Effects of Local Land-Surface Processes on Heavy Rainfall in the Baltimore-Washington Metropolitan Area

A case study of heavy rainfall over the Baltimore-Washington metropolitan area is presented to highlight the effects of local land-surface processes on the water cycle of a heterogeneous region with dense urbanization. The Weather Research and Forecasting (WRF) model is employed and different types of urban parameterizations are applied. The simulation results are compared with a variety of observational datasets, including high-resolution radar rainfall fields; water vapor flux derived from Doppler radar wind profiles, Global Positioning System (GPS) precipitable water, and radiosonde water vapor density profiles; and surface observations. In general, the model adequately captures the rainfall distribution and the direction and magnitude of water vapor flux. Even though the heavy rainfall is primarily induced by large-scale forcing, the rainfall distribution is strongly dependent on the way in which urban land-surface processes are represented. When compared with the Noah land surface model, which adopts a simple representation of urban land-surface processes, a more realistic single-layer urban canopy model shows better performance in reproducing storm cells in terms of location near the urban area. As compared with results of an additional experiment in which the Baltimore-Washington metropolitan area is assumed to be a natural area, it is found that the temperature excess over the urban area and the modified water vapor transport by urban-modified winds play a role in the initiation and development of storm cells even under strong synoptic forcing conditions. Another sensitivity experiment is conducted to examine the effects of the Chesapeake Bay, and the results show that the evaporation and transport of water vapor from the Chesapeake Bay have a considerable impact on the rainfall amount particularly over the Baltimore region.