SINGULARITIES OF TURBULENT URBAN HEAT FLUXES IN ŁÓDŹ

K. Kłysik, J. Wibig, K. Fortuniak (red.) - Klimat i bioklimat miast
Wydawnictwo Uniwersytetu Łódzkiego, Katedra Meteorologii i Klimatologii UŁ
Łódź 2008, s. 233–242
SINGULARITIES OF TURBULENT URBAN HEAT FLUXES
IN ŁÓDŹ
CECHY CHARAKTERYSTYCZNE
TURBULENCYJNYCH STRUMIENI CIEPŁA W MIEŚCIE
NA PRZYKŁADZIE ŁODZI
Krzysztof Fortuniak1, Sue Grimmond2, Brian Offerle3,
Włodzimierz Pawlak1, Mariusz Siedlecki1
1
Department of Meteorology and Climatology, University of Łódź
Narutowicza 88, 90-139 Łódź, Poland
e-mail: [email protected]
2
Department of Geography, King’s College London
Strand, London WC2R 2LS, England, UK
3
Earth Sciences Centre, Göteborg University
P.O. Box 100, SE-405 30 Göteborg, Sweden
Energy balance components measured at two points near the centre of the city of Łódź,
central Poland, are analysed. Both points were located in comparable environment dominated by
3–5 stories in height buildings. Eddy-covariance systems were mounted at high masts about 25 m
above roof level. Turbulent fluxes were calculated for two non-overlapping periods: 2000.11.05
– 2003.09.02 at Lipowa point and 2005.06.14 – 2008.01.31 at Narutowicza point. The monthly
ensemble diurnal cycles of the turbulent fluxes at two points are similar for months with similar
mean weather conditions. The latent heat flux (QE) accounts approximately 50% of sensible heat
flux (QH) in noon hours. During the night QH can be both positive and negative within the range
–60 ÷ +60 W·m–2 whereas QE rarely drops below zero. Turbulent fluxes vary in annual course
with winter minimum and summer maximum. Transient weather situations can significantly
modify typical pattern of energy partitioning. In extreme case related with warm advection
a negative QH can be observed all over 24 hours. In opposite case QH can be even larger than net
radiation in noon hours.
INTRODUCTION
Knowledge about turbulent fluxes is fundamental to advances in understanding energy exchange of urban surfaces. Due to high costs of sensors and
methodological problems experimental activities in the field on flux measurements in cities were relatively sparse up to the year 2000. Most of the early
studies investigated the surface energy balance over suburban residential areas
in relatively short periods (O k e 1978; C l e u g h , O k e 1986; R o t h ,
O k e 1993; G r i m m o n d , O k e 1995, 1999; O k e et al. 1999). After the
[233]
234
K. Fortuniak, C.S.B. Grimmond, B. Offerle, W. Pawlak, M. Siedlecki
year 2000 number of measurements of turbulent fluxes in the cities grew significantly and many of them have been ongoing over more extended periods
(F o r t u n i a k et al. 2001; G r i m m o n d et al. 2002; R o t a c h 2002;
N e m i t z et al. 2002; V o g t et al. 2006; A l - J i b o o r i et al. 2002;
S o e g a a r d , M o l l e r - J e n s e n 2003; M o r i w a k i , K a n d a 2004;
O f f e r l e et al. 2006a, 2006b). Contrary to the early experiments which
focused on the energy partitioning under favourable weather conditions
(“golden days”), long term measurements provide more comprehensive information about ensemble diurnal patterns and about seasonal changes in energy
balance components. Large campaign-style urban climate experiments allowed
also studying intra-urban differences of surface energy fluxes and their vertical
profiles. In spite of these studies knowledge on turbulent fluxes variability,
their extremes and respond to the weather conditions is still very little. The
objectives of this paper are to present both typical patterns of turbulent fluxes
evolution and their untypical course under specific weather situations. Data
from two urban locations in Łódź, Central Poland, are used to meet this goal.
STUDY AREA AND SIDES CHARACTERISTIC
Łódź is the second largest city in Poland with population about 800 000.
It is situated in central Poland (19°27’E, 51°56’N) in relatively flat area slightly
inclined south-easterly (altitudes from 180 m to 235 m). Lack of significant
geographical peculiarities like lakes, rivers, valleys, mountains, or sea allows
analysis a clear urban effect. Rapid urban development in the second half of
19th century is a reason for the regular arrangement in the old centre where
adjoined buildings, 3–5 stories in height, create real urban canyons. Apart from
a few high-rise buildings roof-level and zero-plane displacement can be easily
defined in the old centre. Because of all these features Łódź is a good place for
urban climate studies.
Measurements of energy balance components with eddy-covariance
method started in Łódź in November 2000. First, an urban energy balance system has been developed at Indiana University and installed at Lipowa 81
(F o r t u n i a k et al. 2001; O f f e r l e 2006a) by Prof. Sue Grimmond and
Dr Brian Offerle (Fig. 1). This site worked continuously up to the end of August 2003. Second, a very similar system, has worked at Narutowicza 88 since
May 2005 (F o r t u n i a k et al. 2006). Both measurement points are located
in the core of old centre in separated by 2.8 km. Surface cover characteristics of
the surrounding neighbourhoods are very similar, with comparable proportions
of roofs, roads, pavements, lawns and trees.
Singularities of turbulent urban heat fluxes in Łódź
235
Fig. 1. Map of Łódź with location of measurement sites:
1 – Lipowa 81, 2 – Narutowicza 88
Rys. 1. Mapa Łodzi z zaznaczoną lokalizacją punktów pomiarowych:
1 – ul. Lipowa 81, 2 – ul. Narutowicza 88
Eddy-covariance and radiation sensors are mounted near the top of high masts
(20 m – Lipowa, 25 m – Narutowicza) placed on the flat roofs of buildings
(17 m – Lipowa, 16 m – Narutowicza). Average roof level can be estimated as
~11 m at first site and as ~16 m at second, which according to rule-of-thumb
(G r i m m o n d , O k e 1999), gives the zero-plane displacements as zd =
7.7 m and zd = 11.2 m respectively. Roughness lengths for momentum calculated in close to neutral stratification (–0.05 < ζ < 0.01) are z0m = 2.0 m and
z0m= 1.9 m.
METHODOLOGY AND DATA USED
At each site eddy-covariance system consists of three-dimensional sonic
anemometer-thermometer (SWS–211/3K ATI – Lipowa, R.M. Young 81000 –
Narutowicza) and krypton hygrometer (KH20 Campbell Sci.) connected to
data-logger. Data sampled at 10 Hz were used to calculate fluxes with simple
236
K. Fortuniak, C.S.B. Grimmond, B. Offerle, W. Pawlak, M. Siedlecki
box-averaging for 1 hour periods. For each averaging period vertical wind
velocity was nullified with double rotation (K a i m a l , F i n n i g a n 1994).
Fluctuations of sonic temperature were converted into actual temperature. Humidity fluxes were corrected for oxygen cross-sensitivity of krypton hygrometer. Fluxes were also corrected for density fluctuation (WPL–correction;
W e b b et al. 1980) and for spatial separation of sensors. Additionally, all
components of radiation balance were measured independently by Kipp &
Zonen CNR1 net radiometer.
In the present study data from Lipowa for the period 2000.11.05 –
2003.09.02 and from Narutowicza for the period 2005.06.14 – 2008.01.31 are
used to study singularities of turbulent urban heat fluxes.
Fig. 2. Ensemble diurnal energy balance fluxes at two measurements sites in Łódź for July
calculated for hours in which valid observations exist (Lipowa point: years 2000–2003,
Narutowicza point years 2005–2007)
Rys. 2. Średni dobowy przebieg składników bilansu cieplnego na dwóch punktach pomiarowych
w Łodzi obliczony na podstawie wszystkich poprawnych danych godzinnych
z okresu pomiarowego (Lipowa: lata 2000–2003, Narutowicza lata 2005–2007)
RESULTS
Conventionally, a typical diurnal pattern of any meteorological parameter
can be obtained by averaging available data over the selected period of observation or by analysing selected days. Examples of ensemble diurnal pattern in
fluxes for July are presented at Fig. 2. It shows high consistency of fluxes at
both measurement sites. This consistency is a consequence of the locations in
comparable environments and similar weather conditions for the Julys during
the periods of observation. Depending on prevailing weather conditions for
other months the average diurnal fluxes evolution vary between the sites or are
Singularities of turbulent urban heat fluxes in Łódź
237
similar. Nevertheless, high consistency of fluxes on both stations, located in
these similar environments, gives a hope that sensors are located above roughness sub-layer, and fluxes are representative of the constant flux layer. As is
typical for urban areas, in noon hours the turbulent sensible heat flux (QH) is
larger than the latent heat flux (QE) with Bowen ratio close to 2 at both stations.
In the majority of summer days QH is in the range of 150–250 W·m–2 at both
sites with generally higher values at Lipowa than at Narutowicza (Fig. 3).
However, as net radiation (Q*) is a main forcing controlling the sensible heat
flux, QH can rise in hot summer days above 300 W·m–2. The largest QH recorded in a 1 h averaging period reached 445 W·m–2 on May 25th, 2001 (Lipowa). Whereas, during cloudy weather there are very small turbulent fluxes
(even below 50 W·m–2). A second consequence of the steering role of radiation
in sensible heat flux evolution is the annual course of QH gradually increasing
from winter minimum to summer maximum.
Fig. 3. Monthly box-and-whiskers plots for sensible (QH) and latent (QE) turbulent heat fluxes
calculated for midday hours (10–14 h) for days possessing valid observations for QH and QE. The
tops and bottoms of each “box” are the 25th and 75th percentiles of the samples, respectively.
The line in the middle of each box is the sample median. Dots mark outlier (values that are more
than 1.5 times the interquartile range away from the top or bottom of the box).
Rys. 3. Wykres pudełkowy miesięcznych wartości turbulencyjnego strumienia ciepła jawnego
(QH) i ciepła utajonego (QE) dla godzin okołopołudniowych (10-14) dni z poprawnymi wynikami
pomiarów QH i QE. Góra i dół „pudełka” oznaczają 25% i 75% percentyl z próbki (dolny i górny
kwartyl). Linia na środku pudełka oznacza medianę. Kropki oznaczają wartości skrajne
(odstające ponad 1,5 wielkości międzykwartylowej od góry i dołu pudełka).
238
K. Fortuniak, C.S.B. Grimmond, B. Offerle, W. Pawlak, M. Siedlecki
More factors determine the latent heat flux. Besides radiation and precipitation,
vegetation plays an important role in latent heat flux intensity as it pumps water
from deep soil to the atmosphere. Even if vegetation is reduced in urban areas,
it rapidly develops in May, resulting in a jump of QE from low winter values to
(typically in range 0–50 W·m–2 from October to April) to summer ones (Fig. 3).
Typical summer values of QE are in a range 70–120 W·m–2. Latent heat flux
extremes are difficult to estimate due to limitations of measurement technique
because during wet weather or immediately following, when one can expect the
highest water vapour flux from the ground, raindrops deposit on the sensor
windows which leads to spurious results. Only a few episodes of high QE are
observed. These are situations when the krypton hygrometer dried quickly after
rain which could be recorded. Still, it is unlikely that they represent real
maxima.
The example diurnal patterns (Fig. 2) illustrate another feature of urban
climate, that of the sensible heat flux is close to zero through the night indicating a near-neutral surface layer at the measurement height. However, distributions of all nocturnal fluxes (Fig. 4) suggest that this is a mean value rather that
an every night occurrence. Standard deviations (σ), 10%, and 90% quantiles
(q10%, q90%) additionally show high diversity of night-time QH. Their values at
Lipowa are: σ = 17 W·m–2, q10% = –19 W·m–2, q90% = 20 W·m–2 and at Narutowicza: σ = 15 W·m–2, q10% = –29 W·m–2, q90% = 9 W·m–2. Additionally,
negative QH characterise more than 50% of night at both stations – 64% at Lipowa and 75% at Narutowicza. Latent heat flux is positive on the majority of
nights (88% at Lipowa and 92% at Narutowicza) and it never drops below
–15 W·m–2. Larger values of QE are mainly associated with post-rain situations
which create the right tail of the distributions.
Fig. 4. Probability density functions of night-time (for all night hours) values of sensible (QH)
and latent (QE) turbulent fluxes at two measurements sites in Łódź
Rys. 4. Funkcje gęstości prawdopodobieństwa średnich nocnych wartości (dla każdej nocy
średnia po wszystkich godzinach nocnych) turbulencyjnego strumienia ciepła jawnego (QH)
i ciepła utajonego (QE) na dwóch punktach pomiarowych w Łodzi
Singularities of turbulent urban heat fluxes in Łódź
239
Figure 5 presents a typical evolution of Q*, QH and QE in five days of fine
weather in July 2006. As in mean diurnal variation case (Fig. 2), the sensible
heat flux dominates over a latent heat. Because of increased heat capacity of
urban structures QH remains positive late in evening, when Q* drops below
zero. However, large radiative cooling leads to a negative QH in the analysed
nights. Latent heat flux is positive in the analysed case varying from about
100°W·m–2 at midday to a few W·m–2 at night. This classical pattern can dramatically change under specific weather conditions, especially during winter
when radiative forcing is relatively weak. Warm advection after a period of
cold weather decreases the sensible heat flux. An extreme case (Fig. 6), was the
two days 15–16 November 2002 when QH was negative not only at night but
also in the middle of the day. During this period the air temperature rose to
10–12°C whereas just a few days earlier it had remained close to 0°C for more
than a week. However under cold advection (Fig. 7), as occurred during the
period 16–17 January 2001 the air temperature dropped from about 0°C to
about –6°C. The sensible heat flux during this period was positive throughout
24 hours per day and even greater than the net all wave radiation during the
daytime.
Fig. 5. Energy balance components evolution under fine weather conditions
– days with one of the greatest QH at Narutowicza
Rys. 5. Przebieg składników bilansu cieplnego w kilkudniowym okresie ładnej pogody
– dni z jednymi z największych wartościami QH na posterunku przy ul. Narutowicza
240
K. Fortuniak, C.S.B. Grimmond, B. Offerle, W. Pawlak, M. Siedlecki
Fig. 6. Example of two days with negative sensible heat flux for all 24 hours
in a day: 15–16 November 2002 at Lipowa
Rys. 6. Przykład dwudniowego okresu z ujemnymi wartościami strumienia
ciepła jawnego w ciągu całej doby: 15–16 listopada 2002 na stacji Lipowa
Fig. 7. Example of two days with sensible heat flux greater than net radiation
at noon hours: 17–18 January 2001 at Lipowa
Rys. 7. Przykład dwudniowego okresu, kiedy strumień ciepła jawnego był w godzinach
południowych większy od salda promieniowania: 17–18 stycznia 2001 na stacji Lipowa
CONCLUSIONS
Long term measurements of turbulent fluxes in Łódź allow analysis of the
urban energy partitioning beyond the “golden days” to transient weather situations. The individual diurnal pattern of turbulent fluxes can differ significantly
from the mean evolution. Day-time turbulent fluxes vary in a wide range for
each month depending not only on the present but also on the past weather
conditions. The high heat capacity of a town can amplify sensible heat fluxes
due to heat accumulation or reduce this flux due to large surface cooling. This
must be considered in numerical model evaluations.
Singularities of turbulent urban heat fluxes in Łódź
241
REFERENCES
C l e u g h H. A., O k e T. R., 1986, Suburban-rural energy balance comparisons in summer
for Vancouver, B.C. Boundary-Layer Meteorology, 36, p. 351–369.
F o r t u n i a k K., O f f e r l e B., G r i m m o n d S., O k e T., K ł y s i k K.,
W i b i g J., 2001, A system to observe the urban energy balance: Initial results from winter-time measurements in Łódź. Annales Universitatis Mariae Curie-Skłodowska, Sectio B
Geographia, Geologia, Mineralogia et Petrographia, Vol. LV/LVI, 19, p. 167–176.
F o r t u n i a k K., K ł y s i k K., S i e d l e c k i M., 2006, New measurements of the
energy balance components in Łódź. [in:] 6th International Conference on Urban Climate,
June 12–16 2006, Göteborg, Sweden. Preprints. Ed. Urban Climate Group, Department of
Geosciences, Göteborg University, p. 64–67.
G r i m m o n d C. S. B., O f f e r l e B. D., H o m J., G o l u b D., 2002, Observations of
local-scale heat, water, momentum and CO2 fluxes at Cub Hill, Baltimore. [in:] Preprints
4th Symposium on the Urban Environment, Amer. Meteor. Soc., p. 117–118
G r i m m o n d C. S. B., O k e T. R., 1995, Comparison of heat fluxes form summertime
observations in the suburbs of four North American cites. Journal of Applied Meteorology,
34, p. 873–889.
G r i m m o n d C. S. B., O k e T. R., 1999, Heat storage in urban areas: Local-scale observations and evaluation of a simple model. Journal of Applied Meteorology, 38, p. 922–940.
K a i m a l J. C., F i n n i g a n J. J., 1994, Atmospheric boundary layer flows: Their structure
and measurement. Oxford University Press, New York, NY, 289 p.
M o r i w a k i R., K a n d a M., 2004, Seasonal and diurnal fluxes of radiation, heat, water
vapor and CO2 over a suburban area. Journal of Applied Meteorology, 43, p. 1700–1710.
N e m i t z E., H a r g r e a v e s K. J., M c D o n a l d A. G., D o r s e y J. R.,
F o w l e r D., 2002, Micrometeorological measurements of the urban heat budget and
CO2 emissions on a city scale. Environmental Science & Technology, 36, p. 3139–3146
O f f e r l e B., G r i m m o n d C. S. B., F o r t u n i a k K., K ł y s i k K., O k e T. R.,
2006a, Temporal variations in heat fluxes over a central European city centre. Theoretical
and Applied Climatology, 84, p. 103–115.
O f f e r l e B., G r i m m o n d C. S. B., F o r t u n i a k K., P a w l a k W., 2006b, Intraurban differences of surface energy fluxes in a central European city. Journal of Applied
Meteorology and Climatology, 45, p. 125–136.
O k e T.R., 1978, Surface heat flux and the urban boundary layer. [in:] WMO Symp. Boundary
Layer Physics Applied to Specific Problems of Air Pollution, Norköping, Sweden, WMO,
p. 63–69.
O k e T. R., S p r o n k e n - S m i t h R. A., J á u r e g u i E., G r i m m o n d C. S. B.,
1999, The energy balance of central Mexico City during the dry season. Atmospheric Environment, 33, p. 3919–3930.
R o t a c h M. W., 2002, Overview on the Basel urban boundary layer experiment – BUBBLE.
[in:] Preprints 4th Symposium on the Urban Environment, Amer. Meteor. Soc., p. 25–26
R o t h M., O k e T. R., 1993, Turbulent transfer relationships over an urban surface. I: Spectral characteristics. Quarterly Journal of the Royal Meteorological Society, 119, p. 1071–
1104.
S o e g a a r d H., M o l l e r - J e n s e n L., 2003, Towards a spatial CO2 budget of a metropolitan region based on textural image classification and flux measurements. Remote
Sensing of Environment, 87, p. 283–294.
V o g t R., C h r i s t e n A., R o t a c h M. W., R o t h M., S a t y a n a r a y a n a A. N.
V., 2006, Temporal dynamics of CO2 fluxes and profiles over a Central European city.
Theoretical and Applied Climatology, 84, p. 117–126.
242
K. Fortuniak, C.S.B. Grimmond, B. Offerle, W. Pawlak, M. Siedlecki
W e b b E. K., P e a r m a n G. I., L e u n i n g R., 1980, Correction of the flux measurements for density effects due to heat and water vapour transfer. Quarterly Journal of the
Royal Meteorological Society, 106, p. 85–100.
STRESZCZENIE
W pracy przeanalizowano wyniki pomiarów składników bilansu cieplnego na dwóch punktach pomiarowych położonych w centralnych dzielnicach Łodzi – punkt przy ul. Lipowej 81
i punkt przy ul. Narutowicza 88. Oba punkty zlokalizowane były w miejscach o podobnym typie
zabudowy zdominowanym przez 3–5 piętrowe budynki. Pomiary strumieni turbulencyjnych
wykonano z zastosowaniem metody kowariancji wirów. Czujniki zamontowane były na wysokich masztach, ok. 25 m nad średnim poziomem dachów. Turbulencyjne strumienie ciepła jawnego (QH) i ciepła utajonego (QE) oraz składniki bilansu radiacyjnego mierzono na posterunku
przy ul. Lipowej w okresie 2000.11.05 – 2003.09.02, a na posterunku przy ul. Narutowicza
w okresie 2005.06.14 – 2008.01.31. Średnie dobowe przebiegi składników bilansu cieplnego
w poszczególnych miesiącach cechuje duże podobieństwo dla obu punktów pomiarowych
(rys. 2). W godzinach południowych strumień ciepła utajonego stanowi ok. 50% strumienia
ciepła jawnego. W ciągu nocy (rys. 4) QH może być zarówno dodanie jak i ujemne zmieniając się
w granicach –60 ÷ +60 W·m–2, z wartością średnią zbliżoną do 0 W·m–2. Natomiast nocny strumień ciepła utajonego jedynie w wyjątkowych wypadkach spada poniżej zera. Turbulencyjne
strumienie ciepła charakteryzuje wyraźna zmienność sezonowa z maksimum w lecie i minimum
zimą (rys. 3). Typowa dobowa zmienność składowych bilansu cieplnego pojawiająca się
w radiacyjnym typie pogody (rys. 5) może ulec znacznemu zaburzeniu podczas pogody adwekcyjnej. W wyjątkowych przypadkach adwekcji ciepła strumień ciepła jawnego może być mniejszy od zera przez całą dobę (rys. 6). W okresie zimowym, podczas napływu mroźnego powietrza
nad miasto, QH może być nie tylko dodatnie przez całą dobę, lecz nawet w południe przyjmować
wartości większe od salda promieniowania (rys. 7).