Evaluation of Upper Tropospheric Humidity in WRF Model during Indian Summer Monsoon
In this work, we evaluate the upper tropospherichumidity (UTH) in a regional atmospheric model in conjunction with remote sensing observations and reanalysis products during the Indian summer monsoon (ISM). We performed continuous Weather Research and Forecast (WRF) model simulations from 1st May to 1st October for every year during 2001 to 2011 at 45 km spatial resolution. The maximum UTH zones viz. Bay of Bengal, and central and north-east Indian regions are well represented in WRF model when compared to the satellite observations and European Centre for Medium-Range Weather Forecasts (ECMWF) interim reanalysis (ERAI) during ISM season. Analyses found that ERAI exhibits higher magnitudes of UTH over the ISM region compared to that of satellite observations (Atmospheric Infrared Sounder:AIRS) and WRF. In terms of spatial distribution, WRF is in close agreement with satellite observations as compared to ERAI, is further supported by the pattern correlation coefficients. In addition to that, our analysis suggested that WRF model could simulate the seasonal evolution of the northward movement of maximum UTH band as in AIRS and ERAI. However, UTH variability over the equatorial Indian Ocean and western north Pacific (north of Madagascar region) is underestimated (overestimated) in the WRF model compared to the observations. Nevertheless, the model is able to represent high (low) UTH over the north Indian Ocean region during active (break) period, unable to capture the northward propagation of UTH well. This indicates that the model has considerable discrepancies in simulating UTH over the deep convective monsoon region during the ISM season. It is suggested that in order to improve the UTH representation in the model over the ISM region, it is essential to reduce biases over the equatorial and southern tropical regions. Thus, this study emphasized the variations in UTH at different time scales during monsoon season along with the credibility of remote sensing observations in WRF model.