direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Inhalt des Dokuments

W3 - Heat and vapor fluxes of urban vegetation patterns – a remote sensing based approach

Doctoral student: Stenka Vulova

Supervisors: Prof. Dr. Birgit Kleinschmit, Prof. Dr. Doerthe Tetzlaff, Prof. Dr. Chris Soulsby

Background

Water resources planning and management, which is essential for human and ecological sustainability, demands a deeper understanding of water flows. Water flows can be partitioned into blue-water flows, which flow through rivers and recharge groundwater, and green-water flows, constituting transpiration from plants and evaporation from soil and other surfaces (Falkenmark & Rockström 2006). While research has traditionally focused on blue water flows, understanding the partitioning of rainfall into blue- and green-water flows across spatial and temporal scales is increasingly important for water resources management (Falkenmark & Rockström 2006). Vegetation is therefore a key component of the water cycle, driving water flow across the soil-plant-atmosphere interface and responding to fluctuations in water availability (Damm et al. 2018).

In urban environments, vegetation is generally spatio-temporally heterogeneous, with variations in vegetation type, species, vegetation density, vegetation height, leaf area, microclimate, water accessibility, and soil and water characteristics (Nouri et al. 2013a). Evapotranspiration (ET) is a crucial parameter with a high dynamic component in the urban water and energy regime, yet a sufficient understanding of the role of vegetation in the urban water cycle is still lacking. Quantifying ET accurately is a necessity in order to understand and manage climate change, irrigation, water use, and land use (Nouri et al. 2013b). As the majority of studies investigating ET have thus far focused on rural and agricultural regions, the estimation of ET in urban environments with heterogeneous land cover remains is a critical area of research (Bartesaghi Koc et al. 2018). Unravelling the resulting complex feedbacks and interactions between the plant-water system and environmental change is essential for any modelling approaches and predictions, but still inadequately understood due to currently missing observations (Damm et al. 2018).

Combined observational and modelling approaches incorporating remote sensing offer the potential to advance the understanding of heat and vapor fluxes at the soil-vegetation-atmosphere interface (Damm et al. 2018). ET estimation using remote sensing technology is the most efficient and cost-effective method that can be utilized for large spatial areas (Nouri et al. 2013a). Remote sensing is particularly well-suited for heterogeneous vegetated areas, due to its capacity to quantify spatio-temporally variable vegetation characteristics (Nouri et al. 2013a). Remote sensing can improve existing ET estimation methods, provide broad spatio-temporal coverage and facilitate continuous updates (Nouri et al. 2013a; Nouri et al. 2013b).

Aims

The main aims of this project are to 1) characterize the seasonal and diurnal variation of heat and vapor fluxes of urban vegetation using a combination of remote sensing (UAV, satellite) and field-based data; 2) validate UAV-based evapotranspiration (ET) estimates with in-situ ET estimates; and 3) to upscale ET estimates of urban vegetation spatially (urban scale) and temporally (multi-year) using satellite data.

Methods

In order to assess the diurnal and seasonal variability of ET from urban vegetation types, an interdisciplinary sampling campaign integrating remote sensing and field-based methods was conducted at a study site in Berlin, Germany.

Fig. 1: Sampling locations of different urban covers in Steglitz, Berlin.
Lupe

The sampling campaign lasted one year (spring 2019 to spring 2020) in order to characterize the seasonality in hydroclimatic drivers and phenological effects on ET. Three main vegetation types (urban grassland, trees, and shrubs) were sampled. Drone flights with multispectral and thermal cameras were undertaken. Leaf Area Index and soil moisture were measured, at monthly and 15-minute resolution, respectively. Using LI-6800, the diurnal and seasonal variation of gas exchange and stomatal conductance was measured. ET estimates from drone data were derived using algorithms, such as the “Triangle Method” and 3T model and validated with field-based ET estimates. Eddy flux data will provide hydroclimatic data. Sap flow monitoring will quantify tree transpiration. In the first stage of this study, a time series of ET across seasonal and diurnal scales of a single year will be generated for the local study area. Then, the field-validated UAV data will be combined with satellite data to upscale ET spatially and temporally. Machine learning methods will incorporate thermal and multispectral MODIS and Landsat imagery, Penman Monteith reference evapotranspiration (ET0), and other climatological variables in order to model latent and sensible heat fluxes measured by two eddy flux towers in Berlin.

On the city-scale, a parallel study predicts areas that will be most affected by higher air temperature using crowdsourced air temperature, machine learning, and satellite data. A crowdsourced air temperature dataset (“Internet of things”) from the TU Berlin Climatology department with high spatial and temporal resolution provides the unique opportunity to understand the factors influencing urban air temperature. Landsat 8 data (Land Surface Temperature, albedo, Normalized Difference Vegetation Index (NDVI), and Normalized Difference Built-up Index (NDBI)) were used as predictors, in addition to other geodata (including building height, vegetation height, and impervious surface fraction). Crowdsourced air temperature data from the past is also integrated into the models, allowing for forecasting of high-resolution air temperature one to two days in advance. Two machine learning algorithms were compared (generalized boosted regression models (GBM) and Random Forest). Combining a massive crowdsourced dataset with more than a thousand stations in Berlin with remote sensing data and machine learning methods is an innovative approach which will allow the city to better prepare for future heat waves.

Preliminary results

Thermal UAV imagery from the field-scale study already indicates that there is a huge variation in evapotranspiration of different vegetation types (forest vs. meadow) and even between different grass species, which has a direct effect on the land surface temperature. This unique dataset can quantify the heat and vapor fluxes of urban vegetation at a high spatial resolution, with practical recommendations for urban planning. In a future of cities faced with climate change and water scarcity, greening the urban environment with the most heat-reducing plants will improve the quality-of-life of urban residents.

Concerning the large-scale study results so far indicate that it is possible to model the air temperature of Berlin with a high spatial resolution and high accuracy. “High risk” areas for higher air temperatures in the metropolitan area of Berlin can be identified, which can be used to protect vulnerable residents and predict energy consumption for air conditioning. We also reveal cooling effects of urban green infrastructure.  This approach can enhance traditional DWD forecasting, with the benefit of more than 1000 stations in the Berlin area.

Collaborations

  • Fieldwork in Steglitz and conference paper with W1
  • Eddy flux and crowdsourced air temperature data exchange with Dr. Fred Meier and Daniel Fenner (Chair of Climatology, TU Berlin)
  • Methods and data exchange with W1, W2, and F4 (including collaboration within the “Interfaces in urban watersheds” Common Topic group)
  • Machine learning-related issues with Dr. Elena Matta and Dr. Mohammad Zounemat-Kermani
  • Lysimeter data exchange with Dr. Basem Aljoumani
  • ET methods exchange with Dr. Hamideh Nouri (Georg-August-Universität Göttingen)

References

Bartesaghi Koc,C., Osmond,P., & Peters,A. (2018): Evaluating the cooling effects of green infrastructure: A systematic review of methods, indicators and data sources. Solar Energy, 166, 486–508.

Bowler,D.E., Buyung-Ali,L., Knight,T.M., & Pullin,A.S. (2010): Urban greening to cool towns and cities: A systematic review of the empirical evidence. Landscape and Urban Planning, 97(3), 147–155.

Damm,A., Paul-Limoges,E., Haghighi,E., Simmer,C., Morsdorf,F., Schneider,F.D., van der Tol,C., Migliavacca,M., & Rascher,U. (2018): Remote sensing of plant-water relations: An overview and future perspectives. Journal of plant physiology, 227, 3–19.

Falkenmark,M., & Rockström,J. (2006): The New Blue and Green Water Paradigm: Breaking New Ground for Water Resources Planning and Management. Journal of Water Resources Planning and Management, 132(3), 129–132.

Nouri,H., Beecham,S., Kazemi,F., & Hassanli,A.M. (2013a): A review of ET measurement techniques for estimating the water requirements of urban landscape vegetation. Urban Water Journal, 10(4), 247–259.

Nouri,H., Beecham,S., Kazemi,F., Hassanli,A.M., & Anderson,S. (2013b): Remote sensing techniques for predicting evapotranspiration from mixed vegetated surfaces. Hydrology and Earth System Sciences Discussions, 10(3), 3897–3925.

Wang,X., Cheng,H., Xi,J., Yang,G., & Zhao,Y. (2018): Relationship between Park Composition, Vegetation Characteristics and Cool Island Effect. Sustainability, 10(3), 587.

Zusatzinformationen / Extras

Quick Access:

Schnellnavigation zur Seite über Nummerneingabe

Auxiliary Functions