reference values for application and verification of indirect methods for the T-year flood discharge estimation, as well as for rainfall-runoff model verification. In the paper there has been analysis carried out for examining the following statistical distributions: a) the Pearson type 3 – P3 (i.e. 3-parameter gamma) distribution and, b) the log-normal distribution – LN. For the statistical calculation the IMGW (2005; Institute of Meteorology and Water Management) computer program, which applies the method of maximum likelihood – MML, was used to estimate the parameters of the above distributions. Also commonly applied in engineering practice in Poland, the Pearson type 3 distribution with the method of quantile – MQ, for estimating the distribution parameters was additionally used. Data of 34-year daily discharges (1969–2002) of small river basin of Zagożdżonka were used for the analyze. 28 K. Banasik, A. Byczkowski DATA USED Data of discharge of the Zagożdżonka river in the gauge of Plachty Stare, where area of the basin is 82.4 km2, has been colleted by the Department of Water Engineering and Environmental Restoration (former Department of Hydraulic Structures) of the Warsaw Agricultural University – SGGW since 1962. Until 1980 water stages had been measured by observer three times a day, when additionally water stage recorder was installed. Location of the river basin is shown in the Figure 1. The mean annual precipitation and runoff are estimated at 610 mm and 109 mm respectively. The Zagożdżonka watershed is of lowland type. The absolute relief is 37 m. The mean slops of the main streams are from 2.5 to 3.5‰. The land use is dominated by arable land (small grain and potatoes). Sandy soils are the dominant soil types in the watershed (Byczkowski et al. 2001). Hydrometric discharge measurements, carried out at the Płachty Stare gauge up to ten times a year, have been basis for rating curve estimation and continue its verification. Annual maximum flood flows from the period 1969–2002 were used for the statistical analysis. This period has been accepted, as the whole 40-year, i.e. 1963–2002 (Banasik et al. 2003) period of records did not fulfill the postulate of homogeneity according to Kraskal-Wallis (IMGW, 2005). RESULTS OF WQp% (T-YEAR FLOOD DISCHARGE) ESTIMATION AND CONCLUDING REMARKS Using the IMGW computer program (Ozga-Zielińska et al. 1999; IMGW, 2005), which apply the method of maximum likelihood (MML) for parameter estimation, two distribution function were selected from four for further analyze, based on the Akaike criterion. The selected distributions have been as follow: FIGURE 1. Map of the upper part of the Zagożdżonka River basin Estimation of T-year flood discharge for a small lowland river... • the Pearson type 3 (P3; i.e. 3parameter gamma) distribution, and • the log-normal (LN) distribution. Very close values of the Akaike criterion for the both distribution (i.e. 167,04 and 166,03 for the Pearson type 3 and the log-normal distributions respectively) influence the decision to include both of the functions in the analyze. The statistical series of the 34-year period have been also approximated using the most frequently used in Poland probability distribution according to the Pearson type 3, with the method of quantiles (MQ; Kaczmarek, Trykozko 1964). The calculations of the T-year discharges (or discharges with p% probability of exceedance) are based on quantiles of the following range: 10%, 50%, 90% and 100%. The maximal discharges are estimated from the equation (Byczkowski 1999): WQp = WQ50%[1 + cvФ(p; s)] (1) where: WQp – maximal annual discharge with the probability of exceedance p%, 29 WQ50% – maximal discharge with the probability of exceedance p = 50%, cv – coefficient of variation of the series of maximal discharges, Ф(p; s) – probability function for a given type of distribution. The results of the calculations for the three distributions are shown on the fig. 2 and given in the Table 1. Results presented on the Figure 2 and in the Table 1 show relative significant differences in the flood discharges of small probability of exceedance, estimated according the two different statistical distributions, i.e. the lognormal distribution and Pearson type 3 distribution. As the difference of the Akaike criterion for the above statistical distribution was relatively small (i.e. 166.03 for the lognormal and 167.04 for the Pearson type 3 distribution), one may consider to select also other criterion at choosing the propel statistical distribution. Visual assessment of agreement of the theoretical distributions with the empirical data, shown on the Figure 2, indicates also that the TABLE 1. Flood flows according various distributions and methods of parameter estimation for the Zagożdżonka River at Płachty Stare gauge station Return period T (year) Probability p (%) (1-CDF) Flood flows (m3/s) according to distribution and method of parameter estimation Lognormal -LN Pearson type 3 - P3 Pearson type 3 – P3 method of quantiles method of maximum likelihood – MML - MQ 71.4 33.3 34.3 1000 0.1 100 1.0 31.1 21.8 22.1 20 5.0 14.9 13.9 13.8 10 10 10.2 10.5 10.3 2 50 2.79 3.20 2.80 1.11 90 0.98 0.94 0.75 1.01 99 0.59 0.63 0.59 30 K. Banasik, A. Byczkowski 1 2 5 10 20 50 100 200 500 T 60 1000 55 50 LN-MML 45 P3-MML 40 Q [m3/s] 35 30 25 P3-Mq 20 15 10 5 100 99 90 80 70 60 50 40 30 20 10 5 2 1 0.5 0.2 p [%] 0.1 FIGURE 2. Annual peak discharges frequency curve for the Zagożdżonka River at Płachty Stare gauge for the period of 1969–2002 lognormal distribution fit better with the measurement results. The similar results of flood discharges of various probabilities, estimation according to Pearson type 3 distribution, with the use of the two various method of parameter estimation (i.e. method of maximum likelihood and method of quantiles) indicate that the method has had small influence on the results of computation. As the five largest flood flows, shown on the Figure 2, indicated heterogeneity with the remaining data, an analyze in which a division of the floods caused by rainfall events and snowmelt events, as various genesis for the river basin responses, seems to be needed. Acknowledgment. The study described in this paper has been carried out within research project Nr 1010/P01/2006/30, founded by Ministry of Science and Higher Education. The financial support provided by this organization is gratefully acknowledged. The assistance of Mr J. Gładecki in the process of preparing the material for the analyses is also gratefully acknowledged. REFERENCES BANASIK K., BYCZKOWSKI A., GŁADECKI J. 2003: Prediction of T-year flood discharge for a small river basin using direct and indirect methods. Ann. Warsaw Agricult. University – SGGW, Land Reclam. No 23, p. 3–8. Estimation of T-year flood discharge for a small lowland river... BYCZKOWSKI A. 1999: Hydrologia t. 2. Wyd. SGGW, Warszawa. BYCZKOWSKI A., BANASIK K., HEJDUK L., MANDES B. 2001: Wieloletnie tendencje zmian procesu opadu i odpływu w małych zlewniach nizinnych - na przykładzie rzeki Zagożdżonki (Long-term tendency in changes of precipitation and runoff in a small lowland river basin of Zagożdżonka). Instytut Meteorologii i Gospodarki Wodnej. Atlasy i monografie. Warszawa, p. 43–52. IMGW, 2005: Zasady obliczania największych przepływów rocznych o określonym prawdopodobieństwie przewyższenia – Długie ciągi pomiarowe przepływów (Guidelines for computation of annual flood discharges with small probability of exceedance – long data sets). Instytut Meteorologii i Gospodarki Wodnej, Warszawa. KACZMAREK Z., TRYKOZKO E. 1964: Application of the method of quantiles to estimation of the Pearson distribution. Acta Geoph. Polon. T. 12, z. 1. MAIDMENT D.R. (ed.) 1993: Handbook of hydrology. McGraw-Hill, Inc. New York. OZGA-ZIELIŃSKA M., BRZEZIŃSKI J., OZGA-ZIELIŃSKI B. 1999: Zasady obliczania największych przepływów rocznych o określonym prawdopodobieństwie przewyższenia – przy projektowaniu obiektów budownictwa hydrotechnicznego. (Guidelines for computation of annual flood discharges with small probability of exceedance – for designe of hydrotechnical structures). Materiały Badawcze, Seria: Hydrologia i Oceanologia. IMGW, Warszawa. PILGRIM D.H., DORAN D.G. 1993: Practical criteria for the choice of method for estimating extreme design floods. Extreme Hydrological Events: Precipitation, Floods and Droughts (Proceedings of the Yokohama Symposium, July 1993). IAHS Publ. no 213, p. 227–235. 31 Streszczenie. Wyznaczenie przepływów maksymalnych prawdopodobnych w małej zlewni nizinnej przy zastosowaniu metody statystycznej. Codzienne przepływy rzeki Zagożdżonki w Płachtach Starych (A = 82,4 km2), położonej na Równinie Radomskiej z okresu 34 lat były podstawą wyboru wartości maksymalnych rocznych, wykorzystanych do wyznaczenia przepływów maksymalnych o małym prawdopodobieństwie przekroczenia. Rozkłady prawdopodobieństwa – gamma i logarytmiczno-normalny, wybrano do opisu własności losowych ciągu maksymalnych przepływów rocznych. Przeprowadzone obliczenia za pomocą programu komputerowego IMGW, wykorzystujacego metodę największej wiarygodności do wyznaczania parametrów, wykazały lepszą zgodność rozkładu lograytmiczno-normalnego z danymi empirycznymi niż rozkładu gamma (Pearsona typu 3), zarówno przy zastosowaniu kryterium Akaike, jak również według oceny wizualnej (subiektywnej). Przy rozkładzie tym (logarytmiczno-normalnym) uzyskano znacznie wyższe wartości przepływów maksymalnych o niskim prawdopodobieństwie przekroczenia (1% i 0,1%). Wyznaczając przepływy maksymalne prawdopodobne metodą dotychczas powszechnie stosowaną w praktyce inżynierskiej tj. przyjmując rozkład Pearsona typu 3, z parametrami wyznaczonymi metodą kwantyli, wykazano znikomy wpływ metody szacowania parametrów na postać rozkładu prawdopodobieństwa. MS. received November 2006 Authors’ address: Kazimierz Banasik, Andrzej Byczkowski Wydział Inżynierii i Kształtowania Środowiska SGGW 02-787 Warszawa, ul. Nowoursynowska 166 Poland e-mail: [email protected] [email protected] Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 33–42 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Curve Number update used for runoff calculation DONALD E. WOODWARD*, CLAUDIA C. SCHEER* RICHARD H. HAWKINS** *Natural Resources Conservation Service, USA **University of Arizona, USA Abstract: Curve Number update used for runoff calculation. The Natural Resources Conservation Service (NRCS), formerly the Soil Conservation Service (SCS), developed the runoff curve number procedure for estimating direct runoff from rainfall on ungaged agricultural watersheds in the late 1950s. This procedure is being used world wide to estimate direct runoff from rainfall events and, more recently, applied to continuous simulation models. In 1990 a work group of hydraulic engineers from NRCS and the Agricultural Research Service (ARS) began an effort to update the procedure for estimating direct runoff. The results of this joint effort will be explained in this paper. Key words: NRCS runoff curve, event analysis, model fitting. INTRODUCTION The NRCS runoff curve number procedure was developed in the late 1950s to estimate the runoff volumes. It was developed as a simple procedure for estimating direct runoff for use in the design of conservation practices. Curve numbers represent the runoff potential of various soil-cover combinations. Originally, curve numbers were developed by utilizing the maximum annual daily runoff and associated rainfall data from ARS watersheds located across the United States. The original curve numbers encompassed a wide variety of hydrologic soil groups, agricultural land uses and hydrologic conditions. Curve number values are found in the NRCS (NEH-630 1985). Over the years, additional curve numbers have been developed for new land uses, tillage practices and cover conditions. Since adequate rainfall-runoff data was not always available for these new situations, alternative methods of developing curve numbers were utilized. Many of the newer curve numbers were developed by comparison of values for similar land uses, or by weighting current land use curve numbers, a method particularly favored in the development of urban curve numbers. In the 1980s, NRCS recognized that an in-depth review of the runoff curve number procedure was needed in order to adequately address concerns with curve number usage. These concerns included the development of curve numbers for additional land use and cover conditions from limited data, the need of curve numbers for new land uses and cultivation practices, regional and seasonal variation of curve numbers. At the 1989 ARS/NRCS Hydraulic Engineers Workshop, the principle subjects discussed were the use and abuse 34 D.E. Woodward et al. of curve numbers, the application of curve numbers in continuous hydrologic models, and the future of the runoff curve number procedure. Meeting participants recommended that an ARS/NRCS Work Group be established to further the study of curve numbers and investigate the potential need to develop curve numbers for major regional land areas. At the 1992 meeting, the Curve Number Work Group decided that the procedures used to determine the first curve numbers needed to be revised and updated. A limited test indicated that plotting the ordered values rather than the paired or natural values provided realistic results. In the initial and subsequent analyses, annual maximum daily rainfall and runoff values were used. It was noted that while paired values matched nature, ordered values better matched design and evaluation watershed studies. The Work Group concluded that they did not have the resources to analyze all of the data available for small watersheds in an orderly and timely manner. Hawkins (1998) had the most complete, organized set of ARS data from small agricultural watersheds and spent considerable time and effort to remove errors from this data and was interested in dedicating the time needed to complete the required analysis. About this same time period, the American Society of Civil Engineers (ASCE) established a task group whose purpose it was to review the NRCS rainfall-runoff estimation procedure known as the curve number system. The Curve Number Work Group revised eight chapters of National Engineering Handbook (NEH-630) Revised version l of NEH-630 may be found at: http://www.wcc.nrcs.usda. gov/water/quality/common/neh630/ 4content.html . During the revision of the NEH630 chapters incorrect or misleading statements were corrected and, in some cases, certain points were reemphasized for clarity. The most important changes were: • The reference to Antecedent Moisture Conditions (AMC) was removed and the terminology was changed to Antecedent Runoff Condition (ARC). • It was recognized that ARC II represents the average watershed condition of the watershed when flooding occurs. This means the average watershed conditions could be different for different geographical areas when flooding occurs. In other words, a particular runoff curve number for ARC II in Arizona represents a different condition than the same curve number for ARC II in Oregon. • Explicit expression of the Runoff Curve Number Equation as a transformation of the rainfall frequency distribution to runoff frequency distribution was reemphasized. • ARC I and ARC III were expressed as measures of dispersion about the central tendency (ARC II). This is a corollary to treating curve number as a random variable. • Mathematical proof and demonstration that S does not include Ia was documented. • The desirability of locally determined curve numbers was reiterated. Although this was part of the original documentation in NEH-630, local calibration has been seldom pursued. Curve Number update user for runoff calculation ORDERED PAIRS AND ASYMPTOTES One of the goals of the Work Group was to standardize the procedure for calculating curve numbers from rainfall-runoff data. The accepted handbook method was to plot the annual series of rainfall-runoff on a scatter diagram and select the curve number that best fit the data. However, this method ignored the many storms that were not the largest annual events. The runoff curve number equation is often used to transform a rainfall frequency distribution into a runoff frequency distribution. For example, the 100-year rainfall is used to determine the 100-year runoff, etc. This practice, called frequency matching, led to the idea of ordered pairs (Hjelmfelt 1980). Hawkins (1993) followed this idea by sorting the rainfall and runoff depths separately and re-aligning them on a rank order basis, creating new sets of rainfallrunoff pairs. These rainfall-runoff pairs can be thought of as having equal return periods with the individual runoffs not necessarily associated with the original causative rainfall. Using all the storms in the data set, Hawkins calculated a curve number for each of the ordered pairs and plotted them against the rainfall. The curve number was found to vary with storm depth, but for most cases approached a constant value at higher rainfalls. The resulting curves were fitted with asymptotic equations to approach this constant curve number value. The limiting curve number, approached as rainfall approaches infinity, is taken as the best fit curve number for the watershed. 35 This method of determining a watershed curve number has the advantage of being mathematical and therefore programmable. Results are mostly influenced by the largest event, which is in keeping with the usual intended applications of the curve number. . This method appears to give results consistent with the present procedures and curve number tables in NEH-630. The curve number Work Group has adopted this procedure for future determination of curve numbers from local data. MODES OF APPLICATION OF THE CURVE NUMBER METHOD The Work Group recognized three distinctly different modes of application for curve numbers: 1. Determination of runoff volume for a given return period, given total event rainfall for that return period. 2. Determination of direct runoff for individual events. This acknowledges the variation between events and is the basis for the initial development. 3. Process models, an inferred application as an infiltration model, or a soil moisture-curve number relationship, or as a source area distribution. The first application is the most widely used in engineering and uses the curve number to transform the rainfall frequency distribution into a runoff frequency distribution. The runoff volume that is computed is often overlooked and the peak discharge, which is more frequently the desired value, is calculated using a unit hydrograph model. 36 D.E. Woodward et al. The second application, runoff from individual rainfall events, is the basis for the original development, as the rainfall versus runoff plots led to the curve number concept. There is a wide variation of runoff from rainfalls of the same magnitude which forces us to acknowledge that curve number varies among storms for a wide range of reasons. The original handbook, developed in large part for conditions in the humid east, south, and mid-west, designated Antecedent Moisture Condition, or AMC, as the most significant variable in explaining this. Average moisture condition was called AMC II and applied to the curve number when flooding occurs. Dry conditions (AMC I) applied to the low curve number, and wet conditions (AMC III) applied to the high curve number. This condition is most often determined by prior rainfall since soil moisture conditions are not frequently monitored. The Work Group studied the effect of soil moisture on curve numbers by looking at infiltrometer studies (Van Mullem 1992). Four studies with 162 data pairs on 86 different soils from across the United States were used. Although average curve number increased from 9% to 40% between the studies from the initial test (dry to average condition) to the wet condition test (average to wet condition), no significant relationship was found between soil moisture and curve number. The study indicated that the difference in curve number that might be related to soil moisture is much less than the variation between ARC I and ARC III. Similarly, Hawkins and Cate (1998) showed that 5-day prior rainfall was the only consistent factor in explaining deviations from the central trend of runoff in 11 of 25 agricultural watersheds which were studied, and at levels far below handbook expectations. Because prior rainfall explains only part of the variation of curve number, the terminology has been changed to Antecedent Runoff Condition, or ARC. More importantly, the ARC I and ARC III conditions have been shown to be the bounds on the distribution of curve number. Figure 1 shows the ARC I, II, and III curve numbers plotted on a rainfallrunoff scatter diagram. Upon review of this by Hjelmfelt (1980) indicated that there appears to be a correlation between exceedance limits and ARC. Figure 2 shows that ARC I represents the 10% exceedance limit and ARC III represents the 90% exceedance limit for a number of agricultural watersheds (Hjelmfelt et al. 1982). It might be inferred that the curve number model is an infiltration model because in its application it is used to determine runoff incrementally over the duration of the storm for input into a unit hydrograph model. With this use it becomes a surrogate for an infiltration model. This has created much confusion. The model doesn’t behave like most infiltration models/equations because the curve number losses don’t always decline with time or with prior rainfall and may actually increase when rainfall intensity increases (Hawkins 1980). The curve number model behavior in this regard is the same as from a partial area saturation model. Additionally, with the curve number model, the ultimate steady state rate is zero. It does not approach a steady-state non-zero infiltration rate with time, as do the Horton or Green- Curve Number update user for runoff calculation 37 Hastings, Nebraska WS44028 (1941–1954) 5 CN(III) = 94 3 CN(II) = 85 Q(in) 4 2 CN(I) = 70 1 0 0 1 2 3 4 5 P(in) FIGURE 1. Rainfall-Runoff scatter diagram showing ARC I, II, and III curve numbers FIGURE 2. The 10% and 90% exceedance values compared to the AMC I and AMC III curve numbers Ampt Equations. There doesn’t seem to be any consensus as to whether this is the way watershed losses occur. It should be emphasized that the curve number model is not a point infiltration model and the difference between rainfall and runoff is better defined as watershed “losses.” Watershed indices, such as the curve number, are lumped expressions of net basin performance. In this regard, the runoff curve number model performs well as an integrator of 38 D.E. Woodward et al. all the losses from all the processes over the watershed, which was the original intention. It is also sometimes inferred that the parameter S, defined as the potential maximum retention, is a physical property of the site like a soil moisture storage parameter, and the water in it can be accounted for. This is has not been shown with any certainty. The parameter S (or curve number) is a model variable and is only constant for a particular storm. Although it is related to soil and cover characteristics, it is not an identifiable physical property. VARIATION OF CURVE NUMBER WITH SEASON AND LAND USE Curve numbers were derived from rainfall-runoff data for 15 distinct land uses on 177 small watersheds in the United States (Rietz and Hawkins 2000). Curve numbers for each land use on each watershed were calculated using the asymptotic method and evaluated at the local, regional and national scale. Significant differences at the 5% level were found between the curve numbers of almost all of the different land uses tested. Significant differences in curve number were also found on grazed and ungrazed paired watersheds, and on watersheds that had undergone land use conversions. The general magnitudes and rank order of the average land use curve numbers were in general agreement with expected handbook values. Meadows almost always produced the lowest curve number at both the local and regional level. Forestland produced the lowest overall average curve number at the national level, but also displayed the largest variability. No significant differences could be determined between curve numbers for pasture and rangeland at the regional scale or between row crops and small grain at any scale. Where comparable, pastures usually had higher curve numbers than meadows. None of these comparisons considered hydrologic soil group or any other soil parameter. Seasonal variation of curve number has also been noted. This is seen more readily in the more humid settings, and is rare in arid and semi-arid watersheds. Where evident it follows a pattern with higher curve numbers in the dormant season when the ground has less cover and is likely to be wetter; and lower curve numbers during the summer when the ground is dryer and vegetation is in a high growth stage. Also, seasonal variation in forest curve numbers may be associated with leafing stages (Price, 1998). THE INITIAL ABSTRACTION RATIO (IA/S RATIO) The Initial Abstraction Ratio (Ia/S or λ) was assigned a value of 0.2 in the original development of runoff curve numbers. Data from NEH-630 (1985) has been used to support the value of 0.2. As part of the effort to develop the required documentation for the curve number study, it became apparent that additional review of λ would be appropriate. Data analysis from NEH-630 (1985) indicated the statistical average (mean) of the plotted values was about 0.05. Curve Number update user for runoff calculation Two techniques, Event Analysis and Model Fitting, were used for determining Ia/S from field data sets. Event Analysis. Event analysis requires concurrent synchronized breakpoint records of both rainfall and runoff depth. The event rainfall depth recorded when the direct runoff hydrograph begins is taken as Ia. Knowing the total event rainfall P and the direct runoff Q, equation 1a is solved for S, and the ratio simply taken Ia/S = λ. Each event gives a separate value for λ, and the median for a large number of events is taken as the representative watershed value. General Model Fitting: Here the value of λ is simply determined by iterative least squares procedure fitting for both λ and S of the general equation. Q = (P – λS)2/(P + (1 – λ)S) 39 for P ≤ λS Q=0 The objective of the fitting is to find the values of λ and S such that Σ{Q – [(P – λS)2/(P + (1 – λ)S)]}2 for P > λS is a minimum. Here each P:Q data set gives only one value of λ. An illustration of such fitting is given in Figure 3. In each of the above two methods, only larger storms were used. This was done to avoid the biasing effects of small storms towards high curve numbers. With Event Analysis, only events with Pe = P – λS ≥ 1 inch were used. With Model Fitting, only events with P ≥ 1 inch were used. As shall be seen, resulting values of Ia were often quite small, so that this difference between the two techniques was slight. For statistical analysis, only watersheds for P ≥ λS ARS W S26030 Coshocton, Ohio 6 Direct Runoff Q (inch) 5 4 Q= P Ordered 3 Natural 2 1 0 0 1 2 3 4 5 6 Rainfall P (inch) FIGURE 3. Model Fitting by least squares for WS26030 located at Coshocton, OH with a drainage area of 303 acres. For the natural data (squares): S = 4.0974 inches, CN = 70.8, λ = 0.0179, R2 = 50.50%, and SE = 0.32 inch. For the ordered data (triangles): S = 2.0943 inches, CN = 82.6, λ = 0.1364, R2 = = 99.17%, and SE = 0.0372 inches 40 D.E. Woodward et al. with more than 20 events with P ≥ 1 inch or Pe ≥ 1 inch were used. In addition, for the model fitting determinations, both “natural” and “ordered” data sets were used. Natural data pairs the P and Q as they naturally occurred in time, and thus displays considerable variety in runoff with rainfall. Ordered data matches (usually) rank-ordered P and Q values, so that each has approximately the same return period. This is in keeping with a major application of the method, which in design work at least - matches the frequency of the rainfall with the frequency of the runoff. For example, the 100-year rainfall is assumed to produce the 100-year runoff. Rainfall-runoff data from 307 watersheds or plots were used, originating from USDA-Agricultural Research Service, US Forest Service, US Geological Survey, and New Mexico State University. The data sets covered 23 states, mainly in the east, Midwest, and south of the United States. There was no data from the northwestern 1/3 of the country, from roughly California to Minnesota. A total of 28,301 events were available that met the rainfall depth (P and Pe) criteria. For event analysis, only ARS data was applicable, insofar as it alone contained the needed detailed in- storm break-point information. All others were only rainfall and runoff depths P and Q. This is summarized in Table 1. These 307 watersheds all had 20 or more events which met the storm size criteria. The ARS data is available from Web site ftp://hydrolab.arsusda.gov/ pub/arswater/. The “USLE” plot data had been used in the development of the Universal Soil Loss Equation, and was downloaded from the web site: http:// Forest topsoil.nserl.purdue.edu/usle/. Service data was in large part supplied in reduced form to Hawkins (1980, 1993) by Hewlett (1977, 1984), who used it in earlier papers (Hewlett, et al., 1977; Hewlett and Fortson, 1984). The Jornada plot data, from a site north of Las Cruces NM, was supplied by Ward and described in Hawkins and Ward (1998). The USGS data was supplied from local sources for a number of urban and urbanizing watersheds in the Tucson area. In general, the results showed that λ is not a constant from storm to storm, or watershed to watershed, and that the assumption of λ = 0.20 is unusually high. In 202 out of the 307 watersheds studied, Ia/S = 0.05 had a lower standard error in predicted Q that Ia/S = 0.2. ( Jiang 2001) TABLE 1. Data sets and sources Data source ARS USLE (ARS) USFS Jornada (NMSU) USGS # Watersheds (w) or plots (p) 134 (w) 137 (p) 26 (w) 6 (p) 4 (w) Method used Event Analysis, Model Fitting Model Fitting Model Fitting Model Fitting Model Fitting Curve Number update user for runoff calculation CONCLUSIONS 1. Additional work is needed to develop a revised set of curve numbers.. An up-to-date soils maps needs to be prepared for each of the experimental watersheds used in the study. NRCS is reviewing the impacts of changing the Ia/S ratio. 2. The use of curve numbers as a moisture counting technique needs additional study. While it appears to work, additional documentation needs to be developed. REFERENCES Agricultural Research Service Water Data Center, 1995: ARS Water Data: ARS/ Access CD. USDA-ARS Hydrology Lab, Beltsville, Maryland. Agricultural Research Service Water Data Center, http://www.hydrolab.arsusda.gov/ arswater.html. HAWKINS R.H. 1980: Infiltration and Curve Numbers: Some Pragmatic and Theoretic Relationships”, Proceedings of Symposium on Watershed Management 1980, ASCE, 925–37. HAWKINS R.H. 1993: “Asymptotic determination of runoff curve numbers from data”. Journal of Irrigation and Drainage Engineering. Amer Soc Civ Eng. 119(2): 334–345. HAWKINS R.H., WARD T.J. 1998: “Site and cover effects on event runoff, Jornada Experimental Range, New Mexico”. Proceedings from American Water Resource Association Conference on Rangeland Management and Water Resources. Reno, NV, 361–370. HAWKINS R.H., CATE A. 1998: “Secondary Influences in Curve Number Rainfall Runoff”, Proceeding of the International Water Resources Engineering Conference ASCE. 41 HEWLETT J.D., CUNNINGHAM. G.B., TROENDLE C.A. 1977: “Predicting Storm Flow and Peak Flow from small basins in humid areas by the R-index method”. Water Resources Bulletin. 13(2): 231–253. HEWLETT J.D., FORTSON J.C. 1984: “Additional tests on the effect of rainfall intensity on storm flow and peak flow from wild-land basins”. Water Resources Research. 20(7): 985–989. HJELMFELT A.T. 1980: “Empirical investigation of curve number technique”. Journal of the Hydraulics Division. American Society Civil Eng 106 (HY9): 1471–1476. HJELMFELT A.T., KRAMER L.A., BURWELL R.E. 1982: “Curve Numbers as Random Variables, Rainfall-Runoff Relationship´ Resources Publications, Littleton, CO., 365–370. HJELMFELT A.T., WOODWARD D.A., CONAWAY G., QUAN Q.D., Van MULLEM J., HAWKINS R.H. 2001: “Curve Numbers, Recent Developments”. XXIX IAHR Congress Proceedings, Beijing, China. JIANG R. 2001: Investigation of Runoff Curve Number Initial Abstraction Ratio. MS thesis, Watershed Management, University of Arizona, 120 pp. PRICE M. 1998: “Seasonal Variation in Runoff Curve Numbers”. MS Thesis, Watershed Management, University of Arizona. 189 pp. RIETZ P.D., HAWKINS R.H. 2000: “Effects of land use on runoff curve numbers”. Watershed Management 2000, American. Society. Civil Engineers. Proceedings Watershed Management Symposium, Fort Collins CO. (CD ROM) Soil Conservation Service 1985: National Engineering Handbook, Section 4, Hydrology (NEH-630). USDA/USLE data web site: http://topsoil. nserl.purdue.edu/usle/. USDA, Agricultural Research Service: ftp:// hydrolab.arsusda.gov/pub/arswater/. 42 D.E. Woodward et al. Van MULLEM J.A. (1992). “ Soil Moisture and Runoff–Another Look”. ASCE Water Forum ‘92, Proceedings of the Irrigation and Drainage Session, Baltimore, MD. Streszczenie: Aktualizacja parametru CN metody obliczania opadu efektywnego. W drugiej połowie lat 50. Służba Ochrony Gleb (ang. Soil Conservenation Sernice ) w USA (obecnie Służba Ochrony Zasobów Naturalnymi ang. Natural Resources Conservation Service – NRCS) opracowała procedurę nazywaną metodą Curve Number do wyznaczania opadu efektywnego dla zlewni nieobserwowanych. Procedura ta jest sto- sowana na całym świecie do wyznaczania opadu efektywnego ze zdarzeń opadowych oraz częściej w modelach do symulacji ciągłych. W 1990 roku zespół z NRCS i Służby Badań Rolniczych (ang. Agricultural Research Service) rozpoczął działania nad aktualizacją procedury określania opadu efektywnego. W artykule przedstawiono wyniki przeprowadzonych prac. MS received July 2006 Authors address: Donald E. Woodward Natural Resources Conservation Service, USA e-mail: [email protected] Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 43–54 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Assessment of hydrologic regime changes induced by the Jeziorsko dam performance and morphodynamic processes in the Warta river TOMASZ DYSARZ, JOANNA WICHER-DYSARZ Department of Hydraulic Engineering, Agricultural University of Poznań, Poland Abstract: Assessment of hydrologic regime changes induced by the Jeziorsko dam performance and morphodynamic processes in the Warta river. In the paper the overview of morphological processes occurring in the selected Warta river reach is given. The processes were induced by the performance of the Jeziorsko dam, which was built in 1986. The main purpose of the presented research is the assessment of these processes impact on the hydrologic conditions in the river system. First time the Jeziorsko reservoir it was filled up to the maximum water level in 1991. In 1995 the hydropower plant was put into operation. The sediment particles transported with flowing water were accumulated in the inlet part of the reservoir. In this area very convenient conditions for wild life, especially for water plants and birds, appeared as the result of water stages increase. However, the risk related to seasonal floods and dike break possibility have threaten the local societies until now. The historical data and hydrodynamic simulation model are used for the assessment of the hydrologic changes. The data consists of field measurements of the river morphology and hydrologic data from the selected river gages. The hydrodynamic model was used to reconstruct the flow conditions in particular states of the river system evolution. The statistical analysis of the obtained results enabled the assessment of the hydrologic characteristics transformation in the selected Warta river reach. Key words: hydrodynamic, simulation model, river morphology, river hydrology. INTRODUCTION In the paper the overview of morphological processes occurring in the selected Warta river reach is given. The processes were directly or indirectly induced by the performance of the Jeziorsko dam, which was built in 1986. The main purpose of the presented research is the assessment of these processes impact on the hydrologic conditions in the river system. Since the dam was built the velocity characteristics in the analyzed Warta river reach were changed what induced the deposition of the transported sediment in the inlet part of the reservoir. Next the sediment accumulation in the river channel induced one more process affecting hydrologic regime. Due to the increase of floodplain inundation frequency, the luxuriant vegetation start to grow there. These two processes, sediment accumulation and vegetation growth, cause additional increase of floodplain inundation. The processes induced by the dam built and its performance seems to be self-driven cycle causing irreversible alterations of hydrologic regime. The problem under consideration draws attention of many researchers and engineers today. The relationships between hydrologic regime conditions and the riparian vegetation are crucially important for the robust river management. The examples of such studies were given by Keeland et al. (1997), Surtevant (1998), Stromberg (2001), Nilsson and 44 T. Dysarz, J. Wicher-Dysarz Svedmark (2002) and many others. Amongst many anthropogenic actions in the water environment the damming of rivers plays special role due to complexity of induced ecological responses. One of the most interesting aspects of these responses is the dam impact on the upstream hydraulic conditions and related changes of hydrologic regime. The example of studies on this aspect was given by Mumba and Thompson (2005). They analyzed the changes in the floodplains located along the Kafue river (Zambia) reach between two dams, Itezhi-tezhi (upstream) and Kafue Gorge (downstream). The Authors proved that dams totally changed the hydrological and ecological conditions in the region. Almost permanent inundation of the area located between the dams was the direct reason. Such conditions were imposed by parallel influences of releases from upper dam and high water levels in the lower reservoir. The methodology used for assessment of hydrologic regime changes ranges from simple approach based on equal discharge studies (Pinter and Heine 2005) to complex analysis including studies of discharges and water table frequency (e.g. King et al. 1998, Maingi and Marsh 2002, Magilligan and Nislow 2005, Wellmeyer et al. 2005). This is also quite common to simulate the behavior of the water system by means of mathematical model (e.g. Kite 2001, Maingi and Marsh 2002). In each case the methodology applied should match the specific features of the system, processes occurring there as well as data availability. The main purpose of the presented paper is to present the hydrological regime changes in the Warta river reach located upstream of the Jeziorsko reservoir. According to the performed observations the hydrologic conditions are caused by three factors. The primary reason is water management in the reservoir. Two next elements are (1) sediment accumulation in the river and (2) vegetation growth. The field measurement data and hydrodynamic model are used for our studies. Steady and unsteady flow simulation were run to reconstruct the flow conditions in the system before and after the analyzed changes. The final results are presented in two way. First method consists of water stages in the selected river cross-section related to some flows determined for the inflow gauge station. The chosen flows are related to the specific probabilities of excedance, e.g. 1%, 10%, or maximum observed discharge in the system is used. The second method is the statistical analysis of unsteady flow simulations presented as relative and cumulative frequency functions related to the particular stages of river evolution. The paper includes into six main parts, acknowledgments, references and two appendixes. In the first part brief introduction to the investigated problem is given. In the second part the study area and processes occurring there are described. The data collected are briefly presented in third part. Then the basic elements of the applied methodology are presented in fourth sections. In fifth part the obtained results are presented and discusses. Finally the conclusions are formulated in the sixth part. The attached appendixes includes the details of applied methodology. Assessment of hydrologic regime changes induced by the Jeziorsko... CASE STUDY SYSTEM Jeziorsko reservoir was built in 1986 in central part of Poland. It is located in the Warta river, between the Sieradz (upstream) and Uniejów (downstream) gauge stations. The reservoir inlet with Warta river reach are shown in the attached photo map (Fig. 1). The Jeziorsko reservoir is relatively new object. First time it was filled up to the admissible maximum water level (121.50 m a.s.l.) in 1991. The hydropower plant was put into operation in 1995. The Warta catchment area in the inlet of the reservoir is 8450 km2. The 45 average discharge in the nearest Sieradz gauge station is about 45 m3/s, but the discharges variability ranges from 10 m3/s to 440 m3/s (observed in 1997). The wet season during the year is January – May. The lowest discharges are observed during the summer and autumn time, from September to December. The Jeziorsko reservoir is multipurpose reservoir performing in the annual working scheme. During the months from September to December the water stages are kept on the minimum level of 116 m a.s.l. In spring months the reservoir is filled with water. The maximum water level 121.5 m a.s.l. is km 499 km 500 km 501 km 502 km 503 km 504 km 505 km 506 FIGURE 1. Photo-map of Jeziorsko reservoir inlet part and Warta river 46 T. Dysarz, J. Wicher-Dysarz hold during April – June period. In the period from July to August the water stored at the reservoir is used and the reservoir water stages gradually come back to the minimum level. The main purpose of the reservoir performance is the flood protection during wet season. The reservoir is also the source of water for the big cities in Great-Poland province: Poznań, Koło, Śrem and Konin. The water stored is also used for irrigation purposes and to preserve biological life below the dam. The Pątnów-Konin and Adamów power plants used the water from the reservoir for cooling. Other goal of the Jeziorsko performance is production of hydro-power, though, this is not significant due to the small difference between upper and lower water levels in the dam (about 10 m). The reservoir is also used for recreation and inland fishery. Since the very beginning of the reservoir performance the changes of the Warta river bed and floodplains are observed in the inlet part of the Jeziorsko. Due to decrease of flow velocities there the capacity of sediment transport is lower and material is deposited. The most important processes occur in the river reach below the bridge in Warta town, between km 503 and km 500 (see Fig. 1). The annual scheme of reservoir performance foster the vegetation growth in this area. It enabled to establish a nature reserve for birds in the upper part of the reservoir in 1998. Since this time the important reservoir performance purpose is conservation of water conditions sufficient for biological life in this area. Due to the existence of the nature reserve and Environmental Protection Ministry act valid since 1998, the river engineering actions are not allowed in the reservoir. The purpose of this regulation was the protection of birds habitats and biological life in the upper part of the considered object. The regulation causes also some problems indicated in previous papers, e.g. Wicher (2004), Wicher-Dysarz and Przedwojski (2005). The water levels in the backwater part are gradually increasing. Long-term results of these processes may cause degradation of the river in the inlet part of the reservoir as well as the reservoir capacity decreasing. Second risk is related to destruction of flood protecting and backwater dikes by overtopping. The losses occurring in such case are difficult to overestimate but the probability of this danger seems to increase. However, the Ministry act enabled the presented research in some sense. In other conditions the authorities responsible for the reservoir effectiveness and safe use should take steps preventing from these processes. Now, it is possible to observe the hydrologic regime evolution in the pure form. The only influences are reservoir performance, sediment accumulation and vegetation growth. Hence, the chosen study area properly fits the research purposes. COLLECTED DATA The set of available data consists of (a) system geometry, (b) discharges and water stages in the system, (c) state of the vegetation cover. Assessment of hydrologic regime changes induced by the Jeziorsko... The channel and floodplain elevations are observed for long period of time in the area under consideration. The most basic set of data consists of the river regulation design prepared in 1975 (Matan 1975). The regulated river seemed to be stable and was not subject to any changes until the reservoir started to perform in 1991. Hence, this set of data may be considered as the original terrain and river bottom without any significant influence of the analyzed processes. The another sets of applied data consists of measurements done in 1997 and 2004. After a few years of reservoirs performance the Warta river changes were checked and assessed by the team from Department of Hydraulic Engineering, Agricultural University of Poznań. The measurements results were described by Wicher (2004) and Wicher-Dysarz and Przedwojski (2005). Hence, the system geometry data consists of three sets of river bottom and floodplain measurements in 37 cross-sections covering the river reach from Sieradz gauge station (km 520+850) to Jeziorsko dam (km 486+500). These data were used to describe three stages of the river bottom and floodplain evolution. Other data used in our analyses are discharges and water stages observed in Sieradz gauge station as well as water levels measured at the Jeziorsko dam headwater. The set of data from Sieradz gauge station consists of daily measurements done in periods 1963– –1970, 1973–1983, 1993–1995 and 1997–2001. These are 27 years of hydrological data. Some of the data were measured and published by Institute of Meteorology and Water Management (1963–1970, 1973–1983). Others were collected by Regional Board for Water 47 Management in Poznań. The later data are made available under the circumstances of cooperation between Department of Hydraulic Engineering and Regional Board for Water Management. The data collected from Sieradz gauge station are used to describe the hydrologic condition in the inlet of the analyzed river reach which are not affected by reservoir performance. The water stages at Jeziorsko headwater were observed during the period of reservoir performance, from 1992 to 2001. This set of data was collected by Regional Board for Water Management. The data were used to design the average scenario of reservoir performance during the year. The impact of vegetation on the roughness coefficients was assessed on the basis of measurements including plants localization and vegetation cover density. The stalk thickness was estimated in the same way. The necessary measurements were done by researchers from Department of Hydraulic Engineering during 2004–2005. On the basis of collected data hydromorphological stages of the system evolution is classified as four stages of river bottom. The stages of river bottom enable the design of five stages of hydraulic conditions in the water system. First analyzed state of the system is the primary bottom described in Matan (1975). This geometry was considered as pre-dam conditions. The second geometry was prepared on the basis of measurements done in 1997 (Wicher 2004). This stage of the system correspond to the system state a few years after the beginning of reservoir performance. The roughness in the particular crosssections is the same as in the 1975. It 48 T. Dysarz, J. Wicher-Dysarz does not describe the slight changes of vegetation cover in this period. Hence, the impact of pure sediment deposition may be assessed. The third and fourth geometry of the system was prepared on the basis of measurements done in 2004 and corresponds to current conditions. However, the third scheme does not reflex to the roughness changes due to the vegetation growth. This element is included in the fourth system geometry. Hence, the current state of the system may be assessed in two ways: with and without vegetation growth. The prepared stages of river bottom and hydraulic conditions are as follows: 1) initial bottom reflecting pre-dam conditions (1975) with free outflow in the outlet; 2) initial bottom reflecting pre-dam conditions (1975) with outflow corresponding to the average annual scheme of reservoir performance; 3) bottom measured in 1997 with the initial roughness and outflow as above; 4) bottom measured in 2004 with the initial roughness and outflow as above; 5) bottom measured in 2004 with roughness reflecting current state of vegetation and its seasonal variation, outflow corresponding to the average annual scheme of reservoir performance. In the fifth stage the concept of roughness changes due to the vegetation growth is applied. The riparian vegetation impact on the flow conditions may be modeled as the changes of Manning roughness coefficients. Following Klopstra et al. (1997) concept of velocity changes due to the presence of non- submerged vegetation the roughness coefficient may be expressed as 2 n = h3 Cw md 2g (1) where: g – acceleration of gravity, h – the depth, Cw, m and d – the parameters describe the characteristics of riparian vegetation, Cw – the coefficient reflecting seasonal changes of leaves cover. Its values were taken as Cw = 1.05 for winter period and Cw = 1.40 for summer (Armanini et al. 2005). m is average number of plants in one square meter and d is average diameter of bushes stalks. The presented approach was used to describe the changes of roughness in the floodplain. Parameters m and d were determined on the basis of measurements done in 2004 and 2005. The listed stages of river system were used for the simulation of steady and unsteady flow conditions along the reach under consideration. APPLIED METHODOLOGY The main basis for the assessment presented here is the set of collected data described in previous part of the paper. To estimate the hydrologic regime characteristics related to five stages of river system several deterministic and stochastic techniques are used. The methodology applied consists of three main elements. These are (1) maximum flows probability of excedence, (2) hydrodynamic model simulations for Assessment of hydrologic regime changes induced by the Jeziorsko... steady and unsteady conditions, (3) analysis of relative and cumulative frequencies of water stages in selected river cross-sections. The above elements are described shortly in the following parts of the paper. In the first part of the analysis the maximum flows probability of excedence is elaborated. The curve is determined for the Sieradz gauge station located in the inlet of the system. The curve is shown in Figure 2. Crosses represent the annual maximum flows observed and probabilities of excedance assigned to them according to Weibull formula. The continuous line is the Pearson type III curve estimated by means of quintile method (Ozga-Zielińska and Brzeziński 1994). The estimated curve is used to determine the characteristic flows occurring in the system with some specified probability of excedance. The chosen flows are commonly used for the assessment and design purposes. They are presented in Table 1. 100 observations Pearson III curve probability [%] 80 60 40 20 0 0 100 200 300 400 maximum discharge [m3/s] 500 FIGURE 2. Maximum flows probability of excedance for Sieradz gauge station 49 TABLE 1. Selected flows with specific probabilities of excedance determined for the Sieradz gauge station Probability of excedance 10 % 5% 1% 0.5 % Observed maximum Denotation Value [m3/s] Q10 % Q5 % Q1 % Q0.5 % 245.40 263.28 294.34 304.89 Max Q 440.00 The determined probable flows together with minimum and maximum water stages in the reservoir are used to simulate the steady flow conditions in the analyzed Warta river reach. The computations are done for each of five river bottom stages listed in previous part of the paper. The hydrodynamic model is used in this purpose. The routines included in steady flow module of the classic HEC-RAS package are applied. The main basis of the used methodology is energy balance equation written for the compound channel. The complete description of HEC-RAS package may be found in Brunner (2002). The results of the simulations are changes of water levels in selected river cross-sections related to the maximum flows and stages of river bottom. They are presented and discussed in the next part of the paper. The second part of the performed analyses consists of inflow scenarios, unsteady flow simulations and frequency analysis. The used inflow scenarios are the observed discharges in the Sieradz gauge station. They are combined with the average annual scheme of water management in the Jeziorsko reservoir. The unsteady flow module included in 50 T. Dysarz, J. Wicher-Dysarz HEC-RAS package is used to simulate the behavior of the system related to five stages of the river. Then the obtained results are analyzed by means of water stage relative and cumulative frequencies determined in several river cross-sections. The basic definitions of frequency curves are used. More information may be found in Chow et al. (1988). The selected cross-sections are located between km 503 and km 500, where the most intensive morphological and vegetation changes are observed. The results are presented and discussed in the next part of the paper. elevation [m a.s.l.] 124 Q1% in 2004 123 Max Q in 1975 Q1% in 1975 122 121 120 cross-section distance [m] 119 0 The described analyses are done for the all cross-sections in the river reach between the bridge in Warta town (Fig. 1) and the Jeziorsko reservoir. These are sections located at km 503+560, 503+380, 503+140, 502+800, 502+300, 501+270, 500+970, 500+510 and 499+890. The examples of obtained results are shown in Figures 3–10. Figures 3–6 presents the results related to the first part of analyses done with the help of steady flow module. The rest of mentioned figures consists of relative and cumulative frequency curves elaborated on the basis of unsteady flow simulations. The Figures 3 and 4 present the water stages related to analyzed flows Q1% and Max Q. The water stages are estimated for the pre-dam conditions (denoted as ‘in 1975’) and current conditions (‘in 2004’). These stages are drawn on the original cross-sections contours. The 100 200 300 400 500 600 FIGURE 3. Cross-section shape and characteristic water stages in km 503+560 elevation [m a.s.l.] 124 ANALYSIS OF SIMULATION RESULTS Max Q in 2004 Max Q in 2004 123 Q1% in 2004 122 Q1% in 1975 Max Q in 1975 121 120 cross-section 119 distance [m] 118 0 100 200 300 400 500 600 700 FIGURE 4. Cross-section shape and characteristic water stages in km 503+140 given examples present the results for the cross-sections located in km 503+560 and km 503+140. It is clearly visible that the morphological and vegetation changes in the river system caused crucial increase of water stages in the system. The maximum observed flows should are close to the maximum dikes elevations in the system. It is important to indicate that further changes in this area may cause the dike break by overtopping. Assessment of hydrologic regime changes induced by the Jeziorsko... The increase of water stages related to specific maximum flows is also presented in Figures 5 and 6. In this graphs the estimated changes of water stages related to the flows Q10%, Q1% and Max Q are plotted. The squares, circles and triangles represent values estimated for particular hydraulic conditions. Horizontal axis describes the time. The dotted lines and empty figures reflects the changes of water stages without increase of vegetation cover. It is clearly visible that water stages are raising. The vegetation impact causes dramatic intensification of these processes. 123.5 water stage [m a.s.l.] 51 water stage [m a.s.l.] 123.0 122.7 122.4 Max Q Q1% 122.1 Q10% year 121.8 1975 1980 1985 1990 1995 2000 2005 FIGURE 6. Changes of the water stages related to characteristic flows in cross-section km 503+140 124 water surface elevation [m a.s.l.] 123.2 123 122.9 effect of impacts, bottom 2004 Max Q 122.6 122 Q1% 122.3 Q10% floodplain minimum 121 year 122.0 1975 1980 1985 1990 1995 2000 free outflow, bottom 1975 2005 FIGURE 5. Changes of the water stages related to characteristic flows in cross-section km 503+560 In the Figures presenting relative frequency functions (Figs. 7 and 9) the estimated initial stage of hydrologic regime is compared with the estimated current conditions. The initial stage is represented by Pearson type III probability density function fitted to the results of simulation. In these figures the minimum floodplain elevations are denoted as dashed line. The Figures 8 and 10 shows the gradual changes of cumulative frequency functions 120 0.00 0.02 0.04 0.06 relative frequency 0.08 0.10 FIGURE 7. The relative frequency function corresponding to the initial and current state of the hydrologi-cal regime in cross section located at km 503+560 related to analyzed stages of the hydromorphological conditions in the river. Presented results shows two common trends in cross-sections located between km 503 and 501. The most frequent stages become higher and their relative frequency is greater. This means greater frequency of floodplain inundation, what favors the riparian vegetation growth. 52 T. Dysarz, J. Wicher-Dysarz 124 123 water surface elevation [m a.s.l.] 123 water surface elevation [m a.s.l.] bottom 2004 +vegetation bottom 2004 +vegetation 122 bottom 2004 bottom 2004 bottom 1997 122 bottom 1997 121 free outflow, bottom 1975 121 free outflow, bottom 1975 cumulative frequency bottom 1975 120 0.20 0.40 0.60 0.80 1.00 FIGURE 8. The cumulative frequency function corresponding to the analyzed states of the hydrological regime in cross section located at km 503+560 123 120 0.00 0.00 water surface elevation [m a.s.l.] effect of impacts, bottom 2004 122 121 floodplain minimum free outflow, bottom 1975 relative frequency 120 0.00 0.04 0.08 0.12 FIGURE 9. The relative frequency function corresponding to the initial and current state of the hydrologi-cal regime in cross section located at km 503+14 CONCLUSIONS The obtained results shows that potential hydrologic regime in the analyzed area is significantly changed. The impact of each single process as well as the cumulated impact of all processes have cumulative frequency bottom 1975 0.20 0.40 0.60 0.80 1.00 FIGURE 10. The cumulative frequency function corresponding to the analyzed states of the hydrological regime in cross section located at km 503+140 one common trend. The higher water stages become more frequent and their frequency is greater. This is well visible in Figures 7–10. The changes are crucial even in the cross-section located in the longest distance from reservoir (km 503+560). The areas located along river reach were not flooded frequently there (Fig. 7). Now, the inundation is almost permanent. The risk related to the dike break by overtopping is raising, too. This aspect of the hydrologic regime changes is clearly visible in Figures 3–6. The channel capacity was huge enough to pass flows as Q1% and Max Q in the past. Now, the changes of channel geometry and roughness cause huge resistance and increase of water stages. It is quite easy to predict, that further changes will result in natural catastrophe and damages. The presented results shows that the probability of such an event dramatically increases. Assessment of hydrologic regime changes induced by the Jeziorsko... ACKNOWLEDGEMENT This work was supported by Polish Committee for Scientific Research under grant “Analysis of morphodynamic and vegetation impacts on the hydraulic conditions in backwater part of lowland reservoirs”, project no. 2 P06S 034 30. REFERENCES ABBOTT M.B. 1979: Computational Hydraulics, Elements of the Theory of Free Surface Flows. London: Pitman. ARMANINI A., RIGHETI M. & GRISENTI P. 2005: Direct measurement of vegetation resistance in prototype scale. Journal of Hydraulic Engineering, 43 (5): 481–487. BARKAU R.L. 1982. Simulation of the July 1981 flood along Salt River, Report for CE695BV, Special Problems in Hydraulics. Fort Collins, CO.: Department of Civil Engineering, Colorado State University. BRUNNER G.W. 2002: HEC-RAS, River Analysis System hydraulic reference manual, computer program documentation. Davis, CA.: US Army Corps of Engineers, Hydrologic Engineering Center. CHOW V.T., Maidment D.R. & Mays L.W. 1988: Applied Hydrology. McGraw-Hill International Editions, Civil Engineering Series, McGraw-Hill Book Company. CUNGE J.A., HOLLY F.M.Jr & VERVEJ A. 1980: Practical aspects of computational river hydraulics, Pitman Advanced Publishing Program. KEELAND B.D., CONNER W.H. & SHARITZ R.R. 1997: A comparison of wetland tree growth response to hydrologic regime in Louisiana and South Carolina. Forest Ecology and Management, 90: 237–250. KING S.L., ALLEN J.A. & McCOY J.W. 1998: Long-term effects of a lock and dam and greentree reservoir management 53 on a bottomland hardwood forest. Forest Ecology and Management, 112: 213–226. KITE G. 2001: Modelling the Mekong: hydrological simulation for environmental impact studies. Journal of Hydrology, 253: 1–13. KLOPSTRA D., BARNEVELD H.J., NOORTWIJK J.M. von & VELZEN E.H. von 1997: Analytical model for hydraulic roughness of submerged vegetation. The 27th Congress of the International Association for Hydraulic Research, Proceedings of Theme A, Managing Water: Coping with Scarcity and Abundance, San Francisco, California. LIGGETT M.B. & CUNGE J.A. 1975: Numerical methods of solution of the unsteady flow equations, in Mahmood and Yevjevich (1975). MAGILLIGAN F.J. & NISLOW K.H. 2005: Changes in hydrologic regime by dams. Geomorphology, 71: 61–78. MAHMOOD K. & YEVJEVICH V. (eds.) 1975: Unsteady flow in open channels. Fort Collins, Colorado: Water Resources Publications. MAINGI J.K. & MARSH S.E. 2002: Quantifying hydrologic impacts following dam construction along the Tana River. Kenya. Journal of Arid Environments, 50: 53–79. MATAN J. 1975: Jeziorsko reservoir in the Warta river, Hydro-engineering and land improvements structures in the reservoir backwater. Poznań: Hydroprojekt (in Polish). MUMBA M. & THOMPSON J.R. 2005: Hydrological and ecological impacts of dams on the Kafue Flats floodplain system, southern Zambia. Physics and Chemistry of the Earth, 30: 442–447. NILSSON C. & SVEDMARK M. 2002: Basic principles and ecological consequences of changing water regimes: riparian plant communities. Environmental Management, 30 (4): 468–480. OZGA-ZIELIŃSKA M., BRZEZIŃSKI J. 1994: Applied hydrology. Scientific Publishing PWN, Warsaw (in Polish). 54 T. Dysarz, J. Wicher-Dysarz PINTER N. & HEINE R.A. 2005: Hydrodynamic and morphodynamic response to river engineering documented by fixed-discharge analysis, Lower Missouri River, USA. Journal of Hydrology, 302: 70–91. STROMBERG J.C. 2001: Restoration of riparian vegetation in the south-western United States: importance of flow regimes and fluvial dynamism. Journal of Arid Environments, 49: 17–34. SURTEVANT B.R. 1998: A model of wetland vegetation dynamics in simulated beaver impoundments, Ecological Modeling, 112: 195–225. WELLMEYER J.L., SLATTERY M.C. & PHILLIPS J.D. 2005: Quantifying downstream impacts of impoundment on flow regime and channel planform, lower Trinity River, Texas. Geomorphology, 69: 1–13. WICHER J. 2004: Sediment accumulation in lowland reservoirs. Ph.D. thesis in the Agricultural University of Poznan (in Polish). WICHER-DYSARZ J., PRZEDWOJSKI B. 2005: Modeling of sediment accumulation in the inlet part of Jeziorsko reservoir, Annual Reviews of Agricultural University of Poznan, 26: 483–493. Streszczenie: Ocena zmian warunków hydrologicznych wywołanych działaniem zapory Jeziorsko oraz procesami morfodynamicznymi w rzece Warcie. W artykule zostały opisane procesy morfodynamiczne zachodzące na wybranym odcinku rzeki Warty. Zjawiska te zostały wywołane działaniem zapory Jeziorsko wybudowanej w 1986 roku. Głównym celem badań jest oszacowanie wpływu tych procesów na aktualne warunki hydrologiczne panujące w rzece. Pierwsze napełnienie zbiornika do poziomu maksymalnego piętrzenia odbyło się w 1991 roku. W 1995 roku została uruchomiona elektrownia wodna „Jeziorsko”. W wyniku przerwania ciągłości transportu rumowiska, w górnej części zbiornika Jeziorsko nastąpiło akumulacja transportowanych przez rzekę cząstek rumowiska. W efekcie osadzania się rumowiska w górnej części zbiornika, zaczęła intensywnie rozwijać się roślinność oraz ptactwo wodne, które znalazło idealne warunki do rozwoju. Czynniki te spowodowały spiętrzenie wody w tej części zbiornika. Dane historyczne oraz model hydrodynamiczny zostały wykorzystane, aby oszacować zmiany warunków hydrologicznych. Zestaw wykorzystanych danych zawiera: dane z pomiarów terenowych geometrii i morfologii koryta oraz pomiary stanów i przepływów z wybranych wodowskazów. Model hydrodynamiczny został wykorzystany do odtworzenia warunków przepływu w wybranych fazach zmian systemu wodnego. Analiza statystyczna wyników symulacji umożliwiła ocenę zmian charakterystyk hydrologicznych na wybranym odcinku rzeki Warty. MS. received November 2006 Authors’ address: Tomasz Dysarz, Joanna Wicher-Dysarz e-mail: [email protected] Akademia Rolnicza w Poznaniu ul. Wojska Polskiego 73a 60-625 Poznań Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 55–67 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Characteristics of the particulate matter PM10 concentration field and an attempt to determine the sources of air pollution in the living district of Ursynów GRZEGORZ MAJEWSKI, WIESŁAWA PRZEWOŹNICZUK Warsaw Agricultural University – SGGW Department of Water Engineering and Environmental Restoration Abstract: Characteristics of the particulate matter PM10 concentration field and an attempt to determine the sources of air pollution in the living district of Ursynów. The paper presents characteristics of the imission field of PM10 particulate matter with the use of basic statistical data such as mean values, variability ranges, numbers of exceedences of the allowable concentration, frequency distributions of concentration occurrence in individual periods, and time courses. Mean values of basic meteorological elements were calculated, circular graphs for percentiles of dust concentration were prepared, and percentiles of pollution plume rates were made in order to analyse meteorological conditions of dispersal and spreading out of air pollution. A synthetic WZ indicator of pollution has also been calculated that encompasses the joint impact of various meteorological elements on pollution concentration levels. Key words: particulate matter PM10, pollution plume rate, meteorological conditions. INTRODUCTION First, identification of areas where allowable standards of pollution concentration are exceeded, and second, identification of pollution sources being the threat to the environment are basic tasks for the atmospheric air quality monitoring system. The knowledge on the concentration field and emission sources is a basis for setting correction programmes necessary for the improvement of environmental conditions. The order by Minister of Environment as of 5 July 2002 (DzU, No 115, pos. 1003) settles in detail requirements which the air protection programs should answer to. It is said inside that act, among others, that apart of assessing the imission stated on a given area, the inflow of pollution on that area should also be estimated. Discrimination between the emission influence and the impact of meteorological conditions on the rate of pollution concentration registered (Walczewski et al. 2000) is extremely important for proper execution of all tasks of environment protection services, because the increase or the decrease of pollution concentration values depend on both emission and dispersion conditions. The meteorological characteristics of the measurement period makes possible to evaluate properly the imission situation. Therefore, basic meteorological data should be taken into account at the interpretation of imission measurement results, and a detailed analysis of the wind rose from the measurement period would allow conclude on whether the measured concentrations came either from local emission sources or are rather 56 G. Majewski, W. Przewoźniczuk caused by an inflow of pollution from outside the measuring network. The living quarter of Ursynow is one of the three districts of Warsaw, the Polish capital, for which the air protection correctional programs were made. That program was prepared because the allowable level of PM10 particulate matter had been exceeded (Order No 62). The particulate matter is a commonly occurring pollution matter, the concentrations of which maintain in urban agglomerations at a rather high level exceeding frequently allowable values. Miniparticles dispersed in the air are considered to be one of more essential potential threats to human health as related to air pollution (i.a. Juda & Chróściel 1974, Stern 1994, Dockery 1996, Quarg 1996, Niecko et al. 1998, Warych 1999, Jabłońska 2003). It is assessed that about one third of the population in Poland exposed to inhalation of dust pollution is bound to chronic diseases of the breathing system, that leads to circulatory system and cancer diseases (Warych 1999). Ursynów is a district of Warsaw, capital of Poland, situated almost entirely on the Warsaw Lowland, elevated 20–30 m up of the level of water in the Vistula River, and it encompasses the southern regions of the Warsaw city. It covers the area of 44.6 km2 constituting 8.6% of the Warsaw city total area, being at the third place as for the size within the city. Three nature reserves occur within the quarter mentioned; they are: the Natolin Wood, the Ursynów Slope – being the main topographic element shaping the local climate of the region mentioned, and the Kabaty Wood called the Stefan Starzyński Wood – a forest tract over 900 ha in size lying within the Warsaw region; the latter wood is the most essential natural element of the southern aeration system for the capital. The industrial plants and structures equipped in their own boiler-rooms influencing the air quality on the area of the Ursynów quarter are as follows: Oncology Centre, Warsaw Plant of Aggregate Exploitation, Midas Polska Ltd, Volvo Auto Polska, Underground Standstill Station, PBM Południe Co, Radio Ceramics Plant, Stolbud Ltd, Geant Hypermarket, Unikon-Beton, Tesco Hypermarket, and others. The streets with the greatest intensity of car traffic in this quarter are: Pulawska, KEN Avenue, Rodowicza Anody, Rosoła, Płaskowickiej, Ciszewskiego, Roentgena, and Poleczki street. The areas with the occurrence of low communalliving emission are as follows: ‘Green Ursynów’ along the Puławska Avenue, borders of the Kabacki Wood from the Zolna street up to the city boundary, the western side of the Puławska Avenue, the area encompassing the streets of Krasnopolska, Baletowa, Karczunkowska, Sarabandy, Farbiarska and Gawota, the area along the Ursynów Slope, and finally the area along the Prawdziwka street. MATERIAL AND METHODS An attempt has been undertaken in this paper to assess the imission situation in the Ursynow quarter, in the MzWarszUrsynow station area, as well as an attempt to point out the reasons for the existing state. The data for calculations Characteristics of the particulate matter PM10 concentration... were taken from the automatic station of atmospheric air monitoring called the MzWarszUrsynow station (φN = = 52o09’39’’, λE=21o02’03’’, 102 m a.s.l.) working within the regional network of the Masovia Province (WIOS database). Measurements of SO2, NO, NO2, NOx, CO, and PM10 concentrations were made at that station, as well as measurements of basic meteorological elements. The calculations made from measurement data as averages in onehour intervals concerned the period Oct. 2003 – Sept. 2005 that was divided into winter half-years (the period from October to March, covering heating 57 seasons) and summer half-years (the period from April to September) . The MzWarszUrsynów measuring station under discussion is representative for the general urban ambience and it characterises well imission of pollution dust in the area of living quarters exposed to the impact of traffic and communal and industrial emissions. Figure 1 illustrates the location of the station. The concentration of PM10 particulate matter is measured there in continuously with MLU TEOM1400 analyser using the oscillation microweight method. The principle of the TEOM function consists in measuring the frequency of vibration FIGURE 1. The Air Pollution Monitoring Stations in Ursynów 58 G. Majewski, W. Przewoźniczuk of a conical element on which the filter is fixed. The increase of the filter mass due to deposition of dust particles on it causes the change in the frequency of measured vibration. This method ensures the detectability threshold below 0.06 [μg·m–3] for mean one-hour values at the throughflow 3 [l⋅min–1]. The device has got the certificate of conformity with the PM10 measurement relay method according to the European standard (EN12341). The following calculations were the basis for characterising the concentration field of PM10 particulate matter: 1) mean concentration of dust in calculation periods, 2) number of exceedences of the allowable concentration, 3) variability ranges of concentration values in individual calculation periods, 4) concentration histograms based on the Air Quality Scale rates recommended by WIOS, 5) percentiles 25, 50, 70, and 98 of concentration values, 6) mean annual 24-hour courses of dust concentration values. The data registered at MzWarszUrsynów and MzWarszSGGW stations were used for evaluation of meteorological conditions of pollution dispersal. Both those stations laid 800 m apart from each other (correlation coefficient between temperatures measured on both stations is 0,99) are equipped in the automatic outfit registering basic meteorological elements in 10-minute periods. Measurements of air temperature, wind speed, and air humidity at several levels are made at the MzWarszSGGW station. The air temperature values from heights 5 cm and 22 m were used to calculate gradients. The analysis of meteorological conditions of pollution dispersal and spreading out covered the following points: 1. Setting mean monthly values of basic meteorological elements (Table 1), 2. Comparing the course of air temperature in calculation periods with the mean long term course, circular graphs for 3. Making percentiles of particulate matter concentration as related to air inflow directions, circular graphs for 4. Making percentiles of pollution plume rate in order to get additional information on location of emission sources, synthetic WZ 5. Calculating indicator of pollution for assessing meteorological conditions of pollution dispersal. The vicinity of the station was divided into equal sectors with the central α angle constituting 1/16 part of the full angle in order to prepare circular graphs for percentiles of particulate matter concentration. The percentiles of the order p = 25, 50, 70, 98 were calculated basing on the sets of dust concentration values occurring at the air inflow from the direction related to the given sector α. Circular graphs were made for percentiles calculated for all sectors of wind directions; they are graphical presentations of concentration value distributions at various wind directions. Pollution flows were analysed for to get more information on directions of dust inflows over the area under study (Klis, Matejczyk 2002). The pollution flow is a vector. The scalar size of this Calendar year 2005 2004 2003 Mean monthly cocentration of PM10 [μg·m-3] 34,7 41,9 37,1 42,5 33,3 40,5 42,0 27,5 28,7 28,5 34,9 38,7 38,4 29,1 41,2 29,7 43,5 36,3 46,7 28,1 28,5 29,0 34,6 47,3 Month X XI XII I II III IV V VI VII VIII IX X XI XII I II III IV V VI VII VIII IX Mean air temperature [0C] 5,5 5,1 1,0 -4,9 0,0 3,7 8,9 12,2 16,0 18,1 19,5 14,0 10,2 3,4 1,7 0,8 -3,1 0,2 9,5 14,0 16,4 20,8 18,2 16,2 Mean wind speed [m·s-1] 1,5 1,4 1,9 1,7 1,8 1,6 1,2 1,1 1,0 1,1 1,2 1,3 1,5 1,7 1,6 1,9 1,7 1,6 1,3 1,2 1,1 1,0 1,0 0,8 Mean air humidity [%] 84,3 90,5 88,1 88,5 83,4 78,4 63,7 69,8 66,2 70,3 67,6 71,0 81,5 89,1 90,5 84,5 84,8 76,4 59,5 66,6 64,6 58,8 62,9 62,9 Precipitation [mm] 67,2 20,7 41,7 30,9 34,6 40,9 52,2 54,9 42,3 61,3 53,7 20,1 39,7 62,1 11 33,1 27,2 40 18,8 56 45,3 4,7 38,8 33,6 181 180 182 181 6 7 7 10 7 9 10 1 7 7 2 7 4 7 6 12 4 5 14 9,0 7,7 33 35 9,9 5,2 Min. S24 [μg·m-3] 18 46 Mean concentration PM10 [μg·m-3] Date of max. Max. S24 [μg·m-3] 98,0 03.04 35,7 119,2 11.12 36,3 75,2 16.04 33,3 156,4 29.01 38,3 Standard deviation [μg·m-3] 17,2 20,7 12,3 22,7 The 75th percentile [μg·m-3] The 50th percentile [μg·m-3] Percentile of 1-hour The 50th percentile [μg·m-3] The 25th percentile [μg·m-3] The 98th percentile [μg·m-3] The 25th percentile [μg·m-3] 23,3 30,6 44,5 79,9 19,0 29,5 46,4 102,5 21,8 31,5 46,5 100,5 18,9 29,8 46,7 116,2 24,0 31,2 41,7 62,0 19,5 29,1 42,4 93,9 21,2 35,6 50,5 97,7 18,2 32,5 52,0 114,8 Percentile of 24-hour The 75th percentile [μg·m-3] The number of exceedance of D24 TABLE 1. The selected statistical characteristics of PM10 concentration and meteorological elements from October 2003 to September 2005, station MzWarszUrsynów Winter, summer half-year period winter half-year period winter half-year period summer half-year period The number of concentrationS24 summer half-year period The 98th percentile [μg·m-3] 60 G. Majewski, W. Przewoźniczuk vector is equal to the amount of pollution flowing in a time unit through the area unit perpendicular to the flow direction. This is the flow intensity [μg ·m–2·s–1], being the measure of the inflow or outflow of a substance over an area unit situated within a territory (Stull 1995). Circular graphs were made for percentiles of the order p = 25, 50, 70, 98 after calculating the pollution flow intensities. A synthetic WZ indicator of pollution was calculated for discriminating between the impact of the emission and the influence of atmospheric conditions on the size of the imission registered. This indicator encompasses all impacts of various meteorological elements influencing the increase or decrease of air pollution during winter (Walczewski 1997). It is expressed in the formula: m above the ground level at the time 12 UTC (10-minute mean), C(IN) = 0 when there is a lack of lower inversion layer at 12 UTC, and 2 when the layer is present. The data necessary for verifying the existence of the inversion layer were got from measurements of air temperature at two levels on the mast installed in the MzWarszSGGW station. D(d) = 0 for d ≥ 4 [mm], 1 for 0.8 < d ≤4 [mm], 2 for 0 < d ≤ 0.8[mm], E(p) = 1.5 when there is the anticyclone on a given day, and 1 in the rest of cases. The minimum WZ value may amount to 0, and the maximum is 16.5. The WZ index was used with success for the first time by Walczewski (1997) in Kraków for identification of the changes in the level of air pollution with SO2. WZ = [A(T) + B(v) + C(IN) + D(d)] · E(p) where: A(T) – component describing the temperature impact, B(v) – component describing the wind speed impact, C(IN) – component describing the inversion layer impact, D(d) – component describing the precipitation impact, E(p) – coefficient taking into account the atmospheric air pressure. The individual articles for the WZ indicator as adopted in the paper were defined in the following way: A(T) = 0 for T > 0 [°C], 1 for –5 < T ≤ [0°C], 2 for –10 < T ≤ –5 [°C], 3 for –15 < T ≤ –10 [°C], 4 for T < –15 [°C], where T is the mean 24-hour temperature of air, B(v) = 0 for v > 2[m·s–1], 2 for 1< v ≤2 [m·s–1], 3 for 0 < v ≤1 0[m·s–1], where v is the wind speed at the height of 22 RESULTS The series of 1-hour concentration measurements of PM10 particulate matter gained in the period under study, i.e. 1.09.2003–30.10.2005 contained 94% of correct results with variability range from 0.8 to 241.8 [μg·m–3]. One can notice the occurrence of a clear cycle of daily dust concentration variability, with two local peaks in morning and evening time – around 9h00 and after21h00, and the hours of their occurrence vary during the year, on the figure 2 presenting the mean hourly concentrations of PM10 particulate within 24 hours. In the winter season the concentration maximums occur later in the morning and earlier in the evening as related to the summer season. A clear drop of PM10 particulate matter concentration is observed during Characteristics of the particulate matter PM10 concentration... w arm season 61 cold season 60 50 [μg•m-³] 40 30 20 10 23:00 22:00 21:00 20:00 19:00 18:00 17:00 16:00 15:00 14:00 13:00 12:00 11:00 10:00 09:00 08:00 07:00 06:00 05:00 04:00 03:00 02:00 01:00 00:00 0 FIGURE 2. Mean 1-hour concentrations of PM10 in the sequence of 24-hours in the station MzWarszUrsynów, 2004 the day around 13 h 00 and in the night around 5 h 00. The occurrence of these relationships results from not only the daily course of pollution emission to the atmosphere (e.g. traffic peaks) but also 24 hour cycle of changes in the height of mixing layer and further on from the development and disappearing of convection processes, and general and local circulation. Daily concentrations of PM10 particulate matter in the period under study (Tab. 1) acquired the values from the interval 5.2 – 156.4 [μg·m–3], that is from 10.4 to 312.8% respectively of mean daily allowable value (D24 = 50 [μg·m–3]) (Order by Minister of Environment of 6 June 2002). This value has been exceeded in 132 cases within the all measuring period, including 60% exceedences in heating seasons. The mean daily concentrations of dust acquired most frequently the values from 25 to 50 [μg·m–3] as it is shown in the frequency histogram of daily concentration occurrence (Fig. 3), based on intervals of the Air Quality Scale (WIOS) in both winter and summer seasons. The highest concentration values, exceeding D24, occurred most frequently in winter months, this being probably linked with the dropping air temperature and with more intensive work of heating facilities. In summer seasons, the values from the intervals 50–75 and 75–100 [μg·m–3] appeared as well, but the frequency of their occurrence was lesser than in winter seasons, but the values over 100 [μg·m–3] did not appear at all. In the warm season of year, exceedences of mean daily allowable value occurred most frequently in April and September (Tab. 1), but in the months V–VIII they appeared only sporadically. No one case of exceedence of the allowable standard occurred in the periods: V, VI, VIII of 2004, and VII–VIII of 2005. The phenomenon of the occurrence of a considerable number of exceedences in April and September requires a deeper analysis. In the years under study, the reasons of such a state can be seen in low mean speed of wind, low mean humidity of air, low sums of precipitation, and low diversity in the inflows of air masses (source: Daily Synoptic Bulletin of the IMGW 2003–2005). 62 G. Majewski, W. Przewoźniczuk 50 results [%] warm half-year 56,1% 60 cold half-year 43,1% 35,1 40 29,3% 30 20 13,0% 16,6% 10 1,7% 3,3% 1,9% 0 0–25 25–50 50–75 75–100 >100 –3 [μg•m ] FIGURE 3. Frequency distributions in percents of 24-hours particulate matter concentration at MzWarszUrsynów, X 2003–IX 2005 The mean monthly values of PM10 concentration were from 27.5 [μg·m–3] (in May 2004) to 47.4 [μg·m–3] (in September 2005), while the values averaged for winter and summer year halves differed only very slightly from each other, and they were 38.3 [μg·m–3] and 36.3 [μg·m–3] in winter year halves and a bit less – 33.3 [μg·m–3] and 35.7 [μg·m–3] in summer year halves respectively. Circular graphs of concentration percentiles related to individual sectors of wind rose were made and analysed in order to identify the directions of inflows of particulate matter (Fig. 4a). This figure shows the difference in the distribution of percentiles in summer and winter seasons over individual sectors. They were more even in summer seasons, and this fact could evidence the impact of pollution sources located at various directions around the measuring point. The emission from traffic sources prevailed in those periods because a thick network of streets with a high intensity of car traffic occurred around the measuring point. In the summer 2004, when no overstandard values were observed, no dominant direction of pollution inflow was seen. It results from the analysis of pollution plumes that the sources of emission intensities being the greatest in the period mentioned existed from the measuring station toward the South in the sectors 67–270oN (Fig. 4b). A bit otherwise was in the summer 2005. It is seen on Fig. 4a that the highest concentrations occurred at the air inflow from NE direction (the value of percentile of 98% level in the NE sector was by 20% higher than in the remaining sectors), and this can suppose an impact of a large point source of emission. The analysis of pollution plumes has shown that the intensity of the emission was the greatest in the sectors 110–180oN, and moreover in the sectors NE, N, WNW, where the following industrial plants were located: EC Siekierki Heating plant, Urban Construction Characteristics of the particulate matter PM10 concentration... Percentiles of PM10 pollution plume rate - 2003/2004 Perecentiles concentrations of PM10 - 2003/2004 N NNW a) 200 NNE [μg•m ] 150 NW N NNW b) -3 63 NE 250 [μg•m -2 •s-1] 200 NW NNE NE 150 100 WNW ENE WNW 100 50 W ENE 50 E 0 W E 0 50% 50% 70% 70% WSW ESE WSW ESE 98% 98% SW SW SE SSW SE SSW SSE SSE S S Perecentiles concentrations of PM10 - 2004/2005 N NNW 200 -3 [μg•m ] 150 NW Percentiles of PM10 pollution plume rate - 2004/2005 N NNW NNE [μg•m-2 •s -1 ] 200 NW NE 250 NNE NE 150 100 WNW ENE WNW 100 50 W ENE 50 E 0 W E 0 50% 70% 50% 70% WSW ESE WSW ESE 98% 98% SW SW SE SSW SE SSW SSE SSE S S Perecentiles concentrations of PM10 - 2004 Percentiles of PM10 pollution plume rate - 2004 N N NNW 200 [μg•m -3 ] NNW NNE 150 NW NE 250 NNE [μg•m-2 •s -1 ] 200 NW NE 150 100 WNW ENE WNW 100 50 W E 0 W 50% 70% ESE 98% SW WSW SE SSW ESE 98% SW S SSE S Percentiles of PM10 pollution plume rate - 2005 N N NNW 200 [μg•m -3 ] NNE NE WNW NE ENE 50 50 W E 0 50% E 0 -1 100 ENE W NNE -2 [μg•m •s ] 150 100 WNW 250 200 NW 150 NW SE SSW SSE Perecentiles concentrations of PM10 - 2005 NNW E 0 50% 70% WSW ENE 50 70% 50% WSW 70% WSW ESE ESE 98% SW SE 98% SE SW SSW SSW SSE SSE S S FIGURE 4. Circular graphs of percentiles concentrations of PM10 (a) and percentiles of PM10 pollution plume rate (b) in the cold and warm half-years from October 2003 to September 2005 at MzWarszUrsynów station 64 G. Majewski, W. Przewoźniczuk Enterprise, Radio Ceramics Plant, and the Construction Woodworking Plant. A part of the pollution could also flow from the city centre. In the winter season 2003–2004 the distribution of concentrations in individual sectors of the wind rose was almost evenly, with a slight overweight (under 10%) of concentration values occurring in the sectors 180–203oN and 338–360oN. This shows an impact of some sources on the imission field, the sources that were dispersed in various directions around the measuring station. It results from the analysis of pollution plumes that the greatest intensity of emissions came from the sources located in the sector 67–160oN and in the sector 315–360oN where point emission sources existed as it was mentioned in the preceding paragraph. In the season 2004–2005 the highest concentrations occurred at the air inflow from northern direction (Fig. 4a). The analysis of the plumes, similarly as for the preceding season, revealed the most intensive emission from the sector 67–180oN with local heating facilities and areas with low communal and living emissions (Fig. 1). Moreover, in that season one could note an inflow of greater amounts of dust from the sector 248–290oN. The diagram of plumes showed the existence of sources with great concentrations of emission from that direction where pollution sources worked intensively. The calculation of the WZ pollution index described in the preceding section was the next stage of the analysis of meteorological conditions important for pollution dispersal. The calculations were made for both heating seasons mentioned. However, only the results from January 2004 and 2005 were put down in the paper due to its limited size. It is seen on Figure 5 presenting the monthly courses of mean daily PM10 concentrations compared to the analogous course of the WZ index that the maximums of concentration values cover the maximums of the index values. Sometimes a slight retardation of the concentration maximum as compared to the highest index value can be noted, this fact being well grounded because the concentration resulted from unfavourable meteorological conditions. In January 2004 meteorological conditions were less favourable for pollution spreading out than in January 2005, because the mean WZI.2004 index amounts to 5.3 (max = 12), and WZI.2005 is 3.5 (max 8.2), and the mean concentrations were: 42.5 [μg·m–3] and 29,7 [μg·m–3] respectively. A more detailed analysis of those winter months has shown that high air pressure areas shaped the weather in January 2004. Either frosty and dry polar continental air flew from NE and E or arctic air from the north. At that time an increased emission of pollution from heating sources could occur at very low air temperatures. Not only the pressure pattern but also a small amount of precipitation (7 days with precipitation over 0.8 mm) favoured increased concentrations of pollution. However in January 2005 the mean monthly air temperature was higher than in 2004 and that pattern could cause a bit lesser emission from heating sources. The following impacts: occurrence of low air pressure patterns and the fronts accompanying them, the inflow of polar marine air, the increase of wind speed, Characteristics of the particulate matter PM10 concentration... 180 65 14 160 12 140 10 8 100 80 6 WZ indicator PM10 [μg·m-3] 120 60 4 D24 40 2 20 0 0 2 3 4 5 6 7 8 January 2004 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 PM10 D24 - 24-hour permissible concentration WZ 70 14 60 12 D24 PM10 [μg·m-3] 50 10 40 8 30 6 20 4 10 2 0 WZ indicator 1 0 1 2 3 January 2005 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 PM10 WZ FIGURE 5. Comparison of monthly variabilities of air pollution WZ indicator and PM10 concentration (24-hours) at MzWarszUrsynów station and the frequency of precipitation had contributed to better pollution spreading out and they could cause a drop in registered values of pollution concentration. STATEMENTS AND CONCLUSIONS The high values of PM10 particulate concentrations occurring in the Ursynow living quarter were caused, at a considerable rate, by a low emission coming from main communication 66 G. Majewski, W. Przewoźniczuk avenues and from dispersed sources (local heating facilities, individual domestic fire places) occurring in the sectors 67–180oN and 245–360oN. Low sources were of local importance but their impact decided on the scale of the threat from pollution. This threat results just from the fact that the low sources are located near human living areas; therefore dust aerosol breathed in is a ‘fresh’ one with a high share of miniparticles, and at the same time it is very dangerous for human health. This threat has been confirmed by introductory results of the quality analysis of PM10 found at the MzWarszSGGW station that showed the presence of heavy metals in the dust, such as Ni, Pb, Cr, Zn, Fe, Cu (archival data of the Meteorology and Climatology Department of the Warsaw Agricultural University). Since the year 2003, i.e. from starting the measurements of PM10 particulate matter concentration at the MzWarszUrsynow station, a decreasing tendency of this pollution is observed, although the number of allowable exceedences of mean daily values maintains still at a high level (mean daily allowable value of the particulate matter concentration was exceeded in 132 cases in the whole measuring period). The differences in seasonal concentrations were caused by emissions of particulate matter from burning out fuels in low sources being greater in winter seasons at relatively worse conditions of aeration and air self-cleaning that occur in the cold season. Meteorological conditions were the factor deciding on the imission registered in all winter months under study. It was impossible to found for the study period that a pollution emission decrease caused by implementation of the air quality improvement program had been the reason for the betterment of the imission situation in the area under analysis. According to the opinion of the authors, the decrease of particulate matter amount registered in the period 2003–2005 was caused by the betterment of meteorological conditions. REFERENCES DOCKERY D.W., CUNNINGHAM J., DAMOKOSH A.I., NEAS L.M., SPENGLER J.D., KOUTRAKIS P., WARE J.H., RAIZENNE M, SPEIZER F.E. 1996: Health effects of acid aerosols on North American children: respiratory symptoms. Environ. Health Perspect. 104, 500–505. EN12341 – Determination of the PM10 fraction of suspended particulate matter – Reference method and field test procedure to demonstrate reference equivalence of measurement methods. November 1998. JABŁOŃSKA M. 2003: Skład fazowy pyłów atmosferycznych w wybranych miejscowościach Górnośląskiego Okręgu Przemysłowego. Wydawnictwo Uniwersytetu Śląskiego. Katowice. Juda J., Chróściel S. 1974: Ochrona powietrza atmosferycznego. NIEĆKO J., NIEĆKO M., NIEĆKO U. 1998: Charakterystyka pyłu zawieszonego i jego wpływ na organizm ludzki. Air protection in theory and application. Section I, s. 29-49. KLIŚ CZ., MATEJCZYK M. 2002: Ocena wpływu źródeł na jakość powietrza w świetle ustawy „Prawo ochrony środowiska”. Ochrona Powietrza i Problemy Odpadów, vol. 36, nr 3: 95–98. Rozporządzenie Nr 62 Wojewody Mazowieckiego z dnia 8 grudnia 2003. Characteristics of the particulate matter PM10 concentration... Rozporządzenie Ministra Środowiska z dnia 6 czerwca 2002 r. w sprawie dopuszczalnych poziomów niektórych substancji w powietrzu oraz marginesów Tolerancji dla dopuszczalnych poziomów niektórych substancji (DzU z 2002r. nr 87 poz. 798). STERN A.C. et al. 1994: Fundamentals of air pollution. Academic Press, San Diego. STULL R.B. 1995: Meteorology Today For Scientists and Engineers, West Publishing Comp. New York. Quarg, 1996: Airborne Particulate Matter in the United Kingdom. Third Report of the Quality. WALCZEWSKI J. et al. 2000: Wykorzystanie danych meteorologicznych w monitoringu jakości powietrza. Biblioteka Monitoringu Środowiska, Warszawa. Walczewski J. 1997: Wskaźnik meteorologiczny określający prawdopodobieństwo wzrostu zanieczyszczenia powietrza w okresie zimowym. Wiad. IMiGW, t. XX (XLI). WARYCH J. 1999: Zanieczyszczenie powietrza cząstkami aerozolowymi i wynikające stąd problemy. Ochrona powietrza i Problemy Odpadów. Vol. 33, nr 3, majczerwiec. 67 Streszczenie: Charakterystyka pola imisji oraz próba określenia źródeł zanieczyszczenia powietrza pyłem zawieszonym PM10 na obszarze dzielnicy mieszkaniowej Ursynów na podstawie danych pomiarowych z automatycznej stacji monitoringu atmosfery. W pracy przedstawiono charakterystykę pola imisji pyłu zawieszonego PM10 za pomocą podstawowych statystyk, takich jak wartości średnie, zakresy zmienności, liczby przekroczeń dopuszczalnego stężenia, rozkłady częstości występowania stężeń w poszczególnych przedziałach, przebiegi czasowe. W celu przeanalizowania warunków meteorologicznych dyspersji i rozprzestrzeniania się zanieczyszczeń, obliczono średnie wartości podstawowych elementów meteorologicznych, sporządzono wykresy kołowe percentyli stężenia pyłu, oraz percentyli natężenia strumieni zanieczyszczenia. Obliczono również syntetyczny wskaźnik zanieczyszczenia WZ, ujmujący łączne oddziaływanie różnych elementów meteorologicznych na poziom stężeń. MS. received July 2006 Authors’ address: Wydział Inżynierii i Kształtowania Środowiska, SGGW, 02-787 Warszawa, ul. Nowoursynowska 159 Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 69–74 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Runoff volume and slope gradient relationship – laboratory investigations ANNA BARYŁA Department of Environmental Improvement, Warsaw Agricultural University – SGGW Abstract: Runoff volume and slope gradient relationship – laboratory investigations. The objective of the research was to analyze influence of the slope gradient on the volume of surface runoff. Runoff is the primary driving variable in the water – induced erosion process. The overland flow process is strongly affected by the slopes because of the effective rainfall rate at the surface. The laboratory physical model diameters were 120 cm long, 15 cm wide and 120 cm deep. The experiment the measurements of water erosion was done for different slopes: 5%, 10%, and 15%. The rainfall was artificial simulated and his intensity for experimental separate sets. Based on performed research it can be concluded that the volume of the runoff depends on the gradient of the slope. As the gradient increase the volume of the runoff also increase. Key words: rainfall simulation, physical model, experiment. INTRODUCTION The runoff of rainfall water on the slope depends on several factors such as: decline, form, length and plant cover of the slope, amount ant intensity of the rainfall as well as the water properties of the soil (Carson and Kirbky 1972, Słupik 1981, Morgan 1986, Józefaciuk 1996, Kinnell 1997). Among physical soil properties it is the soil infiltration ability that influences the runoff the most. The impact of infiltration on the possibility of surface runoff occurrence is revealed particularly when the surface lay is hardly permeable. The structure and the porosity of the soil have a decisive impact on the amount of infiltration (Józefaciuk A. and Józefaciuk C. 1996). Basic causes of variation of infiltration speed are changes of retentive capacity in the ground, the type of the gradient and the primary capacity value, treated as moisture in a time given (Słupik 1981, Schmidt 1997). Theoretical base of infiltration was formed by Horton (1993); according to him the surface runoff forms on a slope when the intensity of rainfall exceeds the infiltration capacity of the soil. In view of gradual filling of the soil pores with the water, the infiltration capacity of the soil diminishes as the rainfall lasts. When the intensity of the rainfall exceeds the infiltration process, a so called superinfiltration runoff occurs. However, when the surface ground lay is totally saturated with the water, a so called saturated zone runoff is formed (Słupik 1981, Kinnell 1997, Józefaciuk A. and Józefaciuk C. 1996, Rejman and Usowicz 1999). A particular type of it is a runoff in the initial phase of intense rainfall formed thanks to a dried-off and hardly permeable shell or a blockage laying on the surface against the infiltrating water which is created by the air bound in the soil. This type of runoff is called a pre-infiltration runoff (Schmidt 1997, 70 A. Baryła Kinnell 1997). In this case the surface runoff lasts until the soil is softened and the soil moisture is complemented. This type of runoff occurs the most often on heavy soils, thus short and of little intensity rainfalls very often can entirely constitute the runoff (Kinnell 1997). The process of forming the surface runoff depends in a big measure on, among others, the degree of covering of the soil surface. The practice has shown that it is more intense on open soils. FIGURE 1. Schema of laboratory experiment In natural conditions because of their big differentiation, it is extremely difficult to estimate precisely the impact of each of the factors mentioned above on the course of the runoff and its amount. The aim of the research was to determine the impact of the initial soil moisture, the decrease and the intensity of the rainfall on the value of surface runoff. The results of laboratory experiments are presented in the paper. Runoff volume and slope gradient relationship... MATERIALS AND METHODS The research was carried out on a ground model with the following diameters: 136 cm of length, 120 cm of altitude, 15 cm of width (Fig. 1). The inside oh the model was filled with clay sand firmly poured layer by layer and alternately compacted. In order to enable the observation of water flow in the soil the front wall of the model was made of a transparent polymethacrylate board. Inside the model 16 probes TDR were installed in order to measure the soil moisture. 0.8 mm of diameter each of them, the distance between them was 5 mm. They were connected to the TDR (Malicki 1996) set ant to the computer. All the time during the experiment (in ten-minutes intervals) changes of soil moisture were registered. The rainfall was simulated by microsprinklers; the intensity of outflow depended on the water pressure, which enabled modeling of rainfalls of different intensity. The surface runoff water was collected at the end of the model by a trough having an outlet directly connected to the measuring vessel. The measurements of outflow and moisture were made in the same intervals. 12 experiments were carried out on the model with three slopes of the terrain surface – 5, 10 and 15%, TABLE 1. Physical properties of the soil used in experiment Depth [m] Density [g·cm–3] Porous [%] Filtration coefficient [cm·s–1] 0.00–0.15 1.68 34.1 15.2*10–4 0.15–0.45 1.62 35.6 7.84*10–4 0.45–0.80 1.58 36.8 4.2*10–4 71 and different rainfall intensity: 41 mm h–1, 46 mm h–1, 52 mm h–1 and 54 mm h–1. The duration of every experiment was 1 hour. Surface runoff in the time from the physical model was collected at the end by trough, from of which one accompanied to dishes measuring. Physical properties of the soil are set together in the Table 1. Lesser and lesser values of filtration coefficient in subsequent layers may evidence about more compacted lower parts of the ground in the model. RESULTS AND DISCUSSION Knowledge and proper determination of interdependencies between surface runoffs and factors determining their value, intensity and frequency of occurrence is indispensable in the modeling of the value of surface runoff on a slope. On the Figure 2a are shown the results of surface runoff for the slope of 5%. The analyze of the results obtained can lead to the conclusion that at the rainfall intensity of 41 mm h–1 and a slope of 5% the summary runoff was 0.8 mm h–1, instead at a slope of 15% the summary runoff was 5.6 mm h–1. When the rainfall intensity was increased by 5 mm h–1 (46 mm h–1) at a slope 5% the summary surface runoff amounted to 1.2 mm h–1 at 15 mm h–1 – 11.8 mm h–1. The time of starting-up the surface runoff at the intensity of 41 mm h–1 was 22 minutes and was the longest time of starting-up the surface runoff at a slope 10%. The shortest time was observed at the intensity of rainfall of 54 mm h–1 and then the runoff started 10 minutes after the experiment had started. A. Baryła Surface runoff [mm] 12 a) 5% 41 mm/h 10 46 mm/h 8 52 mm/h 54 mm/h 6 4 2 0 0 Surface runoff [mm] 12 10 b) 20 30 40 50 60 30 40 50 60 30 40 50 60 10% 10 8 6 4 2 0 0 10 12 Surface runoff [mm] c) 20 15% 10 8 6 4 2 0 0 10 20 Time [min] FIGURE 2. Dynamics of runoff surface with rain intensity and different slope (a – 5%, b – 10%, c – 15%). As at the 5 and 10% slopes, the highest value of surface runoff during 1 hour was obtained at the rainfall intensity of 54 mm and equaled to 10.86 mm; this value was twice higher than surface runoff obtained at the intensity of 41 mm (3.68 mm). At the intensity of rainfall equal 41 mm h–1 the surface runoff increased after 10 minutes, and at the intensity of 54 mm h–1 – 2 minutes. The research also proved that in the initial rainfall phase there is an intensified runoff volume until a balance is obtained between the amount and the energy of runoff. The surface runoff speed stabilisation depends mainly on the rainfall intensity, which makes a very thin layer on the soil surface. For that reason the runoff value remains at the similar level, although the surface layer is almost fully saturated. The performed research confirmed that there is an important impact of the land slopes on the course and value of surface runoffs (Fig. 3). At a slope of 5% surface runoffs’ value was more than five times less than at a slope of 15% at the same rainfall intensity. The research performed by Szafrański and others (1998) revealed that on grounds at a slope 3–6% surface runoff value is low and therefore the erosion danger is very slight. A severe water erosion danger may occur only on the lands at a slope of more than 10%. None the less during violent summer high-intensity thunderstorms the phenomenon of linear erosion can occur even at slopes of less than 6% (Rejman 1999). 30 Coefficients of runoff C [%] 72 slope 5% slope 10% slope 15% 20 = 99 ;R 6 6*P R=9 646 *P; 3 , 0 434 0 = 3 C ,2 C=0 *P; R=91 03 90 66 C=0,0 10 0 0 10 20 30 40 Rainfall P [mm] 50 60 FIGURE 3. Relationship between rainfall intensity and coefficients of runoff Runoff volume and slope gradient relationship... The Figure 3 presents a relationship between rainfall intensity and coefficients of runoff for the results of the research. The determination coefficient amounted to respectively: for a slope 5%–99, for a slope 10%–96 and for 15% –91. The research performed revealed that when the rainfall intensity volume increases, surface runoff increases too. CONCLUSIONS 1. The research performed on a ground model confirmed an important role of land slopes played in a course and value of surface runoffs. At a slope equal 5% the runoffs’ value was almost five times less than at slopes equal 15%. It was ascertained that there is a rectilinear dependency between surface runoffs volume and land slopes. 2. Surface runoffs volume described by the runoff coefficient was 0,02 to 0,05 at a land slope 5% and 0,13 to 0,21 at a land slope equal 15%. REFERENCES CARSON M.A., KIRBKY M.J. 1972: Hillslope form and process. University Press, Cambridge. JÓZEFACIUK A., JÓZEFACIUK C. 1996: Mechanizm i wskazówki metodyczne badania procesów erozji. Biblioteka Monitoringu Środowiska w Warszawie. KINNELL P.I.A. 1997: Runoff ratio as a factor in the empirical modeling of soil erosion by individual rainstorms. Australian Journal of Soil Research No 25, p 1–6. MALICKI M.A., PLAGGE R., ROTH C.H. 1996: Improving the calibration of dielectric TDR soil moisture determination 73 taking into account the solid soil. European J. Soil Sci. 47, 57–366. MORGAN R.P.C. 1986: Soil erosion and conservation. Longman Scientific & Technical, Essex, UK. REJMAN J., USOWICZ B. 1999: Quantitative description of water ad soil transport in process of water erosion. Acta Agrophysica, no 23, p. 143–148 (in Polish). SCHMID B. 1997: Critical rainfall for overland flow from an infiltrating plane surface. Journal of Hydrology 193, 45–60. SŁUPIK J. 1981: Rola stoku w kształtowaniu odpływu w Karpatach Fliszowych. Prace Geograficzne, 142, 1–94. SZAFRAŃSKI C., FIEDLER M., STASIK R. 1998: Rola zabiegów melioracyjnych w ochronie przeciwerozyjnej gleb terenów bogato urzeźbionych Bibliotheca Fragmenta Agronomica tom 4B/98. WISCHMEIER W.H., SMITH D.D. 1978: Predicting rainfall erosion losses. USDA Agric. Handb. 537. U.S. Gov. Print. Office, Washington DC, p. 1–58. VAN DIJK A.I.J.M., BRIJNZEEL L.A. Modeling runoff and soil loss from bench terraced hillslopes in the volcanic uplands of West Java, Indonesia. YU B., SOMBATPANIT S., ROSE C.W., CIESIOLKA, C.A.A. and COUGHLAN K.J. 2000: Characteristics and modeling of Runoff Hydrographs for Different Tillage Treatments. Soil Science Society of American Journal No 64, p. 1763–1770. Streszczenie: Zależność pomiędzy spływem powierzchniowym a spadkiem terenu – doświadczenie laboratoryjne. Spływ powierzchniowy jest procesem złożonym, a czynniki od których zależy są zmienne zarówno w czasie, jak i w przestrzeni. Poznanie i właściwe określenie zależności pomiędzy spływami powierzchniowymi a czynnikami wpływającymi na ich wielkość, natężenie i częstotliwość występowania jest niezbędne m.in. dla oceny istniejących modeli spływu powierzchniowego. Badania laboratoryjne wykonano na modelu gruntowym o wymiarach: długość 136 cm, wysokość 120 cm, szerokość 15 cm. Wnętrze modelu wypełnione zostało piaskiem gliniastym mocnym 74 A. Baryła zagęszczanym warstwowo. Badania dowiodły, że w początkowej fazie opadu następuje zwiększenie wielkości spływu do momentu ustalenia się stanu równowagi pomiędzy intensywnością i energią opadu a wielkością i energią spływu. Prędkość ustabilizowania się spływu powierzchniowego zależy głównie od intensywności opadu, który powoduje tworzenie się na powierzchni gleby cienkiej warstewki. Wartość odpływu utrzymuje się przez to na zbliżonym poziomie, mimo że wierzchnia warstwa osiągnęła prawie pełny poziom nasycenia. MS. received November 2006 Author’s address: Katedra Kształtowania Środowiska – SGGW ul. Nowoursynowska 166 02-787 Warszawa e-mail: [email protected] Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 75–81 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Influence of sprinkling irrigation and nitrogen fertilization on health status of potato grown on a sandy soil DARIUSZ PAŃKA1, ROMAN ROLBIECKI2, CZESŁAW RZEKANOWSKI2 1 Department 2 of Phytopathology, University of Technology and Agriculture, Bydgoszcz, Poland Department of Land Reclamation and Agrometeorology, University of Technology and Agriculture, Bydgoszcz, Poland Abstract: Influence of sprinkling irrigation and nitrogen fertilization on health status of potato grown on a sandy soil. The field experiment was carried out in the years 2001–2003 at Kruszyn Krajeński near Bydgoszcz on a sandy soil (Typic Hapludolls). The field experiment was done using the split-plot method, in a dependent system with two variable factors (sprinkler irrigation and nitrogen fertilization) and three replications. ‘Drop’ early potato cultivar was taken into consideration. It is the most frequent cultivated cultivar in Kujawy-Pomerania region. The following levels of experimental factors were used: water: O – without sprinkler irrigation and W – sprinkler irrigation according to tensiometer, nitrogen fertilization: N1 – 75 and N2 – 125 kg N ha–1, respectively. Potato tubers were analyzed at once after harvest from the point of view of occurrence of common scab, black scurf, potatotuber dry-rot and potato-tuber soft-rot symptoms. Estimation of degree of infestation was carried out on 50 tubers taken at random from the each plot. Nine-degree scale was used (0–8°), where: 0° means lack of infestation symptoms (sound tubers) and 8° means over 50% of a tuber surface with symptoms of common scab or black scurf. In case of other diseases a percentage of tubers with symptoms of infestation was estimated. On the basis of the investigation results it was found that sprinkler irrigation had a significant influence on the occurrence of common scab symptoms. Stronger infestation was noted on irrigated treatments. A strong influence of the rainfall course in the vegetation period of the potato on the infestation of tubers by Streptomyces sp. was observed. An influence of nitrogen fertilization on the occurrence of common scab and black scurf symptoms was not found. The cultivar tested was characterized by a higher resistance to infestation by Rhizoctonia solani than Streptomyces sp. Key words: potato, sprinkling irrigation, fertilization, nitrogen, Streptomyces sp., Rhizoctonia solani, common scab, black scurf, sandy soil. INTRODUCTION High quality yield is at present the main aim of the table potato production. The quality of potato is affected most strongly by such the factors like cultivar, course of weather conditions during the vegetation period, mineral fertilization, soil cultivation and plant protection against agrophages [3], [7], [9], [16]. Good quality yield can be produced when optimal conditions for the potato growth and development are created. In such the conditions the pressure of agrophages is limited to minimum. Pathogens developing on tubers are included to the most dangerous pathogens. They often decreased the suitability of potato for food processing. In addition these pathogens decreased the quality of seed-potatoes and they can cause considerable storage losses [1], [6], [15]. Common scab, rhizoctoniose, tuber dryrot and tuber soft rot are included to the most dangerous diseases in cultivation of potato. They occur commonly. Their 76 D. Pańka, R. Rolbiecki, Cz. Rzekanowski development are influenced by a number of such the factors like for example moisture, temperature or mineral fertilization [8], [10], [11], [14]. In Poland cultivation of potato is conducted mostly on light soils characterized by a low fertility and limited water-holding capacity. Because of this, higher doses of mineral fertilization and irrigation of potato plantations should be used, especially in the regions characterized by periodical rainfall deficits. All of this with connection of other factors can influence the development of potato pathogens. The aim of the study was to determine the influence of nitrogen fertilization and sprinkler irrigation on the potato tuber health of ‘Drop’ cultivar, grown on the light soil. MATERIAL AND METHODS The field experiment was carried out in the years 2001–2003 at Kruszyn Krajeński near Bydgoszcz on a sandy soil (Typic Hapludolls). The water reserve to 1 m depth of soil at field capacity was 87 mm and the available water 67 mm. The field experiment was done using the split-plot method, in a dependent system with two variable factors (sprinkler irrigation and nitrogen fertilization) and three replications. ‘Drop’ early potato cultivar was taken into consideration. It is the most frequent cultivated in Kujawy-Pomerania region. The following levels of experimental factors were used: – water: O – without sprinkler irrigation and W – sprinkler irrigation according to tensiometer, nitrogen fertilization: N1 – 75 and N2 – 125 kg N ha–1, respectively. The plot area for harvest was 15 m2. The mineral fertilization was applied at following doses: 80 kg P2O5 ha–1 (superphosphate) and 140 kg K2O ha–1 (potash salt). Fertilization was used presowing in early spring. Chemical protection against potato late blight, colorado potato beetle was conducted in particular years of the study. Potato tubers were analyzed at once after harvest from the point of view of occurrence of common scab, black scurf, potato late blight, potato-tuber dryrot and potato-tuber soft-rot symptoms. Estimation of degree of infestation was carried out on 50 tubers taken at random from the each plot. 9-degree scale was used (0–8°), where: 0° means lack of infestation symptoms (sound tubers) and 8° means over 50% of a tuber surface with symptoms of common scab or black scurf. In case of other diseases a percentage of tubers with symptoms of infestation was estimated. Infection degrees were transformed into infection indexes (II) according to Townsend and Heuberger formula [17]. Obtained data were statistically analyzed using analysis of variance. Mean values were verified with Tukey’s test. – RESULTS AND DISCUSSION In all of the study years, symptoms of infestation by Streptomyces sp. and Rhizoctonia solani, occurred on potato tubers most often. Other pathogens were noted sporadically and they amounted less than 1% tubers with disease symptoms. Influence of sprinkling irrigation and nitrogen... Significant influence of sprinkler irrigation on the occurrence of common scab symptoms on tubers was observed (Tab. 1). Tubers from irrigated plots were significantly higher infected by Streptomyces sp. every year. The strongest symptoms of infestation occurred in 2001, on plots fertilized with the higher nitrogen dose. The lowest infestation by this pathogen occurred in exceptional dry year 2003 (Tab. 2). In this year the infection index on irrigated plots was lower than 24,5%. Similar results obtained Sadowski et al. [10], [11]. They observed a decrease of common scab symptoms occurrence on the very light soils, in years characterized by small amounts of rainfall. In such the 77 conditions, irrigation was favourable to the development of Streptomyces sp. An infestation increase of tubers by this pathogen on irrigated plots, in years with lower rainfall, noted also Gładysiak and Czajka [5] as well as Borówczak and Gładysiak [2]. In years with higher rainfall amounts they observed a decrease of the common scab occurrence. The first two years of the study (2001 and 2002) were characterized by higher rainfall amounts, especially year 2001 (Tab. 2). In spite of this, the infection increase of tubers by Streptomyces sp. on irrigated treatments was obtained. Reason of this phenomenon can be a low rainfall in June in these years. Amounts of rainfall in this month TABLE 1. Effect of nitrogen fertilization and irrigation on cv. ‘Drop’ tubers infection with Streptomyces sp. (Infection index in %). Kruszyn Krajeński 2001–2003 Years Water variant Non irrigated 2001 Sprinkler irrigation Mean Non irrigated 2002 Sprinkler irrigation Mean Non irrigated 2003 Sprinkler irrigation Mean Non irrigated 2001–2003 Sprinkler irrigation Mean 1 Nitrogen fertilization 125 kg·ha–1 75 kg·ha–1 18.7 a1 18.5 a a a 52.8 a 53.3 a b b 35.8 a 35.9 a 15.7 a 8.7 a a a 43.4 a 50.1 a b b 29.6 a 29.4 a 1.4 a 2.5 a a a 21.3 a 24.5 a b b 11.3 a 13.5 a 11.9 a 9.9 a a a 39.2 a 42.6 a b b 25.6 a 26.3 a Mean 18.6 a 53.1 b 12.2 a 46.8 b 1.9 a 22.9 b 10.9 a 40.9 b Mean values followed by the same letter in columns and rows are not significantly different at α = 0.05, according to Tukey’s test. 78 D. Pańka, R. Rolbiecki, Cz. Rzekanowski TABLE 2. Rainfall conditions recorded over vegetation period Period Rainfall (mm) April May June July August Total 1951–2000 26 40 56 70 48 240 2001 45 30 49 106 27 257 2002 13 50 44 108 41 256 2003 18 18 30 106 18 190 2001–2003 25 33 41 107 29 234 were, like in 2003, lower than the manyyear average. According to Sawicka [13] potato is the most susceptible to infection in the period from the beginning of tuber formation to the stage when diameters of tubers achieve 1,5–2 cm. Small amounts of rainfalls in this period are favourable to the development of common scab. ‘Drop’ is the very early cultivar, characterized by a short vegetation period, and because of this its largest susceptibility to infection occurred in the years, exactly in the period with low rainfall. Sprinkler irrigation was then favourable to the pathogen development. According to Szutkowska [14], the occurrence of common scab is mostly affected by the soil moisture and the soil temperature in the period of the early growth of tubers. The highest infection was observed when the soil temperature was about 21°C, and its moisture around 6%. Favourable arrangement of these factors in the years 2001 and 2002 in the remaining months of vegetation was probably the reason of the stronger infection of tubers on irrigated treatments as compared to the year 2003. Significant influence of the nitrogen fertilization dose on the common scab occurrence on tubers was not observed. Czajka et al. [4] observed an increase of common scab infection between treatments with doses 0 and 80 kg N ha–1. Doses which were increased above 80 kg N ha–1 had no larger effect. In the own experiments the occurrence of common scab was also on a similar level in case of both the doses tested. Significant influence of sprinkling irrigation on infection of tubers by Rhizoctonia solani was observed only in 2003 (Tab. 3). This year was characterized by the lowest amount of rainfall during the vegetation period of potato. Sprinkler irrigation in such the condition was favourable to the development of the pathogen as compared to control treatments (without irrigation). According to Weber [16], the high moisture of soil is favourable to sclerotia formation on tubers. This dependence was observed also in our earlier investigations [8]. Similar results were obtained also by Sadowski et al. [12]. In the remaining years of the study, with higher rainfall during the period of the potato growth, no differences in black scurf occurrence between irrigation and control treatments were observed. A significant effect of the nitrogen fertilization dose on black scurf of Influence of sprinkling irrigation and nitrogen... 79 TABLE 3. Effect of nitrogen fertilization and irrigation on cv. ‘Drop’ tubers infection with Rhizoctonia solani (Infection index in %). Kruszyn Krajeński 2001–2003 Years Water variant 2001 2002 2001–2003 Mean 75 kg·ha–1 125 kg·ha–1 Non irrigated 21.81 24.2 23.0 Sprinkler irrigation 20.1 20.8 20.5 Mean 21.0 22.5 Non irrigated 13.0 13.7 13.4 17.1 Sprinkler irrigation 15.8 18.4 Mean Mean 14.4 12.0 a a 20.6 a b 16.3 a 16.1 10.9 a a 25.2 a b 18.1 a Non irrigated 2003 Nitrogen fertilization Sprinkler irrigation 11.5 a 22.9 b Non irrigated 15.6 16.3 16.0 Sprinkler irrigation 18.9 21.5 20.2 Mean 17.2 18.9 1 Mean values without letters or followed by the same letter in columns and rows are not significantly different at α = 0.05, according to Tukey’s test. potatoes was not observed, too. In investigation of Czajka et al. [4] the significantly highest level of black scurf was noted on treatment with 80 kg N ha–1, and the significantly lowest level of black scurf of potatoes – on treatments with 0 and 160 kg N ha–1, in the years with average amount of rainfall. Higher nitrogen fertilization decreased the pathogen occurrence in warm and dry years. A strong influence of the rainfall course in the vegetation period of the potato on the infestation of tubers by Streptomyces sp. was observed. An influence of nitrogen fertilization on the occurrence of common scab and black scurf symptoms was not found. The cultivar tested was characterized by a higher resistance to infestation by Rhizoctonia solani than Streptomyces sp. CONCLUSIONS REFERENCES On the basis of the investigation results it was found that sprinkler irrigation had a significant influence on the occurrence of common scab symptoms. Stronger infestation was noted on irrigated treatments. 1. BOLIGŁOWA E., ŁABZA T., GLEŃ K., PUŁA J. 1999: Infection of potato tubers by storage diseases depending on organic fertilization. Progres in Plant Protection, Vol. 39 (2): 898–901. 2. BORÓWCZAK F., GŁADYSIAK S. 1999: Disease infestation of potato tubers 80 D. Pańka, R. Rolbiecki, Cz. Rzekanowski depending on irrigation and cultivation system. Progress in Plant Protection/ Postępy w Ochronie Roślin, Vol. 39 (2): 786–788. 3. CZAJKA W. 1988: Badania nad występowaniem ważniejszych bakteryjnych i grzybowych chorób ziemniaka na tle wybranych czynników agrotechnicznych oraz zabiegów chemicznych. Acta Acad. Agricult., Tech. Olst. Agricultura Supplement, C 44: 1–58. 4. CZAJKA W., CWALINA B., CZAJKA M., FABISIEWICZ M. 1999: Porażenie bulw ziemniaka patogenami w zależności od nawożenia mineralnego. Progress in Plant Protection/ Postępy w Ochronie Roślin, 39 (2): 852–855. 5. GŁADYSIAK S., CZAJKA M. 1996: Sprinkling influence on potato cultivars infestation by diseases. Nowe Kierunki w Fitopatologii, Materiały z Sympozjum, Kraków: 217–221. 6. JABŁOŃSKI K., CZERKO Z. 1995: Zbiór i przechowywanie ziemniaków. Fundacja Rozwój SGGW, Warszawa: 62–69. 7. MATKOWSKI K., PROŚBA-BIAŁCZYK U., PLĄSKOWSKA E. 2002: Influence of cultivar mixture of potato and chemical control against Phytophthora infestans (Mont.) De Bary on the health status of tubers. Advances of Agricultural Sciences Problem Issues, 489: 261–268. 8. PAŃKA D., PIŃSKA M. 2004: Influence of potassium fertilization on health status of Barycz and Triada potato cultivars. Prace Komisji Nauk Rolniczych i Biologicznych, BTN, 39, Seria B, 52: 261–269. 9. RUDKIEWICZ F., SIKORSKI L., ŚLĄZAK I. 1983: Wpływ rodzaju gleby, nawożenia i zwalczania Phytophthora infestans na rozwój niektórych chorób na roślinach i bulwach ziemniaka. Biul. Inst. Ziemn., 30: 157–170. 10. SADOWSKI CZ., GRABARCZYK S., RZEKANOWSKI CZ. 1988: Wpływ nawadnianianawystępowanieStreptomyces scabies (Taxter) i Rhizoctonia solani Kühn na bulwach ziemniaków uprawianych na glebie bardzo lekkiej. Acta Acad. Agricult. Techn. Olst., Agricultura 47: 45–54. 11. SADOWSKI CZ., PESZEK J., RZEKANOWSKI CZ., SOBKOWIAK S. 1996: Effect of irrigation and different nitrogen fertilization rates on the occurrence of Streptomyces scabies (Taxter) on potato cultivated on very light soil. Plant Breed. and Seed Sci. 40, No. 1–2: 45–49. 12. SADOWSKI CZ., KORPAL W., LENC L., KAWALEC A. 2003: Health status of tubers of potato cultivated under organic and integrated conditions. In: Obieg Pierwiastków w Przyrodzie, Gworek, B. and Misiak, J. (eds). Monografia tom II: 682–686. 13. SAWICKA B. 1995: The influence of selected elements of meteorological conditions on potato tubers with common scab. Zesz. Probl. Post. Nauk Roln. 419: 89–93. 14. SZUTKOWSKA M. 1999: Czy można ograniczyć porażenie bulw parchem zwykłym. Ziemn. Pol. 3: 5–9. 15. ŚNIEG L. 1992: Wpływ nawadniania i nawożenia azotem na występowanie niektórych chorób bulw ziemniaka i ich odporność na uszkodzenia mechaniczne. Fragmenta Agronomica, 3/92: 49–57. 16. WEBER Z. 1976: Wpływ przedplonu i innych czynników na występowanie rizoktoniozy ziemniaka (Rhizoctonia solani Kuhn). Rocz. Nauk Roln., Ser. E, 2: 45–65. 17. WENZEL H. 1948: Zur Erfassung des Schadenausmasses in flanzenschutzversuchen. Pflanzenschutzberichte 15: 81–84. Streszczenie: Wpływ deszczowania i nawożenia azotowego na zdrowotność ziemniaka uprawianego na glebie bardzo lekkiej. Ziemniak jest bardzo często atakowany przez liczne patogeny, mogące powodować znaczne straty plonu. Do najgroźniejszych chorób należą: Phytophthora infestans, Rhizoctonia solani, Streptomyces sp., Erwinia carotovora subsp. carotovora i Fusarium Influence of sprinkling irrigation and nitrogen... spp. Ich rozwój, a tym samym szkodliwość, zależy od wielu czynników. W największym stopniu na przebieg procesu chorobowego wpływają opady, temperatura, dostępność składników pokarmowych i stosowane zabiegi pielęgnacyjne. Celem przeprowadzonych badań było określenie wpływu nawadniania i zróżnicowanego nawożenia azotowego na porażenie bulw ziemniaka przez najgroźniejsze patogeny. Doświadczenie z uprawą ziemniaka odmiany ‘Drop’ wykonano w latach 2001–2003 jako dwuczynnikowe, na glebie lekkiej, należącej do kompleksu żytniego, w trzech powtórzeniach. Pierwszym czynnikiem było nawadnianie deszczowniane stosowane w zależności od potrzeb roślin oraz występujących niedostatków wody, a drugim nawożenie azotowe na poziomie 75 i 125 kg N ha–1. Bezpośrednio po zbiorach analizowano bulwy pod kątem występowania objawów parcha zwykłego, rizoktoniozy, zarazy oraz suchej i mokrej zgnilizny. W doświadczeniu zaobserwowano w kolejnych latach trwania badań duże nasilenie występowania na bulwach parcha zwykłego i rizoktoniozy. Deszczowanie powodowało wzrost porażenia bulw przez S. scabies i spadek zanieczyszczenia bulw sklerotami R. solani. Stopień porażenia bulw różnicowało także nawożenie azotowe. Pozostałe choroby notowano sporadycznie, a ich nasilenie zależało od przebiegu warunków atmosferycznych w poszczególnych latach trwania eksperymentu. MS. received November 2006 Authors’ addresses: Dariusz Pańka Department of Phytopathology, University of Technology and Agriculture, Bydgoszcz, Poland Kordeckiego 20, 85-225 Bydgoszcz, Poland e-mail: [email protected] Roman Rolbiecki Department of Land Reclamation and Agrometeorology, University of Technology and Agriculture, Bydgoszcz, Poland Bernardyńska 6, 85-029 Bydgoszcz, Poland e-mail: [email protected] Czesław Rzekanowski Department of Land Reclamation and Agrometeorology, University of Technology and Agriculture, Bydgoszcz, Poland Bernardyńska 6, 85-029 Bydgoszcz, Poland e-mail: [email protected] 81 Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 83–92 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Heat balance and climatic water balance in vegetation period of spring wheat ELŻBIETA MUSIAŁ1, JOANNA BUBNOWSKA1, EDWARD GĄSIOREK1, LESZEK ŁABĘDZKI2, 1 Agricultural 2 University of Wrocław, Departament of Mathematics IMUZ – Bydgoszcz Abstract: Heat balance and climatic water balance in vegetation period of spring wheat. This paper characterizes the changes of heat and climatic water balance structure during the growing season of spring wheat in four observatories: WrocławSwojec 1964–2000, Bydgoszcz 1946–2004, Gorzów Wielkopolski 1970–1995 and Łódź 1954– –1995. Study of changes and trends contains the following elements of heat balance: net radiation, latent heat flux,sensible heat flux, soil heat flux and their contribution in net radiation. Climatic water balance is defined as a difference between precipitation and potential evapotranspiration reckoned by the Penmann method. Key words: heat balance, spring wheat, sensible heat flux, latent heat flux, climatic water balance. INTRODUCTION Energy and water relations in different ecosystems are best described by water and heat balance structure. Heat and water balance are connected through a streamvapour, that transports a huge quantity of energy to the atmosphere. Because of it any change in water balance influences heat balance and vice versa. Climate changes with their consequences are connected with heat exchange between active surface and atmosphere. Therefore, searching for climate variability concentrates on studying heat and water balance of different ecosystems. Different surfaces transform this energy in a particular way [Bubnowska, Gąsiorek, Łabędzki, Musiał 2005], [Bubnowska, Gąsiorek, Łabędzki, Musiał, Rojek 2005], [Kędziora, Olejnik, Kapuściński 1989], [Kapuściński 2000], [Olejnik 1996]. The knowledge of heat balance for various kinds of surface allows characterizing changes in flux distribution. METHODS The calculation of components of active surface heat balance was carried aut by using the BMC model, worked out by Olejnik and Kędziora [1991, 1999] The heat balance of ecosystems is described by the following equation [Kapuściński 2000; Kędziora 1989]: Rn + LE + H + G = 0 where: Rn – net radiation [Wm–2] G – soil heat flux [Wm–2] H – sensible heat flux [Wm–2] LE – latent heat flux [Wm–2] The exact description of applied calculation method could be found in a previous paper by the same authors [Musiał 2001; Bubnowska, Gąsiorek 84 E. Musiał et al. Łabędzki, Musiał, 2005] and in [Bowen 1926], [Karliński, Kędziora 1968], [Shuttleworth, Wallace 1985]. Water balance in this paper is defined as a difference between precipitation and potential evapotranspiration reckoned by Penmann method. Penman showed that the latent heat flux could be expressed as: Δ ( Rn + G ) + Ea γ LE = ⎛ Δ⎞ ⎜1 + γ ⎟ ⎝ ⎠ where: Ea – ability of air evapotranspiration [Wm–s2], Ea = 7,44(1 + 0,54v)d v – wind speed at 2 m height [ms–1], d – vapour pressure deficit [hPa], γ – psychrometric constant γ = 0,655 [hPaK–1], ∆ – mean rate of change of saturated vapour pressure with temperature [hPaK–1], LE, Rn, G – like above. A simple relationship exists between evapotranspiration ETP expressed in [mm] and latent heat flux LE expressed in [Wm–2]: ETP = n LE 28,34 Where n – number of days in decade, in month. RESULTS The study was carried out on data from the following meteorological observatories: Wrocław-Swojec (1964–2000), Bydgoszcz (1946-2003), Łódź (1954-1995) and Gorzów Wielkopolski (1975–1995). The variability of heat and water balance components was analyzed in given perennials. HEAT BALANCE IN VEGETATION PERIOD OF SPRING WHEAT Mean values of net radiation calculated for vegetation period of a spring wheat (IV–VIII) in ssubsequent years in Gorzów Wielkopolski fluctuated between 94 and 112 Wm–2. The lowest value was obtained in 1962 when the sum of precipitation in vegetation period was also the lowest. Latent heat flux absorbed from 61 to 68% of net radiation, and sensible heat flux – from 24 to 61%. The soil heat flux accounted for around 8% of net radiation. Values of heat balance components in other observatories had similar variability. Net radiation in Łódź changed from 90W/m2 in 1960, when the precipitation was the highest (476.7 mm), to 115 W/m2 in 1983, when the precipitation was the lowest (171.7 mm). In Wrocław and Bydgoszcz net radiation fluctuated from 89 to 112 [Wm–2] and from 89 to 109 [Wm–2] respectively. Latent heat flux absorbed from 58% in Bydgoszcz to 69% of net radiation in Łódź. The highest values of net radiation during the growing season were observed in Łódź and Gorzów Wielkopolski, whereas in Wrocław and Bydgoszcz the values of Rn were lower. Latent heat flux values were the highest in Łódź, where net radiation was the highest. This may be due to the fact that in Łódź more energy was supplied and therefore, more of it could be used for evapotranspiration. Heat balance and climatic water balance... 85 TABLE.1. Components of heat balance during the growing season of spring wheat (IV–VIII) in Gorzów Wielkopolski (1970–1995) Rok Rn -LE -G -H -LE/Rn -H/Rn -G/Rn P 1970 101 66 8 27 0,65 0,27 0,08 215,3 1971 105 64 8 33 0,61 0,31 0,08 256,4 1972 97 62 7 28 0,64 0,29 0,07 306,2 1973 105 66 8 31 0,63 0,30 0,07 273,7 1974 94 59 8 27 0,63 0,29 0,08 330,6 1975 110 71 8 31 0,65 0,28 0,07 233,0 1976 110 72 8 30 0,65 0,28 0,07 172,7 1977 98 65 8 25 0,66 0,26 0,08 460,0 1978 101 64 8 29 0,63 0,29 0,08 279,0 1979 102 65 8 29 0,63 0,29 0,08 225,5 1980 95 59 8 28 0,62 0,30 0,08 320,1 1981 104 63 9 32 0,61 0,31 0,08 306,9 1982 112 70 9 33 0,63 0,30 0,07 153,6 1983 109 71 8 30 0,65 0,28 0,07 214,4 1984 95 60 7 28 0,63 0,29 0,08 353,1 1985 103 65 8 30 0,63 0,29 0,08 229,6 1986 106 68 8 30 0,64 0,28 0,08 255,2 1987 94 62 7 25 0,66 0,27 0,07 365,3 1988 95 59 8 28 0,62 0,30 0,08 259,7 1989 107 70 8 29 0,65 0,27 0,08 190,9 1990 108 66 9 33 0,61 0,31 0,08 285,4 1991 102 69 8 25 0,68 0,24 0,08 221,0 1992 108 72 8 28 0,67 0,26 0,07 170,0 1993 100 64 8 28 0,64 0,28 0,08 268,0 1994 107 71 8 28 0,66 0,26 0,08 259,0 1995 108 69 8 31 0,64 0,29 0,07 320,0 –2 –2 –2 Rn – net radiation [Wm ] , LE – latent heat flux [Wm ], G – soil heat flux [Wm ], H – sensible heat flux [Wm–2], P – sum of precipitation in period IV–VIII [mm] Regarding the Figure 3, the following conclusion could be drawn: in Łódź and Gorzów Wielkopolski more energy was transferred to the atmosphere from the active surface of spring wheat than in Wrocław and Bydgoszcz. In 1982–1995 perennial, common for all observatories, there was a distinct increasing tendency for sensible heat flux. Thus, the amount of energy used for heating atmosphere is growing. Due to the fact that the values of heat balance components depend on the net radiation value, it is worthy looking into contribution of each component to net radiation (Rn). 86 E. Musiał et al. FIGURE 1. Variation of mean ten-days values of net radiation (Rn) during the growing season of spring wheat in Wrocław – Swojec, Bydgoszcz, Gorzów Wielkopolski and Łódź FIGURE 2. Variation of mean ten-days values of latent heat flux (LE) during the growing season of spring wheat in Wrocław – Swojec, Bydgoszcz, Gorzów Wielkopolski and Łódź FIGURE 3. Variation of mean ten-days values of sensible heat flux (H) during the growing season of spring wheat in Wrocław – Swojec, Bydgoszcz, Gorzów Wielkopolski and Łódź Heat balance and climatic water balance... 87 FIGURE 4. Variation of mean ratios of latent heat flux and net radiation (LE/Rn) during the growing season of spring wheat in Wrocław – Swojec, Bydgoszcz, Gorzów Wielkopolski and Łódź FIGURE 5. Variation of mean ratio values of sensible heat flux and net radiation (H/Rn) during the growing season of spring wheat in Wrocław – Swojec, Bydgoszcz, Gorzów Wielkopolski and Łódź Looking at the course of mean values of latent heat flux contribution in net radiation, distinct decreasing tendency for this contribution in the last 20 years is seen. Therefore, less and less energy is used for evapotranspiration, especially in Wrocław and Bydgoszcz. Increasing share of sensible heat flux in net radiation in all observatories stress that more and more energy in all regions is used for heating atmosphere. The above mentioned results are concordant with those by Musiał [Musiał, Gąsiorek, Rojek 2004], Trepińska[1997], [Ryszkowski Kędziora, 1995] and [Kożuchowski 2004]. The temperature increase is a consequence of the growing contribution of latent heat flux to net radiation. Regression equations in Table 2 confirm this. CLIMATIC WATER BALANCE The net climatic water balance determines conditions of plant vegetation [Bac, Rojek, 1979, 1982]. This index may be the source of information on climate change effects. 88 E. Musiał et al. TABLE 2. Mean parennial air temperature values for periods IV–VIII and IV–IX with regression equations Obserwatory Tweg [°C] Tpc [°C] ssr [°C] Linear regression equation Tendency [°C/10 years] Bydgoszcz 15.22 15.0 0.92 y = 0.0176x + 14.7 0.18* Gorzów Wielkopolski 14.57 14.4 0.85 y = 0.0477x + 13.9 0.48* Łódź 14.32 14.1 0.88 y = 0.0129x + 14.0 0.13** Wrocław 14.93 14.7 0.83 y = 0.0347x + 14.3 0.35* Tweg – mean seasonal yearly air temperature (IV–VIII), Tpc – mean seasonal, yearly air temperature (IV–IX), ssr – standard devation of temperature *) – statistically significant for α = 0.05 **) – statistically significant for α = 0.3 FIGURE 6. Variation of climatic water balance during the growing season of spring wheat (IV–VIII) in Bydgoszcz (1946–2003) FIGURE 7. Variation of precipitation (P) and climatic water balance (CWB) during the growing season of spring wheat (IV–VIII) in Wrocław (1964–2000) Heat balance and climatic water balance... In Bydgoszcz climatic water balance was positive in two years during the period 1946–2004. Those two years were chracterized by high sums of precipitation: 599,1 mm in 1980 and 506,6 mm in 1985. In years 1980 and 1985 the spring wheat vegetation period was characterized by rather low (390,56 and 419,35 mm, respectively) evapotranspiration values, whereas mean evapotranspiration value for years 1946–2004 was 456,5 mm. 89 In Wrocław climatic water balance was positive in 1980 (11,3 mm) and in 1986 (9,2 mm). During the spring wheat vegetation period in 1980, the precipitation value was 438 mm and the potential evapotranspiration reached the level of 426,7 mm. The year 1997 is noteworthy due to the flood in Wrocław. During that spring wheat vegetation period the precipitation sum (494 mm) was 52% higher than the mean value. In spite of such high precipitation value, FIGURE 8. Variation of precipitation (P) and climatic water balance (CWB) during the growing season of spring wheat (IV–VIII) in Łódź (1954–1995) FIGURE 9. Variation of precipitation (P) and climatic water balance (CWB) during the growing season of spring wheat (IV–VIII) in Gorzów Wielkopolski (1970–1995) 90 E. Musiał et al. TABLE 3. Evapotranspiration tendencies in 4 examined regions Tendency [°C/10 years] P [mm] ETP [mm] Linear regression equation Bydgoszcz 281 457 y = 0.29x + 443.4 2,9 Gorzów Wielkopolski 266 477 y = 4.91x + 466.7 49,1* Łódź 312 505 y = 2.0x + 470.6 20** Wrocław 326 500 y = 2.0x + 468.2 20* Obserwatory P – mean seasonal, yearly precipitation (IV–VIII) ETP – mean seasonal, yearly evapotranspiration (IV–VIII), *) – statistically significant for α = 0.05 **) - statistically significant for α = 0.01 the net climatic water balance in the 1997 growing season remained negative (–31 mm). In Łódź, during the 1954–1995 perennial, there were 3 years with positive net climatic water balance: 1960 (precipitation 477 mm, climatic water balance 20 mm), 1980 (458 mm and 38,9 mm, respectively) and 1985 (464 mm and 17 mm, respectively). Mean precipitation sum in the growing season for spring wheat in Łódź was 312 mm. In Gorzów Wielkopolski, the climatic water balance was positive only once and reached the value of 48,2 mm in 1977, when the precipitation sum in vegetation period was the highest (460 mm). All analyzed regions were characterized by continuous water shortage in vegetation period. During the period 1970–1995 the worst water conditions were observed in Gorzów Wielkopolski (the highest water shortage sum – 5474 mm), whereas the best terms were seen in Wrocław (water shortage sum of 4640 mm was 15% lower than that in Gorzów). During the wheat spring growing season the values of climatic water balance were positive only in years with the highest precipitation sums. CONCLUSIONS The enlargement of negative net climatic water balance, along with increasing potential evapotranspiration, confirms diminishing atmospheric precipitation. The increase of temperature during the spring wheat growing season in all discussed regions is a consequence of enlarging sensible heat flux, used for heating atmosphere. The growing temperature values enlarge the saturation deficiency, thus allowing more water to evapotranspirate (PET increase). Consequently, growing precipitation enlarge the negative net climatic water balance. The tendencies observed among heat balance components are concordant with changes seen among the components of climatic water balance. The research supported by KBN grant in years 2004–2007. REFERENCES BAC S., ROJEK M. 1979: Klimatyczny bilans wodny a odpływy w Polsce, Przegl. Geofiz. 24(3) 293–297. Heat balance and climatic water balance... BAC S., ROJEK M. 1977: Metodyka oceny stosunków wodnych obszarów rolniczych na podstawie danych klimatycznych, Zesz. nauk ART Olszt. Nr 21, 13–24. BAC S., ROJEK M. 1982: Klimatyczne bilanse wodne w Polsce [w:] Bac S. (red.) Agrometeorologiczne podstawy melioracji wodnych w Polsce. PWRiL Warszawa. BOWEN I.S. 1926: The ratio of heat losses by conduction and by evaporation from any water surface. Phys. Rev., 27, p. 779–787. BUBNOWSKA J., GĄSIOREK E., ŁABĘDZKI L., MUSIAŁ E., ROJEK M. 2005: Bilans cieplny lasu iglastego w latach o ekstremalnych opadach i jego wieloletnie zmiany w rejonie Bydgoszczy i Wrocławia. Woda-Środowisko-Obszary Wiejskie t. 5, z. spec. (14) s. 69–82. BUBNOWSKA J., GĄSIOREK E., ŁABĘDZKI L., MUSIAŁ E. 2005: Struktura bilansu cieplnego łanów w ekstremalnych warunkach opadowych na tle wielolecia. Woda-Środowisko-Obszary Wiejskie, t. 5 z. 2(15) s. 31–52. ŁABĘDZKI L., BĄK B. 2004: Standaryzowany klimatyczny bilans wodny jako wskaźnik suszy. Acta Agrophys. Vol. 3(1) s. 117–124. KAPUŚCIŃSKI J. 2000: Struktura bilansu cieplnego powierzchni czynnej na tle warunków klimatycznych środkowozachodniej Polski, Rocz. Nauk. AR Pozn. Rozpr. Nauk. AR Poznań, 303, s. 250. KARLIŃSKI M., KĘDZIORA A. 1968: Rozważania metodyczne przy układaniu kalendarza przyrody dla województwa szczecińskiego, Pr. Kom. Nauk Roln. Kom. Nauk Leśn. PTPN 24, s. 159–175. KĘDZIORA A. 1999: Podstawy agrometeorologii, Poznań, PWRiL. KĘDZIORA A., OLEJNIK J., KAPUŚCIŃSKI J. 1989: Impact of landscape structure on heat and water balance. Ecol. Intern. Bull. 17 ss. 1–17. KOŻUCHOWSKI K. 2004: Skala, uwarunkowania i perspektywy współczesnych 91 zmian klimatycznych w Polsce, Łódź, Biblioteka s. 170. MUSIAŁ E. 2001: Modelowanie procesu ewapotranspiracji rzeczywistej i prognozowanie jego tendencji. Zesz. Nauk. AR Wroc., nr 412, Rozpr. 182, ss. 116. MUSIAŁ. E. , GĄSIOREK E., ROJEK M.S. 2004: Zmienność temperatury powietrza w obserwatorium Wrocław–Swojec w latach 1964-2001. Acta Agroph. nr 105 Vol. 3(2) s. 333–342. OLEJNIK J., KĘDZIORA A. 1991: A model for heat and water balance estimation and its application to land use and climate variation., Earth Surface Processes Landforms Vol.16, ss. 601–617. OLEJNIK J. 1996: Modelowe badania struktury bilansu cieplnego i wodnego zlewni w obecnych i przyszłych warunkach klimatycznych, Rocz. Nauk. AR Poznań z. 268, ss. 125 RYSZKOWSKI L., KĘDZIORA A. 1995: Modification of the effects of global climate change by plant cover structure in an agricultural landscape. Geogr. Pol. Vol. 65, s. 5–34. RYSZKOWSKI L., KĘDZIORA A. 1993: Rolnictwo a efekt szklarniowy. Kosmos 42 s. 123–149. SHUTTLEWORTH W.J., WALLACE J.S. 1985: Evaporation from sparse crops-an energy combination theory, Quart. J. R. Met. Soc. 111, pp. 839–895. TREPIŃSKA J. 1997: Wahania klimatu w Krakowie (1792–1995). Kraków Inst. Geogr. UJ, s. 1–204. Streszczenie: Bilans cieplny i klimatyczny bilans wodny dla okresu wegetacji pszenicy jarej. Praca charakteryzuje zmiany struktury bilansu cieplnego i klimatycznego bilansu wodnego w czasie okresu wegetacji pszenicy jarej dla czterech stacji: Wrocław-Swojec 1964–2000, Bydgoszcz 1946– 2004, Gorzów Wielkopolski 1970–1995 i Łódź 1954–1995. W pracy przeanalizowano zmiany i trendy takich składowych bilansu cieplnego, jak: saldo promieniowania i strumienie ciepła jawnego, ciepła utajonego przeznaczonego na parowanie, ciepła wymienianego z podłożem oraz ich udział w saldzie promieniowania. 92 E. Musiał et al. MS. received November 2006 Authors’ addresses: E. Musiał, J. Bubnowska, E. Gąsiorek Agricultural University of Wrocław Departament of Mathematics ul. Grunwaldzka 53 50 357 Wrocław Poland (071)3205659 Leszek Łabędzki ul. Glinki 60 85 174 Bydgoszcz Poland [email protected] Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 93–100 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Climatic and agricultural water balance for grasslands in Poland using the Penman-Monteith method WIESŁAWA KASPERSKA-WOŁOWICZ, LESZEK ŁABĘDZKI Institute for Land Reclamation and Grassland Farming, Regional Research Centre in Bydgoszcz Abstract: Climatic and agricultural water balance for grasslands in Poland using the Penman-Monteith method. The aim of the paper was to estimate the spatial variability of climatic and agricultural water balance for two-cut meadows (for two hay yield: 5 and 7 Mg ha-1) in different agro-climatic regions in Poland. It let to estimate meadow water needs under different meteorological conditions. The climatic water balance is a difference between precipitation and reference evapotranspiration calculated according to the PenmanMonteith formula. The agricultural water balance is a difference between precipitation and potential crop evapotranspiration. Potential grassland evapotranspiration is the result of multiplication crop coefficient and reference evapotranspiration according to the Penman-Monteith formula. The crop coefficient values were calculated on the basis of multiannual lysimeter experiments of meadow evapotranspiration. The analysis was carried out in the growing season (from April to September) in the years 1970–1995 for 17 meteorological stations located in different regions of Poland. The balances were done for ten-day periods, months and the whole growing season. The climatic water balance indicates the risk of agro-climatic drought. The agricultural water balance indicates water deficit and irrigation water needs of two-cut meadows in different regions of Poland. The negative values of this balance mean the scarcity of water for grasslands. The values of agricultural water balance are shown on the maps. Key words: climatic and agricultural water balance, two-cut meadow. INTRODUCTION The estimation of drought hazard and irrigation needs for plants can be estimated using crop coefficients method according to the Penman-Monteith formula. This method for calculation reference evapotranspiration ETo is commonly recommended and used in the word To estimate crop water requirements, one can relate potential evapotranspiration from the cropped soil with an optimum water supply under consideration to an estimated reference evapotranspiration by means of crop coefficient [1, 3]. Evaluation of crop water demands, agricultural droughts and irrigation requirements can be made with the help of climatic water balance and agricultural water balance. The climatic water balance can be interpreted as the indicator of climate dryness [7]. The agricultural water balance indicates the values of irrigation water needs for plants. These two balances are needed to the preliminary estimation of irrigation needs. 94 W. Kasperska-Wołowicz, L. Łabędzki The aim of the paper was to estimate the spatial variability of climatic and agricultural water balance for grasslands in different agro-climatic regions in Poland and to show these values on the maps. These maps let to estimate water needs for 2-cut meadow with the different yield in the years with different meteorological conditions. The climatic water balance CWB is the difference between precipitation sum P and reference evapotranspiration ETo calculated according to the PenmanMonteith formula [1, 6]: (1) TABLE 1. The crop coefficients kc for the PenmanMonteith equation for 2-cut meadow in Poland Month IV V VI VII VIII IX Ten-day period 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 AWB = P – ETP kc coefficient for hay yield: 7 Mg⋅ha–1 5 Mg⋅ha–1 0.50 0.45 0.75 0.70 0.95 0.80 1.00 0.90 1.15 1.00 1.20 1.10 1.30 1.20 0.55 0.45 0.65 0.55 0.80 0.70 0.90 0.80 1.10 1.00 1.30 1.15 1.20 1.10 1.35 1.25 1.10 1.10 1.10 1.10 1.10 1.10 (2) where: ETP – is evapotranspiration of wellwatered crop calculated from the formula: ETP = ETo⋅kc METHODS AND MATERIAL CWB = P – ETo The agricultural water balance AWB is the difference between precipitation sum P and potential crop evapotranspiration ETP: (3) where: kc – crop coefficient, ETo – reference evapotranspiration (mm). The kc coefficients values were calculated on the basis of multiannual lysimeter experiments and depended on the hay yield (Tab. 1) [4, 6, 8]. The analysis was carried out in the growing season (from April to September) in the years 1970–1995 for 17 meteorological stations. The chosen stations are situated in different agroclimatic regions of Poland (Tab. 2). According to Bac, Koźmiński, Rojek [2] taking into account water conditions one can distinguish regions: A – dry, B – moderately wet, C – wet. Taking into account thermal and solar energy conditions one can distinguish regions: 1 – hot and sunny, 2 – hot and moderately sunny, 3 – hot and cloudy, 4 – moderately hot and sunny, 5 – moderately hot and moderately sunny, 6 – moderately hot and cloudy, 8 – cold and moderately sunny. The agricultural water balance AWB was calculated for 2-cut meadows with the hay yield of: 7 Mg⋅ha–1 and 5 Mg⋅ha–1. The negative values (–) of climate and agricultural water balance showed Climatic and agricultural water balance... TABLE 2. Meteorological stations and agroclimatic regions according to Bac [2] Number of the station 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Name of the station Koszalin Bielnik Biebrza Chojnice Szczecin Frydrychowo Poznań Warszawa Zielona Góra Łódź Sosnowica Wrocław Opole Częstochowa Zamość Kraków Rzeszów Agro-climatic region C8 B4 B8 B8 A6 A2 A3 A2 A1 A5 A1 B2 B3 B6 B5 C5 C8 the drought risk and water deficit for grasslands irrigation, the positive values (+) – showed the excess of water. The calculation of the balances have been done for ten-day periods, months and the whole growing season (from April to September). The probability distributions of climatic water balance and agricultural water balance were determined. It was assumed that the examined period was classified as wet at the probability of exceedence equal to 25%, mean – 50% and dry – 75%. RESULTS Climatic water balance The climatic water balance CWB is different in Poland and depends on 95 agrometeorological conditions in the examined regions. The mean multiannual values of this balance (at the p = 50%) for growing season (April to September) were negative in all examined stations. In the north (stations 1–3) and south (stations 14, 16, 17) parts of the country the values of climatic water balance were negative but higher than –40 mm. The values of CWB less than –100 mm were in the central Poland, in the regions represented by the stations numbered 5–10. The lowest values of the balance (less than –150 mm) were in the region between the Odra and left side of Wisła rivers, especially represented by the stations 5–7 and 9–10. The highest risk of agro-climatic drought in Poland was observed on the area from the north-west to the central part of Poland (Fig. 1). The lowest monthly values of CWB in central Poland were in May, June and July. It resulted from the monthly distribution of precipitation and reference evapotranspiration. For instance in Poznań countryside (station 7), the multiannual mean monthly sum of ETo reaches the highest values in July (about 100 mm), a little less in June, May and August (about 80 mm) [5]. The higher values of CWB in the average year were found in the south (stations 14, 16, 17) part of Poland, the region situated near the Karpaty mountains and in the north part of Poland – near Bałtyk sea (station 1) and in the Biebrza river valley (station 3) – in the north-east part of the country. Agricultural water balance In the wet growing season (at 25% probability) agricultural water deficit did not occur in the north and south parts of Poland. The negative values of 96 W. Kasperska-Wołowicz, L. Łabędzki FIGURE 1. Mean multiannual precipitation (1) and the climatic water balance (2) in the growing season (IV–IX) in different regions of Poland agricultural water balance for low and high yield meadows were observed in the Warta river basin (stations 6, 7, 10), in the west part of Poland (stations 5, 9) and in the Polesie Lubelskie region (station 11). It means that even in the wet year the big area of Poland is threaten by droughts. The highest water deficit for meadow was in the Szczecin countryside (Fig. 2a). In the mean growing season (at 50% probability) the values of agricultural water balance were negative for 7 Mg⋅ha–1 hay yield meadow in the whole area of Poland. The high water deficits for grasslands (more than 100 mm) were observed on the area from the northwest to the central part of Poland. The highest scarcity of water (more then 150 mm) was observed in Warta and middledown Odra river basins. In mean year, in all regions of Poland, water deficit for 5 Mg⋅ha–1 hay yield meadow was 40–47 mm less than for 7 Mg⋅ha–1 hay yield meadow (Fig. 2b). In the dry growing season (at 75% probability) agricultural water deficit was observed in the whole area of Poland. The highest values of water deficit were in the area between Odra and the left side of Wisła and they exceeded 200 mm for 7 Mg⋅ha–1 yield meadow and 160 mm for 5 Mg⋅ha–1 yield meadow. Water deficit for high yield meadow was less than 100 mm only in south part of Poland in Kraków countryside (Fig. 2c). The values of climatic water balance (CWB) and agricultural water balance (AWB) for the 7 Mg⋅ha–1 hay yield meadow did not differ very much in the whole mean growing season (at 50% probability). These values differed in particular months. They depended on agrometeorological conditions and the time of first and second cut of meadow. The monthly distribution of climatic Climatic and agricultural water balance... a 97 b c FIGURE 2. Agricultural water balance for 2-cut meadow at the yield 7 Mg·ha–1 (upper number, mm) and 5 Mg·ha–1 (lower number, mm) in different agroclimatic regions of Poland in the growing season (IV–IX) a) wet, b) mean, c) dry • 01 – number of meteorological station, 01 – Koszalin and agricultural water balance is shown on the example of Wielkopolska region (dry and hot) – represented by Poznań agrometeorology station (7) and the Biebrza river valley (moderately wet and cold) – represented by station 3. (Fig. 3). The highest water deficits for meadow were in May and September – the months of intensive grass growth. In Wielkopolska region, in these months, they exceed 50 mm and in Biebrza river catchment – 30 mm. In the other months the values of climatic water balance were higher than values of agricultural water balance. In Biebrza region the values of climatic and agricultural water balance were positive in September. In every month they were higher than in Wielkopolska region. CONCLUSIONS 1. The climatic water balance CWB is different in Poland and depends on agrometeorological conditions. The highest risk of agro-climatic drought in Poland was observed in the region 98 W. Kasperska-Wołowicz, L. Łabędzki a b FIGURE 3. Mean monthly multiannual climatic water balance (1) and agricultural water balance for 7 Mg·ha–1 yield 2-cut meadow (2) in the growing season (IV–IX) in a) Poznań, b) Biebrza Climatic and agricultural water balance... 2. 3. 4. 5. between the Odra and left side of Wisła rivers. In the wet year the big area of Poland is threaten by droughts. The negative values of agricultural water balance for low and high yield meadows were observed in the west, central and east part of Poland. In the mean year the values of agricultural water balance were negative for 7 Mg⋅ha–1 hay yield meadow in the whole area of Poland. The highest scarcity of water (more than 150 mm) was observed in Warta and middle-down Odra river basins. Water deficit for 5 Mg⋅ha-1 hay yield meadow was 40–47 mm less than for 7 Mg⋅ha–1 hay yield meadow. In dry year meadow water deficit was observed in the whole area of Poland. The highest water needs were in the area between Odra and left side of Wisła and they exceeded 200 mm for 7 Mg⋅ha–1 yield meadow and 160 mm for 5 Mg⋅ha–1 yield meadow. Water deficits should be satisfied by efficiently performing irrigation systems. REFERENCES 1. ALLEN R.G., PEREIRA L.S., RAES D. & SMITH M. 1998: Crop evapotranspiration – Guidelines for computing crop water requirements. FAO Irrigation and drainage no 56, pp. 300. 2. BAC S., KOŹMIŃSKI C., ROJEK M. 1993: Agrometeorologia. [Agrometeorology] Warszawa. PWN, pp. 250 (in Polish). 3. FEDDES R.A., LENSELINK K.J. 1994: Evapotranspiration. ILRI Publication 16. Drainage Principles and Applications. Wageningen, p. 145–173. 99 4. KACA E., ŁABĘDZKI L., CHRZANOWSKI, S,. CZAPLAK I., KASPERSKA-WOŁOWICZ W. 2003: Gospodarowanie zapasami wody użytecznej gleb torfowo-murszowych w warunkach regulowanego odpływu w różnych regionach agroklimatycznych Polski. Woda, Środowisko, Obszary Wiejskie. Rozprawy naukowe i mono-grafie nr 9, ss. 118. Falenty, Wydaw. IMUZ. [Managing useful water resources in peat-moorsh soils at a regulated outflow in different agro-climatic regions of Poland. Water-Environment-Rural Areas, Treatises & Monographs 9. Falenty: IMUZ, Poland, pp 108] (in Polish, summary in English). 5. KASPERSKA-WOŁOWICZ W., ŁABĘDZKI L. 2004: Porównanie ewapotranspiracji wskaźnikowej według Penmana i Penmana-Monteitha w różnych regionach Polski. Woda, Środowisko, Obszary Wiejskie t. 4 z. 2a (11) s. 123– –136. [A comparison of reference evapotranspiration according to Penman and Penman-Monteith in various regions of Poland. Water-Environment-Rural Areas 4(2a). Falenty: IMUZ, Poland, p. 123– 136] (in Polish, summary in English). 6. ŁABĘDZKI L. 1997: Potrzeby nawadniania użytków zielonych – uwarunkowania przyrodnicze i prognozowanie. Rozpr. Habil. Falenty: Wydaw. IMUZ ss. 120. [Grasslands water requirements – nature conditions and forecasting. Falenty: IMUZ, Poland, pp. 120] (in Polish). 7. ŁYKOWSKI B. 1986: Wskaźnik suchości klimatu P – E w nowym ujęciu. Problematyka melioracji w nauczaniu i badaniach naukowych. Warszawa: Wydaw. SGGW s. 59–68. [Indicator of climate dryness P – E in a new approach. Reclamation problems in teaching and research studies. Warszawa: SGGW, p. 59–68] (in Polish, summary in English). 8. ROGUSKIW., SARNACKAS., DRUPKA S. 1988: Instrukcja wyznaczania potrzeb i niedoborów wodnych roślin uprawnych i użytków zielonych. Mat. instruktażowe nr 66, Falenty: IMUZ ss. 90. [Guidelines 100 W. Kasperska-Wołowicz, L. Łabędzki for estimating water needs and water deficits for cultivated plants and grasslands. Instruction Materials no 66, Falenty: IMUZ, pp. 90] (in Polish). Streszczenie: Klimatyczny i rolniczy bilans wodny dla użytków zielonych w Polsce z wykorzystaniem metody Penmana-Monteitha. Celem pracy była ocena przestrzennego zróżnicowania wodnych bilansów klimatycznego i rolniczego dla dwukośnych użytków zielonych (dla plonów 5 i 7 Mg ha–1). Autorzy przeprowadzili analizę bilansów sporządzonych w okresie dekady, miesiąca oraz wegetacji dla 17 stacji meteorologicznych w różnych regionach Polski. W przeciętnym okresie wegetacji IV–IX wartość rolniczego bilansu wodnego jest ujemna dla plonów w wysokości 7 Mg ha–1 w całym kraju. Największe niedobory (powyżej 150 mm) obserwuje się w dorzeczu Warty. Niedobór dla plonów wysokości 5 Mg ha–1 jest o 40–47 mm mniejszy w porównaniu z plonami użytków zielonych w wysokości 7 Mg ha–1. MS. received December 2006 Authors’ address: Wiesława Kasperska-Wołowicz Leszek Łabędzki IMUZ Glinki 60, 85-174 Bydgoszcz Poland e-mail: [email protected] Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 101–110 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Variation of climatic water balance and heat balance for various ecosystems in Wrocław in the years 1964–2000 ELŻBIETA MUSIAŁ, JOANNA BUBNOWSKA, EDWARD GĄSIOREK Agricultural University of Wrocław, Department of Mathematics Abstract: Variation of climatic water balance and heat balance for various ecosystems in Wrocław in the years 1964–2000. The study contains a description of water relations, characterized by cumulative climatic water balance, in the region of Wrocław-Swojec. Heat relations during the vegetation period were described by heat balance of various active surfaces. The evaluation of climatic water balance variation was based on monthly sums of precipitation and potential evapotranspiration. The latter was calculated according to the Penman’s model for the warm half of the year (IV–IX) and according to the Tichomirow’s and Iwanow’s model for the cold half of the year (X–IX). The values of climatic water balance were assessed for the period from X to IX in every year, as well as for all years in common during the period 1964–2000. The perennial 1964–2000 in Wrocław-Swojec was characterized by potential evapotranspiration increase and a decreasing trend in yearly sums of atmospheric precipitation. The calculated values of cumulative climatic water balance proved the deepening water shortage in this area. Due to aggravating water deficit in Wrocław latent heat flux decreases and sensible heat flux, used for heating the atmosphere, increases, thus causing the warming up in this region. Key words: potential climatic water balance, potential evapotranspiration, heat balance, latent heat flux, sensible heat flux, Bowen’s ratio. about changes of climatic water balance in this region. The importance of climatic water balance in meteorology is due to the fact that net climatic water balance defines conditions for plant vegetation, influences changes of outflow indexes in river cachments and retention. The investigation was performed in years 1964–2000. Every analyzed year started in October and ended in September. Creation of cumulative climatic water balance from October to September gave the possibility to evaluate water overflow or deficit after autumn and winter, at the beginning of vegetation period and during this period until October. Due to the fact that climatic water balance is connected with heat balance through the stream vapour, the values of balance components were evaluated for coniferous forest, spring wheat and potatoes. The main aim of the study was to describe variations of water overflow and deficit after autumn and winter in subsequent years and consequences of water deficits in heat balance component values for examined ecosystems. INTRODUCTION Investigations on variation of heat balance components for various ecosystems, performed in the WrocławSwojec region, made the authors think METHODS The term of climatic water balance was first used in Polish literature by Bac and Rojek in 1970 [Bac, Rojek 102 E. Musiał, J. Bubnowska, E. Gąsiorek 1977, 1979, 1982]. The authors defined climatic water balance as a difference between precipitation and indicatory evapotranspiration [Rojek 1994], [Rojek, Wiercioch 1994, 1995]. The name climatic water balance is mainly connected with net balance assessed for the warm period (IV–IX). Contradictory to the above mentioned investigations, the authors of this study created a term of potential climatic water balance (PCWB). PCWB is defined as a difference between precipitation and potential evapotranspiration calculated by Penman, assessed in the warm half of the year (IV–IX). When the cold half of the year (X–III) is taken into account, PCWB is a difference between precipitation and potential evapotranspiration calculated by Iwanow and Tichomirow [Kędziora 1999], [Olejnik, Kędziora 1991]. Penman [Penman 1948, 1950, 1956, 1963] assumed that the density of latent heat flux used for evapotranspiration can be calculated in the following way: LE = ⎤ γ ⎡Δ ( Rn + G ) + Ea ⎥ = ⎢ Δ+γ⎣γ ⎦ Δ ( Rn + G ) + Ea γ = ⎡ Δ⎤ ⎢1 + γ ⎥ ⎣ ⎦ where : Rn – net radiation [Wm–2], G – soil heat flux [Wm–2], H – sensible heat flux [Wm–2], LE – latent heat flux [Wm–2], Ea – ability of air evapotranspiration [Wm–2], Ea = 7.44(1 + 0.54v)d, v – wind speed at 2 m height [ms–1], d – vapour pressure deficit [hPa], ∆ – mean rate of change of saturated vapour pressure with temperature [hPaK–1], γ – psychrometric constant γ = 0.655 [hPaK–1]. A simple relationship exists between potential evapotranspiration ETP expressed in [mm] and latent heat flux LE expressed in [Wm–2]: ETP = nLE/28.34, where n is a number of days in decade, in month. Climatic water balance, through the stream vapour transporting huge amount of energy to the atmosphere, is connected with active surface heat balance. The components of heat balance are not independent due to the rule of evapotranspiration priority in nature. Therefore, both deficit and excess in water, defined by cumulative potential climatic water balance at the beginning and during the vegetation period, decide on values of various heat balance streams. Thus, heat balance should be the next step of investigation. Heat balance of any active surface may be expressed by the following equation: Rn + LE + H + G = 0 where H – sensible heat flux [Wm–2], Rn, LE, G – like above. Heat balance components for various ecosystems were calculated according to the MBC [Kędziora 1999], [Olejnik, Kędziora 1991]. RESULTS AND DISCUSSION Seasonality of precipitation and potential evapotranspiration in the perennial 1964–2000 in Wrocław-Swojec is Variation of climatic water balance... shown in Figure 1 as sums of potential evapotranspiration and precipitation in the following months of the year. In our study we assume that one season is the period from October, when vegetation period ends, until September. Characteristics of precipitation and potential evapotranspiration seasonality is as follows: the highest sums of atmospheric precipitation and potential evapotranspiration in Wrocław-Swojec during the period 1964–2000 were seen in June, July and August. Moreover, apart from January, February, October and November, atmospheric precipitation was 103 lower than potential evapotranspiration. The analysis of mean sums of potential climatic water balance in this perennial revealed that net PCWB was negative in March, April, May, June, July, August and September. This may be due to the fact that the vegetation period in the above mentioned perennial was characterized by continuous water deficits. The description of 1964–2000 perennial in Wrocław-Swojec by yearly precipitation and potential evapotranspiration sums show that the last two decades of the XXth century were characterized by increasing potential FIGURE 1. Sums of precipitation (P) and potential evapotranspiration (ETP) in Wrocław-Swojec (1964–2000) FIGURE 2. Mean monthly sums of potential climatic water balance (PCWB) in Wrocław-Swojec (1964–2000) 104 E. Musiał, J. Bubnowska, E. Gąsiorek evapotranspiration and diminishing atmospheric precipitation (Fig. 3). Precipitation decrease and potential evapotranspiration increase resulted in enlarging net potential climatic water balance during the years 1964–2000 (Fig. 4). Tendencies of yearly changes in precipitation and potential evapotranspiration sums as well as basic statistical characteristics are shown in Table 1. In the 37-year period of observation in Wrocław-Swojec, potential evapotranspiration has shown a growing tendency both during the whole year and during the period from X–III. The last two decades of the XXth century revealed quicker increase of potential evapotranspiration than in the earlier period. However, the potential evapotranspiration rise in the period X–III was less pronounced than in the preceding years. Decreasing trends of yearly atmospheric precipitation sums and the sums in the period X–III, suggest, the aggravating atmospheric precipitation shortage in the last 20 years. The next step was the characteristics of the period 1964–2000 in WrocławSwojec, conducted in the perennial for every season from X to IX separately, by assessing cumulative potential climatic water balance. This characteristics looks as follows: in years 1964–1986 net PCWB, depending on precipitation, was alternately positive and negative FIGURE 3. Yearly sums of potential evapotranspiration and precipitation in Wrocław-Swojec (1964– –2000) FIGURE 4. Yearly sums of potential climatic water balance in Wrocław-Swojec (1964–2000) Variation of climatic water balance... 105 σp Linear regression equation for precipitation Tendency mm/ /10 year ETP (mm) Linear regression equation for evapotranspiration Tendency mm/10 year σETP 1980– –2000 P (mm) 1964– –2000 season Years TABLE 1. Basic statistical charateristics of evapotranspiration and precipitation in Wrocław-Swojec (1964–2000) X–IX I–XII X–III X–IX I–XII X–III 571 570 200 550 555 196 92 95 55 83 104 52 y = –1.8x+604 y = –1.9x+606 y = –0,5x+210 y = –2.3x+580 y = –3.6x+600 y = –0,7x+205 –18* –19* –5 –23* –36* –7 746 745 180 777 772 204 y = 4x+670 y = 3,9x+672 y = 2,8x+128 y = 5,7x+670 y = 6,9x+640 y = 1,4x+190 40** 39** 28** 57** 69** 14** 69 73 45 73 79 38 P – mean precipitation, σp – standard deviation for P, ETP-mean evapotranspiration σETP – standard deviation for ETP **) – statistically significant for α = 0.01 *) – statistically significant for α = 0.2 throughout the period X–IX. It was often seen that within the analyzed season net PCWB was positive until April, which means that precipitation covered the needs of potential evapotranspiration. Contrarily, since May net PCWB became negative, which was the sign of deepening water deficit. The detailed analysis starts in the year 1964/1965. Net potential climatic water balance (PCWB) was positive for the whole period from X to IX (precipitation 661 mm). Other seasons with positive net PCWB were as follows: 1967/1968 (precipitation 706 mm), 1970/1971 with 693.5 mm of precipitation and 1976/1977 (precipitation 722.6 mm). In the above mentioned years precipitation was higher than the mean value in the perennial (571 mm). The following seasons had negative net potential climatic water balance 1969/1970 with 525.6 mm of precipitation, 1972/1973 with 425 mm of precipitation, 1973/1974 with 431.4 mm of precipitation, 1977/1978 with 599.2 mm of precipitation and 1983/1984 with 544.3 mm of precipitation. In other seasons of the perennial 1964-1986 potential climatic water balance was positive until April, whereas after that period became negative. Since 1986/1987, when net potential climatic water balance was positive, in every coming year water deficit was growing. Even in 1997, the year of flood in Wrocław, precipitation in July did not manage the water shortage to vanish. Deepening water deficits are clearly seen in the last two decades of the XXth century (Fig. 5a, b, c, d, e, f). Water deficit or excess during vegetation period have significant influence on the values of heat balance components. It follows from the rule of evapotranspiration priority, that if there is sufficient amount of water in the substrate, the excess of energy is used for evapotranspiration prior to heating air and soil. Therefore, this rule connects 106 E. Musiał, J. Bubnowska, E. Gąsiorek a 1986/1987 150 100 50 0 X X-XI X-XII X-I X-II X-III X-IV X-V X-VI X-VII X-VIII X-IX X-V X-VI X-VII X-VIII X-IX X-V X-VI X-VII X-VIII X-IX X-V X-VI X-VII X-VIII X-IX -50 b 1987/1988 100 50 0 -50 -100 -150 -200 -250 X X-XI X-XII X-I c X-II X-III X-IV 1988/1989 0 -100 X X-XI X-XII X-I X-II X-III X-IV -200 -300 -400 -500 d 1996/1997 0 X -50 -100 -150 -200 -250 X-XI X-XII X-I X-II X-III X-IV Variation of climatic water balance... e 107 1997/1998 100 0 X X-XI X-XII X-I X-II X-III X-IV X-V X-VI X-VII X-VIII X-IX X-V X-VI X-VII X-VIII X-IX -100 -200 -300 f 1999/2000 50 0 -50 -100 -150 -200 -250 -300 X X-XI X-XII X-I X-II X-III X-IV FIGURE 5. Changes of potential climatic water balance (PCWB) in Wrocław-Swojec for the period X–IX in years 1986–2000 climatic water balance and heat balance. Thus, in the following investigations heat balance components and their contribution in net radiation were calculated for ecosystems of coniferous forest, potatoes and spring wheat (Fig. 6, 7, 8). The consequences of deepening water deficits in Wrocław-Swojec during the last 20 years are as follows : 1. Decreasing latent heat flux used for evapotranspiration; 2. Slow increase of sensible heat flux used for heating atmosphere. Increasing values of Bowen’s ratio (H/LE) [Bowen 1926] for coniferous forest, potatoes and spring wheat prove sensible heat flux enlargement in the last 20 years (Fig. 9). The indicated above tendencies suggest the climate warming up in this area. CONCLUSIONS 1. The 1964–2000 perennial in Wrocław-Swojec is characterized by increased yearly sums of potential evapotranspiration and decreasing trend in yearly precipitation sums. 2. The analysis of a yearly course of potential cumulative climatic water balance in the season from October to September revealed growing water deficits in this region, more evident in the last decades of the XXth century. 3. Growing water deficits in this area are the cause of increasing sensible heat flux in the vegetation period for coniferous forest, spring wheat and potatoes, as well as decreasing tendency for latent heat flux. 108 E. Musiał, J. Bubnowska, E. Gąsiorek FIGURE 6. Variation of mean ten-days values of sensible heat flux (H) during the growing season four coniferous forest, potatoes and spring wheat in Wrocław-Swojec (1964–2000) FIGURE 7. Variation of mean values of sensible heat flux and net radiation ratio (H/Rn) during the growing season for coniferous forest, potatoes and spring wheat in Wrocław-Swojec (1964–2000) Variation of climatic water balance... 109 FIGURE 8. Variation of mean values of latent heat flux and net radiation ratio (LE/Rn) during the growing season for coniferous forest, potatoes and spring wheat in Wrocław-Swojec (1964–2000) FIGURE 9. Variation of mean ten-days values of the Bowen’s ratio (H/LE) during the growing season for coniferous forest, potatoes and spring wheat in Wrocław-Swojec (1964–2000) 110 E. Musiał, J. Bubnowska, E. Gąsiorek 4. Due to increasing values of sensible heat flux used for heating the atmosphere, the air temperature in this region rose 0,03oC per year, which means warming up in this area. REFERENCES BAC S., ROJEK M. 1977: Metodyka oceny stosunków wodnych obszarów rolniczych na podstawie danych klimatycznych, Zesz. nauk ART Olszt. Nr 21: 13–24. BAC S., ROJEK M. 1979: Klimatyczny bilans wodny a odpływy w Polsce, Przegl. Geofiz. 24(3): 293–297. BAC S., ROJEK M. 1982: Klimatyczne bilanse wodne w Polsce. [w:] Bac S. (red.) Agroklimatyczne podstawy melioracji wodnych w Polsce. PWRiL, Warszawa. BOWEN I.S. 1926: The ratio of heat losses by conduction and by evaporation from any water surface. Phys. Rev., 27, p. 779–787. KĘDZIORA A. 1999: Podstawy agrometeorologii, PWRiL, Poznań. OLEJNIK J., KĘDZIORA A. 1991: A model for heat and water balance estimation and its application to land use and climate variation., Earth Surface Processes Landforms vol. 16 ss. 601–617. PENMAN H.L. 1948: Natural evaporation from open water, bare soil and grass. London: Proc.Royal Soc. Vol. 193, 120– 146. PENMAN H.L. 1950: Evaporationover the BritishIsles, Q.J. Roy. Met. Soc., 76 372–83. PENMAN H.L. 1956: Evaporation: an introductory survey, Netherlands J. Agric. Sci. 4, –29. PENMAN H.L. 1963: Vegetation and Hydrology, Tech. Comm. Nr 53, Comm. Bur. of Soils, Harpenden. ROJEK M.M., WIERCIOCH T. 1995: Zmienność czasowa i przestrzenna parowania wskaźnikowego , ewapotranspiracji aktualnej i niedoborów opadowych w Polsce nizinnej w okresie 1951– –1990, ZNAR Nr 268, Monografie VI. ROJEK M. 1994: Time variablity of climatic water balances in selected meteorological stations in Poland. Zesz. Probl. Post. Nauk Rol., 405, 147–153. ROJEK M., WIERCIOCH T. 1994: Indicatory evapotranspiration os summer half-year in various long-term periods. Zesz. Probl. Post. Nauk Rol., 405, 155–161. Streszczenie: Zmienność klimatycznego bilansu wodnego i bilansu cieplnego różnych ekosystemów dla Wrocławia w latach 1964–2000. Praca zawiera charakterystykę składowych bilansu klimatycznego bilansu wodnego we Wrocławiu w latach 1964– –2000. W analizowanym okresie obserwuje się trend wzrostu rocznych sum ewapotranspiracji potencjalnej oraz spadku rocznych sum opadów atmosferycznych. W związku ze wzrostem strumienia ciepła jawnego, ogrzewającego atmosferę obserwuje się wzrost średniej rocznej temperatury powietrza o 0,03°C. Ten wzrost temperatury związany jest ze wzrostem deficytu wody dla lasu liściastego, pszenicy jarej i ziemniaków oraz rosnącym strumieniem ciepła jawnego przeznaczonego na ogrzanie atmosfery. MS received December 2006 Authors address: Elżbieta Musiał, Joanna Bubnowska, Edward Gąsiorek Department of Mathematics Agricultural University of Wrocław ul. Grunwaldzka 53, PL-50357 Wrocław Polska The research supported by KBN grant in years 2004–2007. Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 111–121 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Investigation for biological nitrogen removal from wastewater using simultaneous nitrification/denitrification technology GIEDRÉ VABOLIENÉ Department of Water Supply and Management, Vilnius Gediminas Technical University, Lithuania Abstract: Investigation for biological nitrogen removal from wastewater using simultaneous nitrification/denitrification technology. Biological nitrogen removal from the wastewater is based on the nitrification and denitrification processes at the biological treatment plant with the activated sludge. Different technological schemes can be used for the above mentioned processes. “BioBalance” technology as the newest way of nitrogen and phosphorus removal has been applied at Utena Wastewater Treatment Plant. “BioBalance Symbio” technology for the nitrogen removal is based on the active sludge technology with simultaneous nitrification/denitrification. The aeration zone and the anoxic zone are in one tank. The nitrification and denitrification are carried out during the aeration switching on and off. Nitrification and denitrification processes have been estimated during five experiments in the aeration tanks, when durations of aeration and low aeration were from 120 to 180 min. The impact of aeration regime on biological nitrogen removal has been estimated during this scientific work. Key words: biological nitrogen removal, nitrification, denitrification, nitrate, ammonium nitrogen, biological active potential (BPA). INTRODUCTION Biological nitrogen removal from the wastewater is based on the nitrification and denitrification processes at the biological treatment plant with the activated sludge (Berţinskienë 1999). The nitrification is carry out by the nitrification bacteria, which are divided into two groups. The first group bacteria (Nitrosomonas, Nitrosospira, Nitrosococcus, Nitrosolobus genera) oxidise the ammonium hydrate to nitrite. The second group bacteria (Nitrobacteria, Nitrospira, Nitrococcus genera) oxidise nitrite into nitrate (Bitton 1994). The energy is produced during both stages of the nitrification. Nitrification bacteria consume the energy for CO2 assimilation. Thus nitrification bacteria are autotrophic to the carbon. Nitrification bacteria grow significantly slower than other bacteria during the intensive nitrification process (Droste 1997). The nitrogen cycle ends by its return back to the atmosphere in nature i.e. by denitrification. Denitrification is one of anaerobic respiration variation (nitrate respiration), when NO3 ions are being used as a final electrons acceptor for the oxidation of organic substances. The denitrification is oxygen split from the nitrate, after that – from nitrite by denitrification bacteria. The denitification is being carried out by heterotrophic bacteria: pseudomonas, spirillium, tiobacillus, alkaligenes, bacillus. They use organic pollution, existing in the wastewater as carbon source. There are two main stages in the denitrification. None of the stages will take place till there will be any sufficient amount of the dissolved oxygen. The first is carrying out 112 G. Vaboliené nitrate reduction up to nitrite; the second is reducing nitrite up to nitrogen gas. During the second stage nitrites change nitrate as electrons acceptors, and the nitrogen gas changes nitrite as a product to finish the denitrification reaction. Different technological schemes, following the source of carbon oxidation during denitrification processes, can be used for the above mentioned processes (Henze et al. 1995). The 1st scheme, i.e. carbon source – carbon of raw wastewater. The scheme is composed of the denitrification tank, nitrification tank, and secondary clarifiers. The 2nd scheme, i.e. carbon sources – additionally added easily biodegradable organic substances (methanol, ethanol, etc.). The scheme is composed of the nitrification tank, denitrification tank, additional aeration tank, and secondary clarifiers. The 3rd scheme, i.e. carbon source – carbon of raw wastewater. According to this scheme the aeration zone and the anoxic zone are in one tank. The nitrification and the denitrification is carry out during the aeration switching on and off (Matuzevičius et al. 1998). Usually schemes of the nitrogen removal are combined with phosphorus removal schemes. Different new technologies are used to decrease phosphorus and nitrogen quantity in the wastewater. The mostly advanced is “BioBalance” technology, where nitrogen is removed following the 3rd nitrogen removal scheme, and the anaerobic zone is equipped before the nitrification/ denitrification tank the phosphorus biological removal. The nitrogen removal according to “BioBalance Symbio” technology carried out in a simultaneous way of nitrification and denitrification; controlled the oxygen supply by NADH sensor according to sludge activity. Commonly, there exist co-ferments that function as suppliers of hydrogen and electrons and are nicotinamideadeninedinucleotide NAD+ and its phosphoresced compound derivative NADP+. Reduced NAD(P)H form, can be oxidised once more during the formation of ATP. During the processes of metabolism co-ferments NADH and NADPH are produced in all microorganisms. NAD(P)H quantity in the microorganisms produced depend on their activity. The abovementioned activity depends on existing sludge loads, supplied in the form of nutritious matters. This means that, for example, NAD(P)H quantity will remain stable under the conditions of constant quantity of nutritious matters supplied to the active sludge system, the constant load and sludge activity. The sludge activity and at the same time NAD(P)H production increase gradually at the same time with an increase of the sludge load because of a more intensive supply of nutritious matters. Differently, NAD(P)H production decreases during a decrease of the nutritious matter supply. During research this was determined by linear dependences (Norgard et al. 1996). “BioBalance” measuring equipment NADH fluorescensor is based on the process mentioned above. The fluorecsensor fixes only NADH that gives information on how much energy the microorganisms possess. The information on the energy is defined as the biological active potential (BPA). NADH fluorescensor controls denitrification and nitrification process according to the sludge activity in a single aeration Investigation for biological nitrogen removal... tank. Ammonia is oxidised to nitrates by nitrifying bacteria during the nitrification phase. The nitrates are reduced to molecular nitrogen (N2) by denitrifying bacteria during the denitrification phase. These processes are carried out periodically. Organic matter is oxidised bacterially during both the nitrification and denitrification phases, with oxygen and nitrates, respectively, as the oxidising agents and phosphorus is absorbable with special microorganisms. The aim of the work was to evaluate the impact of aeration regime on biological nitrogen removal from wastewater using simultaneous nitrification/denitrification technology. MATERIALS AND METHODS Researches have been carrying out during period (July–November, 2005) at Utena Wastewater Treatment Plant in Lithuania. Two aeration tanks have been used for the biological wastewater treatment in Utena wastewater treatment plant. The nitrogen removal according to “BioBalance Symbio” technology carried out in a simultaneous way of nitrification and denitrification; controlled the oxygen supply by NADH sensor according to sludge activity. 113 Two aeration tanks worked at the same time. Wastewater-flow after mechanical treatment was divided into both aeration tanks simultaneously. However, the regime of aeration in each aeration tank was different. Aeration can be carried out in one aeration tank, in the meantime low aeration can be carried out in the other aeration tank. The term of aeration and low aeration can be different in both aeration tanks. Because of the abovementioned reasons, this research’s results were analysed of both aeration tanks. Low oxygen concentration 0.1–0.5 mg O2 l–1 has been held until the denitrification carries out and the higher oxygen concentration 0.5–1 mg O2 l–1 has been maintained for the nitrification processes carry out periodically. Conditions for the nitrification process were held from 120 to 180 min, and after that 120–180 min were spent for the denitrification. The duration of the nitrification and the denitrification had been held in five different regimes during experiments (Tab. 1). The researches from 2 to 9 occasions in each aeration tank have been carrying out during different aeration regime. The total nitrogen concentration in wastewater after the mechanical treatment, ammonium nitrogen and nitrate concentration in the beginning of aeration and by the end of aeration rate TABLE 1. The duration of the nitrification and the denitrification during experiments Number of experiment The duration of aeration rate, min The duration of low aeration rate, min 1 150 150 2 4 180 180 120 120 150 120 120 180 5 150 180 3 114 G. Vaboliené in the both aeration tanks, total nitrogen concentration in wastewater after biological treatment had been measured during all experiments. The active sludge concentration and the volatile suspended solids of active sludge in the aeration tank had been measured. The average flow in the aeration tank had been fixed. The efficiency of total nitrogen removal and the active sludge load in the aeration tank had been estimated. Total nitrogen had been estimated adding Kjeldahl nitrogen to nitrites and nitrates nitrogen. All analysis was carried out using standard methods (LST ISO 5815:1989). RESULTS Total nitrogen, [mgN/l] The total nitrogen concentration in wastewater after the mechanical treatment fluctuated from 37 to 52 mg/l, average 44 mg/l during five experiments (Fig. 1). The first experiment carries out 21 day during period (July 19 – August 8). Conditions for the nitrification carried out were held 150 min, and 150 min – for the denitrification. The researches were carrying out 9 times in the both aeration tanks. The gathered results proved that the efficiency of total nitrogen removal fluctuated from 87 to 91%, average 88%. At start of experiment, the concentration of ammonium nitrogen in the beginning of aeration rate fluctuated from 0.84 to 1.4 mg/l, by the end of aeration rate – from 0.47 to 0.91 mg/l (Fig. 2). However, the concentration of ammonium nitrogen 1 mg/l was estimated in one aeration tank by the end of aeration rate after 7 day of experiment start. After that, aeration tanks worked at mentioned above aeration regime yet one week. Then it was estimated that partial nitrification was carried out in the both aeration tanks by the end of aeration rate. The concentration of ammonium nitrogen by the end of aeration rate only increased from 1.4 to 2 mg/l till experiment was finished. Following experimental results can be seen, that the duration of aeration was not enough. The efficiency of total 55 50 45 40 35 30 1 2 3 4 5 6 7 8 9 aeration and reduced aeration rate 150 min aeration rate 180 min, reduced aeration rate 150 min aeration rate 180, 120 min, reduced aeration rate 120 min aeration rate 120 min, reduced aeration rate 180 min aeration rate 150 min, reduced aeration rate 180 min FIGURE 1. Total nitrogen concentration in wastewater after the mechanical treatmen Investigation for biological nitrogen removal... N-NH4 , N-NO3,[mg/l] a 115 3 2,5 2 1,5 1 0,5 0 1 2 3 4 5 6 7 8 9 10 b N-NH4 , N-NO3 [mg/l] N-NH4 in the beginning of aeration rate N-NO3 in the beginning of aeration rate N-NH4 by the end of aeration rate N-NO3 by the end of aeration rate 2,5 2 1,5 1 0,5 0 1 2 3 4 5 6 7 8 9 N-NH4 in the beginning of aeration rate N-NO3 in the beginning of aeration rate N-NH4 by the end of aeration rate N-NO3 by the end of aeration rate FIGURE 2. The concentration of ammonium nitrogen and nitrate in the beginning and by the end of aeration rate at first (a) and second (b) aeration tank when aeration regime: aeration and reduced aeration rate – 150 min nitrogen removal decreased from 91% at experiment start to 87% at the experiment finish. The concentration of nitrate was changing in small interval during first experiment: in the beginning of aeration rate 0.1÷0.6 mg/l, by the end of aeration rate 0.45÷1.2 mg/l (Fig. 2). A low concentration of nitrate in the beginning of aeration rate provided that 150 min. reduced aeration rate was enough for complete denitrification carry out. Consequently results of first experiment indicated the condition of second experiment. The second experiment carries out 11 day during period (August 9–19). Conditions for the nitrification carried out were held 180 min, and 150 min – for the denitrification. It was extended duration of aeration for complete nitrification. The researches were carrying out 8 times in the both aeration tanks. The obtained results proved that the efficiency of 116 G. Vaboliené total nitrogen removal fluctuated from 83 to 89%, average 85%. During two days of the second experiment the concentration of ammonium nitrogen in the beginning of aeration rate decreased from 3,1 to 0.8 mg/l (Fig. 3). Process of nitrification carries out completely. Duration of aeration was enough till the end of experiment. The concentration of nitrate changed more than during first experiment. The concentration of nitrate by the end of aeration rate exceeded 1 mg/l several time. The efficiency of a total nitrogen removal decreased from 89% at experiment start to 83% at the experiment finish. The third experiment carries out during 7 days period (August 22–28). Aeration tanks worked in follow aeration regime: aeration rate – 180 min, reduced aeration rate – 120 min first 5 day of experiment. Later aeration regime was changed: aeration rate – 120 min, reduced aeration rate – 120 min. The researches were carrying out 5 times in first aeration regime and 2 times in second N-NH4 , N-NO3 , [mg/l] 3,5 3 2,5 2 1,5 1 0,5 0 1 2 3 4 5 6 7 8 N-NH4 in the beginning of aeration rate N-NO3 in the beginning of aeration rate N-NH4 by the end of aeration rate N-NO3 by the end of aeration rate b N-NH4, N-NO3 , [mg/l] 2,5 2 1,5 1 0,5 0 1 2 3 4 5 6 7 8 N-NH4 in the beginning of aeration rate N-NO3 in the beginning of aeration rate N-NH4 by the end of aeration rate N-NO3 by the end of aeration rate FIGURE 3. The concentration of ammonium nitrogen and nitrate in the beginning and by the end of aeration rate at first (a) and second (b) aeration tank, when aeration regime: aeration rate 180 min, reduced aeration rate – 150 min Investigation for biological nitrogen removal... aeration regime in the both aeration tanks. The gathered results proved that the efficiency of total nitrogen removal fluctuated from 73 to 80%, average 77%. Though the concentration of ammonium nitrogen in the beginning of aeration rate fluctuated from 0.13 to 2.2 mg/l, but it already decreased and fluctuated from 0.01 to 0.96 mg/l by the end of aeration rate (Fig. 4). However nitrate increased signally by the end of aeration rate. The concentration of nitrate was estimated from 0.47 to 11 mg/l in the beginning of aeration rate and 1.2÷17 mg/l by N-NH4 , N-NO3 , [mg/l] a 117 the end of aeration rate. The reduced aeration rate was preferred for complete denitrification, so the efficiency of total nitrogen removal decreased to 73%. Then the aeration regime was changed and aeration tanks worked in other regime (aeration rate – 120 min, reduced aeration rate – 120 min) last 2 day of experiment for reduction of nitrate. Then aeration rate was changed from 180 min to 120 min concentration of nitrate decreased to 0.75 mg/l in the beginning and by the end of aeration rate during 2 18 16 14 12 10 8 6 4 2 0 1 2 3 4 5 6 7 N-NH4 in the beginning of aeration rate N-NO3 in the beginning of aeration rate N-NH4 by the end of aeration rate N-NO3 by the end of aeration rate N-NH4 , N-NO3 , [mg/l] b 18 16 14 12 10 8 6 4 2 0 1 2 3 4 5 6 7 N-NH4 in the beginning of aeration rate N-NO3 in the beginning of aeration rate N-NH4 by the end of aeration rate N-NO3 by the end of aeration rate FIGURE 4. The concentration of ammonium nitrogen and nitrate in the beginning and by the end of aeration rate at first (a) and second (b) aeration tank, when aeration regime: aeration rate 180, 120 min, reduced aeration rate – 120 min 118 G. Vaboliené day of experiment. The efficiency of total nitrogen removal increased to 73%. Next experiment was carrying out interchanging aeration regime. Aeration rate reduced from 180 min. to 120 min., and reduced aeration rate prolonged from 120 min. to 180 min. for reduction of nitrate. The fourth experiment carries out 7 day during period (August 29–September 4). The researches were carrying out 7 times in the both aeration tanks. The gathered results proved that the efficiency of total nitrogen removal fluctuated from 73 to a 88%, average 83%. The nitrification carried out completely first five day of experiment. The concentration of ammonium nitrogen fluctuated from 0.31 to 1.4 mg/l, 0.02–0.7 mg/l in the beginning and by the end of aeration rate respectively (Fig. 5). However later (two last day of experiment) the concentration of ammonium nitrogen increased in the beginning and by the end of aeration rate. Only partial nitrification carried out. Consequently concentration of nitrate decreased progressively and decreased until 0.5, 0.31 mg/l in the beginning of N-NH4, N-NO3, [mg/l] 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 N-NH4 in the beginning of aeration rate N-NO3 in the beginning of aeration rate N-NH4 by the end of aeration rate N-NO3 by the end of aeration rate b N-NH4, N-NO3, [mg/l] 10 8 6 4 2 0 1 2 3 4 5 6 7 N-NH4 in the beginning of aeration rate N-NO3 in the beginning of aeration rate N-NH4 by the end of aeration rate N-NO3 by the end of aeration rate FIGURE 5. The concentration of ammonium nitrogen and nitrate in the beginning and by the end of aeration rate at first (a) and second (b) aeration tank, when aeration regime: aeration rate 120 min, reduced aeration rate – 180 min Investigation for biological nitrogen removal... aeration rate at the each aeration tanks respectively and 0.89 and 0.33 mg/l by the end of aeration rate at the each aeration tanks respectively during third day of experiment. Enough duration of reduced aeration for completely denitrification had positive impact on efficiency of total nitrogen removal, which increased to 88% later decreased to 82%. The fifth experiment carries out 74 day during period (September 5 –November 18). Conditions for the nitrification carried out were held 150 a 119 min, and 180 min – for the denitrification. The researches were carrying out 7 times in the both aeration tanks. The gathered results proved that the efficiency of total nitrogen removal fluctuated from 80 to 89%, average 85%. Following concentration of ammonium nitrogen 0.44–4.6 mg/l and 0.06÷0.65 mg/l in the beginning and by the end of aeration rate respectively, nitrification process carry out completely (Fig. 6). However concentration of nitrates (3.8 mg/l by the end of aeration rate) showed partial N-NH4, N-NO3, [mg/l] 5 4 3 2 1 0 1 2 3 4 5 6 7 N-NH4 in the beginning of aeration rate N-NO3 in the beginning of aeration rate N-NH4 by the end of aeration rate N-NO3 by the end of aeration rate b N-NH4, N-NO3, [mg/l] 4 3,5 3 2,5 2 1,5 1 0,5 0 1 2 3 4 5 6 7 N-NH4 in the beginning of aeration rate N-NO3 in the beginning of aeration rate N-NH4 by the end of aeration rate N-NO3 by the end of aeration rate FIGURE 6. The concentration of ammonium nitrogen and nitrate in the beginning and by the end of aeration rate at first (a) and second (b) aeration tank, when aeration regime: aeration rate 150 min, reduced aeration rate – 180 min 120 G. Vaboliené denitrification process. It is proved that the duration of reduced aeration was not enough for denitrification. The efficiency of total nitrogen removal decreased from 89% at experiment start to 80% at the experiment finish. The efficiency of total nitrogen removal fluctuated from 73 to 91% during five experiment when aeration tank worked in different aeration regime (Fig. 7). The total nitrogen concentration in wastewater after the biological treatment fluctuated from 3.9 to 12 mgN/ /l during five experiments. CONCLUSIONS Efficiency of nitrogen removal,% 1. Using biological nitrogen removal technologies, when nitrification and denitrification are in process changing aeration intensiveness, it is very important properly evaluate aeration and low aeration duration. 2. When total nitrogen concentrations of mechanically treated wastewater ranged from 37 till 52 mgN/l, it will be proper to use two aeration regimes: 150 minutes aeration and 150 minutes low aeration rates, similarly 180 minutes aeration and 150 minutes low aeration rates. 3. During 120 minutes conditional short aeration rate and prolonged low aeration rate, presently ammonium nitrogen increase, following partial denitrification outcome. Working in this regime it is possible to obtain completed nitrification. 4. During experimental research it was obtained, when aeration and low aeration rates are similar or slightly different approximately 30 minutes, nitrogen removal effectiveness is the biggest. ACKNOWLEDGEMENT This research study is dedicated to EC financed FP6 project MAPO „Enhancing Research and Development Projects to find Solutions to Struggle against various Marine Pollutions”. The author is involved into deliverable on innovation roadmap for technologies based on 95 90 85 80 75 70 1 2 3 4 5 6 7 8 9 aeration and reduced aeration rate 150 min aeration rate 180 min, reduced aeration rate 150 min aeration rate 180, 120 min, reduced aeration rate 120 min aeration rate 120 min, reduced aeration rate 180 min aeration rate 150 min, reduced aeration rate 180 min FIGURE 7. The efficiency of total nitrogen removal when aeration tank worked in different aeration regime Investigation for biological nitrogen removal... gaps within current state-of-the-art. I’m gratefully to all Small Medium Enterprises participated into MAPO project activities; more Small Medium Enterprises working into related to marine pollutions areas are welcome to join future activities. REFERENCES BERŢINSKIENË J. 1999: Water microbiology. Textbook Vilnius: Technika, p. 144. HENZE M., HARREMOES P., JANSEN C., ARVIN E. 1995: Wasterwater Treatment. Biological and Chemical Processes. Springer-Verlag, p. 383. BITTON G. 1994: Wastewater microbiology. Wiley-Liss, New York, p. 456. DROSTE R. 1997: Theory and practice of water and wastewater treatment, p. 954. MATUZEVIČIUS A., PAULAUSKIENË Z. 1998: Experimental researches of phosphorus and nitrogen consumption and nitrogen removal from the wastewater in the biological treatment plant. 3rd International Conference. Cities Engineering and Environment, VGTU, “Technika”, p. 137–142. NORGARD P., HELMO K., SORENSER E. 1996: Purification process for nitrogen removal controlled by NADH. Vand og Jord, Danish, vol 3, p. 126–129. TSAI M.W., WENZEL M.C., EKAMA G.A. 2003: The effect of residual ammonia concentration under aerobic conditions on the growth of Microthrix parvicella in biological nutrient removal plants. Water Research 37, p.p. 3009–3015. HATZICONTINOU G.J., ANDREADKIS A. 2002: Differences in nitrification potential between fully aerobic and nitrogen removal activated sludge systems. Water Science & Technology. Vol. 46, No 1/2, p. 297–189. JENICEK P., SVEHLA P., ZABRANSKA J., DOHAYOS M. 2004: Factors affecting nitrogen removal by nitrification/ 121 /denitrification. Water Science & Technology. Vol. 49, No 5/6, p. 73–79. Lithuanian Ministry of Environment Protection. 1994: Unified methods of the wastewater and surface water quality researches. Chemical analysis methods. Part I. 224 p. Lithuanian Ministry of Environment Protection, 2002: LAND 47-1:2002. (ISO 5815:1989), 17 p. Lithuanian Ministry of Environment Protection, 2000: LAND 32-2000.10 p. Lithuanian Ministry of Environment Protection, 2005: LAND 66-2005. 7 p. Lithuanian Ministry of Environment Protection, 2005: LAND 58:2003. 24 p. Streszczenie: Badania biologicznego usuwania azotu ze ścieków za pomocą technologii jednoczesnej nitryfikacji i denitryfikacji. Metoda biologicznego usuwania azotu ze ścieków oczyszczanych w komorach osadu czynnego opiera się na procesach nitryfikacji i denitryfikacji. Do wyżej wymienionych procesów można stosować różne ciągi technologiczne. Technologia „BioBalance” zaliczana do najnowszych metod usuwania azotu i fosforu została wdrożona w oczyszczalni ścieków w Utenie (Litwa). Technologia „BioBalance Symbio” do usuwania azotu wykorzystuje metodę osadu czynnego z jednoczesnym prowadzeniem procesów nitryfikacji i denitryfikacji. Strefy tlenowa i beztlenowa znajdują się w jednej komorze. Procesy nitryfikacji i denitryfikacji są prowadzone poprzez naprzemienne włączanie i wyłączanie napowietrzania komory. W pracy przedstawiono wyniki badań przebiegu procesów nitryfikacji i denitryfikacji podczas pięciu cykli badań w komorze napowietrzania w warunkach intensywnego i wolnego napowietrzania. W pracy oszacowano wpływ warunków napowietrzania na skuteczność usuwania biologicznego azotu. MS. received July 2006 Author’s address: Giedré Vaboliené Vilnius Gediminas Technical University, Dept. of Water Supply and Management, Saulëtekio al. 11, AIF, LT-10228, Vilnius-40, Lithuania. e-mail: [email protected] Annals of Warsaw Agricultural University – SGGW Land Reclamation No 37, 2006: 123–128 (Ann. Warsaw Agricult. Univ. – SGGW, Land Reclam. 37, 2006) Development of technologies used in agricultural engineering work on an example of selected stages of land consolidation process URSZULA LITWIN*, JAROSŁAW JANUS*, MARIUSZ ZYGMUNT** *Department of the Geodesical Arrangement of Rural Settlements **Department of Geodesy, Agricultural University of Cracow Abstract: Development of technologies used in agricultural engineering work on an example of selected stages of land consolidation process. The land consolidation process is one of the most time-consuming and complicated agricultural engineering tasks performed by geodesists. Their automation and computerisation is not as advanced as for other types of geodetic works. The reason for that is a limited number of works that have been performed in this range since the mid 1980s, i.e. during the period of intensive IT (Information Technology) implementation in geodesy. At present one can observe again the growth in interest in performance of land consolidations processes. The purpose of this elaboration is to introduce the latest solutions IT applied in land consolidation process on an example of the land consolidation in Wojków village. Three characteristic stages of the process are presented herein: creation of assessment register before consolidation, preliminary selection of plots for designed blocks and designing. Proposal for a method for automation of the above-mentioned activities using a digital map developed by means of Bentley MicroStation software, retaining the requirements imposed by land consolidation instruction, was discussed. The above-mentioned stages are illustrated by examples from the land consolidation carried out in Wojków village in years 2001–2004. The work was performed on these premises using the technology presented herein. It allowed reduction of the time of land consolidation process by approx. 30%. Key words: spatial structure of village, land consolidation. INTRODUCTION The efficient use of rural areas has become one of the key problems in the age of Poland’s membership of the European Union. To use the potential of rural areas effectively, among other things a significant improvement in the structure of the areas is required. One of the problems to be solved is an excessive dispersion and fragmentation of farms [Urban 1981, Hopfer and Urban 1984, Harasimowicz 1995]. The stagnation in performance of the land consolidation process that occurred in the 1980s has intensified the problem. Only the increased interest in land consolidation process and related prospects for significant growth in the number of consolidations resulted in appearance of the first IT tools to automate some of the process’s stages [Janus and Zygmunt 2005]. This project describes selected functionalities of the system supporting modern land consolidation process, which is the only system in Poland to treat this process in a comprehensive manner. 124 U. Litwin, J. Janus, M. Zygmunt CREATION OF DESIGNED BLOCKS The land consolidation process the new system was tested on was carried in Wojków village, Mielec district in years 2001–2004. The area of the consolidated land was 670 ha and comprised of 1503 plots. There was no digital map for the area, therefore the first thing was to convert the contents of analogue map into the digital form. Bentley Systems software was used for this purpose. The digital map was checked for correctness by the software the purpose of which was to pick up any topological errors. Another stage of the check was to compare a set of plots on the map to the plots included in the land register survey. The checks ensured that register of lands before consolidation was created properly. To facilitate the work, the software had the functionality of semi-automatic creation of maps with comparative assessment. The borderlines of assessment contours occurred as a result of automatic FIGURE 1. Division of object into 9 parts selection of appropriate borders of plots and facilities, such as: roads, rivers and railway lines. For a digital map prepared like that the software automatically generates register of lands before consolidation. The register is prepared so that it can be printed in A3 format [Litwin and Zygmunt 2005]. To enable multi-station operation the object was divided into 9 designed areas. The software has functionalities allowing this operation to be performed automatically (Fig. 1). Each of the newly created designed areas was then divided into designed blocks (Fig. 2). Division into blocks may also be carried out automatically. Information on how the lines of assessment contours run was used to define boundaries of designed blocks. Relevant function of the software defines a new boundary upon the assessment contour number is indicated. When designing the technology, complete latitude in definition of the boundaries of designed blocks was adopted. The only condition Development of technologies used in agricultural engineering... 125 FIGURE 2. Division of designed blocks within area no 01-Wojków to be met by the designed block’s boundary is topological correctness. The topological correctness is controlled at the operator’s request and any errors are immediately communicated with their precise location. PRELIMINARY SELECTION OF PLOTS FOR DESIGNED BLOCKS Another stage of a land consolidation process that has been automated in a comprehensive manner is the preliminary selection of plots for designed blocks. This procedure is one of more important stages of a land consolidation process. It is performed after the land assessment register is approved and before the stage of designing a new land arrangement is commenced. The purpose of the stage is to make the initial assignment of lands in individual farms to dedicated designed blocks. The number and sizes of the blocks most often result from the run of boundaries of designed invariants, such as boundaries of roads, watercourses, administrative boundaries and boundaries of building sites. The essential problem that occurs during this operation is the necessity of considering all participants’ requests regarding the location of the newly divided lands. The appropriate performance of preliminary selection activities is impeded by the following factors: – large number of participants in the process (in case of land consolidations carried out over large areas it may reach a few thousands), most often exceeding the number of separated designed blocks by several dozen times; – necessity of considering diversification in designed blocks of classes and arable lands when collecting requests; – necessity of as even collection of requests for individual blocks as possible so that further design of a new land arrangement in accordance with this selection would be feasible. 126 U. Litwin, J. Janus, M. Zygmunt To facilitate this operation the process participants submit their requests in more than one version to increase the likelihood of proposing such a land arrangement that the expectations of most participants could be come up. In the time of domination of analogue technologies, this stage of the process was extremely time-consuming and any errors made during the stage were decisive for any other operations carried out during the whole land consolidation process. In particular, making changes to the already submitted requests and control of acquired data took a lot of time. The purpose of the solutions designed is a comprehensive support for the initial selection stage, in particular: – supporting the negotiations with parties to the land consolidation process by visualisation of the farm’s lands against the demarcated boundaries of designed blocks; – ensuring that designed blocks are filled by value assignments as proposed by participants in the land consolidation process; – ensuring that the data being entered are verified for conformity to requirements of the consolidation instruction and the land consolidation and exchange act; delivery of any – immediate information on the registration unit and designed blocks that may be useful on this stage of the land consolidation process, including data from the assessment register before consolidation; – preparation of final summary of equivalents to be designed for every farm and for every block. The following items are used as input data for designed solutions: – data from assessment register before consolidation; – geometrical data to define the boundaries of separated designed blocks; – layer of former condition plots; – layer of classification contours; – layer of designed blocks. From the user’s point of view, the main element of the system to support the initial selection for designed blocks is the specially designed form. The most important components of this form are: – information panel containing, among other things, data on the unit, former condition plots included in it, data on area and value of the unit, and a lot of other information required by the consolidation instruction; – table containing information on collected requests for the unit; – figures and graphical information showing the current filling status of individual designed blocks; – graphical window showing the arrangement of former condition plots, lands of active farm, boundaries of designed blocks and classification contours. The following method for collecting requests for a single farm was proposed: – the first step is to select the active registration unit, which results in filling relevant information fields with data to describe the unit; – the next step is to select the value for which a request in its basic and alternative version is submitted; – for every version the appropriate designed block is selected; Development of technologies used in agricultural engineering... the last step is confirmation of any decisions made. It is possible to cancel each of the decisions for any farm at any time, which makes the specific land value return to the appropriate designed block. The final effect of this stage of a land consolidation process is generating a number of summaries that are required further in the design work carried out in the digital map environment and the purpose of which is to design the target land arrangement. – DESIGN FOR A PRESET VALUE Designing starts with block selection. It requires the number of block to be indicated by the mouse. Next the designing direction is defined and score and number of designed plot are entered. Data are entered using a special form. Coordinates of the newly designed plot are calculated then and boundaries and number are inserted into the block. The designing direction may be altered at any time. The designed boundaries may be deleted and the block that occurs after such a boundary is deleted may be subject to another division. CONCLUSIONS The approach to the land consolidation process presented herein allows achievement of significant economic advantages resulting from, among other things, reduction in duration of individual operations. Choosing Bentley MicroStation software as a system platform enabled using a rich set of CAD 127 tools and allowed significant acceleration of work performance. The functions allowing quick redesign of the plot arrangement made it possible to transfer a lot of design variants to the land consolidation process participants, which effected in lack of complaints, and thus completing the process on time. Simultaneously with reduction in duration of individual stages, conditions were achieved for improvement in quality of a land consolidation process as regards its adjustment to the participants’ requests and through automatic detailed verification of the project for conformity to requirements of the consolidation instruction and the land consolidation and exchange act. The implementation of the discussed IT technologies for land consolidation process in Wojków village confirmed the practical usefulness and correctness of designed solutions. It also needs to be mentioned that individual elements of the proposed method for carrying out this stage are subject to constant modifications as the described technology is implemented in successive land consolidation processes in Poland. REFERENCE HARASIMOWICZ S. 1995: Wpływ cech działki i gospodarstwa na wartość dochodową gruntów. (Influence of plot and farm features on revenue value of lands). Zesz. Nauk. AR w Krakowie, ser. Geodezja No 16, 77–86 (in Polish). HOPFER A., URBAN M. 1984: Geodezyjne urządzanie terenów rolnych. PWN. Warszawa (Geodetic development of agricultural areas) PWN, Warsawa (in Polish) JANUS J., ZYGMUNT M. 2005: Technologia kompleksowej automatyzacji prac 128 U. Litwin, J. Janus, M. Zygmunt scaleniowych (Technology of comprehensive automation of land consolidation process). Materials from the 17th Scientific and Technical Session in the series “Current issues in geodesy and cartography”, Nowy Sącz (in Polish). LITWIN U., ZYGMUNT M. 2005: Nowa technologia generowania rejestru gruntów przed scaleniem (New technology for generation of land register before land consolidation). Zesz. Nauk. AR w Krakowie, ser. Geodezja No 21 (in Polish). URBAN M. 1981: Ekonomika i organizacja gospodarstw rolnych (Economics and organisation of agricutural farms), PWN, Warszawa (in Polish). W pracy przedstawiono trzy charakterystyczne etapy procesu scalenia: tworzenie rejestru szacunkowego przed scaleniem, etap wstępnego naboru do kompleksów projektowych oraz projektowanie. Omówiono propozycję sposobu automatyzacji wymienionych czynności z wykorzystaniem mapy numerycznej opracowanej przy pomocy programu Bentley MicroStation, z zachowaniem wymogów narzucanych przez instrukcję scaleniową. Wymienione etapy zilustrowano przykładami pochodzącymi ze scalenia przeprowadzonego w latach 2001–2004 na gruntach wsi Wojków. Prace na tym obiekcie wykonano z zastosowanej przedstawionej technologii. Pozwoliło to skrócić czas scalenia o około 30%. Streszczenie: Rozwój technologii wykorzystywanych w pracach urządzeniowo-rolnych na przykładzie wybranych etapów scalenia gruntów. Prace scaleniowe należą do najbardziej czasochłonnych i skomplikowanych zabiegów urządzenioworolnych wykonywanych przez geodetów. Proces ich automatyzacji oraz informatyzacji nie jest tak zaawansowany jak w przypadku innych rodzajów prac geodezyjnych. Przyczyną takiego stanu jest ograniczenie ilości prac z tego asortymentu wykonywanych od połowy lat osiemdziesiątych, czyli w okresie intensywnego wdrażania technik informatycznych w geodezji. Obecnie można zaobserwować ponowny wzrost zainteresowania wykonywaniem scaleń gruntów. Celem autorów było przybliżenie najnowszych rozwiązań informatycznych stosowanych w procesie scalenia gruntów, na przykładzie obiektu scaleniowego Wojków. MS. received November 2006 Authors’ addresses: Urszula Litwin, Jarosław Janus, Katedra Geodezyjnego Urządzania Terenów Wiejskich Akademia Rolnicza w Krakowie 30-059 Kraków Al. Mickiewicza 24/28 [email protected] Mariusz Zygmunt Katedra Geodezji Akademia Rolnicza 30-059 Kraków A. Mickiewicza 24/28 [email protected]