Development of an urbanized land data assimilation system RMAPS-LDAS

An urbanized land data assimilation system (LDAS) with a spatial resolution of 1 km is developed to provide a better estimation of land surface elements (e.g., soil moisture and temperature, road and roof temperature, etc.) for a regional weather forecast system named Rapid-refresh Multi-scale Analysis and Prediction System (RMAPS) in order to enhance the precision of short-term regional weather forecasts. The Noah land surface model and the multi-layer building energy model BEP-BEM are coupled to obtain a better representation of land surface states for both natural and urban underlying surfaces. The RMAPS-LDAS runs offline with the input of fused meteorological data from in-situ observations and gridded meteorological data. Some static data (e.g., soil texture classes and monthly green vegetation fraction) are updated using local and satellite-based sources to improve the land modeling accuracy. The structure of the coupled Noah-BEP-BEM modeling system is reorganized to improve the computational and storage efficiency, which makes RMAPS-LDAS an operationally applicable system for regional weather forecasts. Evaluation results show that the land surface states simulated by RMAPS-LDAS improve temporal and spatial patterns of initial conditions and can improve short-term weather forecasts.