Modeling of a severe winter drought in eastern China using different initial and lateral boundary forcing datasets.

This paper describes the performance of the Regional Atmospheric Modeling System (RAMS) in simulating a winter drought event, based on two different forcing datasets. We ran EC (ERA-Interim reanalysis data as initial and lateral boundary forcing conditions) and FNL (NCEP-FNL reanalysis data) simulations for the 2008/2009 winter drought event to quantify the impact of any uncertainty in the different initial and lateral boundary forcing data on regional model outputs. The response of the winter mean atmospheric states to the variations in the initial and lateral boundary conditions was investigated on the basis of these simulation results. The spatio-temporal features of precipitation from the EC and FNL runs closely resembled those measured from the Global Summary Of the Day (GSOD) observations, although the EC run data outperformed the FNL run data in both their spatial distribution patterns and precipitation values. The water vapor flux values explain how the differences in the precipitation values between the EC and the FNL runs were generated, whereas temperature values were not sensitive to any changes in forcing data. The model results from these runs also slightly overestimated temperature on both spatial and temporal scales. For the tropospheric atmospheric data recorded at the Fuyang Meteorological Station in Anhui Province, neither the time series nor the statistical analyses showed any evidence of superiority between the two different driver datasets compared with radiosonde data. However, on closer inspection, the influence of different initial and lateral boundary conditions on modeling the tropospheric atmospheric data appeared to be evident. [ABSTRACT FROM AUTHOR]