Urban roughness parameters estimation from globally available datasets for mesoscale modeling in megacities

Abstract Approaches to understanding urban atmospheric phenomena in cities include mesoscale models coupled with urban canopy models (UCMs). To improve the accuracy of these models, urban morphological parameters of higher spatial resolution are necessary. In this paper, we introduce an approach to derive 1-km scale urban parameters from globally available satellite images of Landsat 8, Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), Global Multi-resolution Terrain Elevation Data 2010 (GMTED2010), and nighttime light (NL) images from the Defense Meteorological Satellite Program Operational Linescan System (DMSP-OLS) for use in numerical weather models. Empirical equations derived from linear fittings were used to calculate urban geometric parameter plane area index from satellite images, which were then compared with values derived from real plane area index. To calculate urban geometric parameters average building height, an artificial neural network fitting was conducted. From the basic urban geometric parameters, other necessary urban parameters were derived. After validating the method, a weather simulation was conducted using the derived parameters for Jakarta, Indonesia. Results showed that urban parameters made from global datasets could improve the performance of the single-layer urban canopy model; and these estimated parameters can be used as a substitute when actual building information are lacking.