Comparative evaluation of performances of the Weather Research and Forecasting Model and an analog ensemble system over the Northwest Himalaya (NWH)

Abstract

An analog ensemble system (AE system) is developed for real-time local-scale prediction of precipitation amount at ten stations in the Northwest Himalaya (NWH) utilizing surface-meteorological observations. Comparative evaluation of performance of the AE system with the other forecasting methods may allow to assess its strengths (weaknesses) and usefulness as a tool for real-time local-scale precipitation forecasting over the NWH. Performance of the AE system is evaluated and compared with the performances of the Weather Research and Forecasting Model (WRF model) and climatological forecast model (CL model). This is done by comparing dynamically downscaled precipitation at 10 km spatial resolution with the help of the WRF model, predicted quantitative precipitation amounts with the help of the AE system and CL model with the corresponding observations on precipitation amount at 10 stations in the NWH. Performances of all the three models are evaluated using measures-oriented and distributions-oriented forecast verification measures to get deeper insight into their performances. Results show that the AE system performs marginally poor as compared to the WRF model and it performs better than CL model for local-scale prediction of precipitation amount and/or weather over the NWH. These findings are interesting and encouraging, and suggest that the AE system can be a useful practical tool for real-time local prediction of precipitation amount and/or weather over the NWH. Local-scale real-time weather forecasts with the help of the AE system can be a large help for various applications directly linked to and influenced by weather over the NWH.