Improving Urban Climate Simulation by Integrating Remotely Sensed Albedo into the WRF Model

With more than 50% of the global population living in cities and the continued urbanization trends, urban areas represent critical hotspots of water, energy, and health challenges facing humanity in the 21st century. A better understanding and prediction of urban microclimate and hydrology within the context of global environmental change plays a key role in tackling these challenges. Correctly characterizing the albedo of building materials is identified as the most important factor in improving urban simulation results. Most urban land surface models used in weather and climate models (e.g., the single-layer urban canopy model in the WRF model) still employ tabulated albedo values, which have limited spatial variability. We used remotely sensed albedo data to improve urban albedo characterization in weather models. We developed a new high-resolution urban albedo dataset based on Landsat and Sentinel-2 data. The new dataset separates the roof from the impervious ground in the NLCD impervious surface dataset using global building footprint data. We then estimated rood and ground albedos at various spatial scales ranging from 1-10 km. The new dataset will improve the characterization of the albedo parameters in the WRF model, improve the simulation of urban meteorological variables at the weather scale, and thus empower stakeholders and researchers to better navigate urban planning and policies in a changing climate.