Abstract

                                Weather forecasting become important tool for the planning of everydayfarming activities, particularly the medium range weather forecast ishighly required to reduce both input and production loss. Tamil NaduAgricultural University is providing block level medium range weatherforecast for next six days with Weather Research Forecast (WRF) Modeland having accuracy of 50 -70 per cent, varies with season. An attemptwas made at Agro Climate Research Centre, Tamil Nadu AgriculturalUniversity, Coimbatore during 2019-21 to improve the accuracy ofmedium range rainfall forecast over different Agro Climate Zones of TamilNadu state by increasing the resolution from block level (9 km) to villagelevel (3km) and altering the microphysics options in WRF v 4.2.1. Theperformance study tested four microphysics schemes viz., Kessler,WSM3, WSM5 and WSM6 for all the four seasons of 2020. The resultsobtained for the Southern Zone (SZ) is discussed in this paper and studyconcluded that the performance of high resolution village level (3km)rainfall forecast at Southern Agro Climatic Zone of Tamil Nadu generatedfrom WRFv4.2.1 model was best with the WSM3 microphysics schemefollowed by Kessler scheme, WSM5 and WSM6. Between the majorrainfall seasons, Forecast Accuracy Index (FAI) was higher in NEM (0.56to 0.73) than SWM (0.38 to 0.70), whereas the Forecast Usability Percent(FUP) was almost similar during both the South West Monsoon (SWM,53.0 to 81.4) and North East Monsoon (NEM, 52.9 to 85.9). There wasoverestimation observed irrespective of season and microphysics options

Key words : Medium range weather forecast; WRF; Microphysics; Accuracy; Verification