Spatial Distribution of Sensible and Latent Heat Flux in the URBANFLUXES Case Study Cities London (U.K) and Basel (Switzerland) Based on a Combined Method Using Micrometeorological Approaches, Earth Observation (EO) Data and GIS-Techniques

Urban surfaces are usually complex mixtures of different land covers and surface materials, the relative magnitude of the surface energy balance components typically will therefore vary widely across a city. Eddy covariance (EC) measurements provide the best estimates of turbulent heat fluxes, but are restricted to their source area. To extrapolate to larger areas land surface modelling with earth observation (EO) data is beneficial as citywide information is possible. Turbulent sensible and latent heat fluxes are calculated using an approach based on the Aerodynamic Resistance Method (ARM). Input data such as land cover fractions and surface temperatures are derived from Landsat 8 OLI and TIRS, urban morphology (mean building height, plan area index, aspect ratio) was calculated from lidar data (high resolution digital building models) and GIS-data layers. Meteorological data are determined from flux towers, at networks of simpler meteorological variables (e.g. air temperature, humidity, air pressure, wind speed and wind direction). The spatial distributions of turbulent heat fluxes were analyzed with respect to the roughness lengths for momentum and heat, the additional aerodynamic resistance when using radiometric surface temperatures, stomatal resistances for different vegetation types and the level of imperviousness. The calculation of surface roughness parameters (roughness length, zero plane displacement height) for real urban surfaces by the method proposed by Kanda et al. (2013) is used. The magnitude of sensible heat flux strongly depends on the surface temperature, because the spatial variation of air temperature during daytime is small when limited to a satellite pixel view. For latent heat flux, the spatial distribution of the saturation deficit of vapor (e.g. above urban water bodies) and the (minimum) stomatal resistance of vegetation types show the largest influence on evapotranspiration rates. The EC fluxes are used to evaluate the ARM results. The latter use the EC source area weights to apply the ARM method.