Performance evaluation of medium-range weather forecast using crop growth simulator

Abstract

Medium range forecast, a recently introduced service in India, has not been evaluated satisfactorily in terms of agricultural applications. In this article an attempt has been made to evaluate the performance of medium range weather forecast for Delhi and Ludhiana environment in the rabi (winter) season of 1994 - 95 by comparing the forecasted with actual weather components, mainly maximum and minimum temperature, wind speed, sunshine hours, and rainfall. Furthermore, yield and growth response of wheat simulated under adequate, moderate, and limited irrigated conditions were compared for forecasted and actual weathers employing wheat growth simulator (WTGROWS). Medium-range forecast performance for predicting maximum and minimum temperatures for both the locations had predictability lying in the range of 70 to 85 percent. The precision was somewhat better in case of maximum temperature assessment. Wind speed for both locations was poorly assessed. Forecasted rainfall matched actual rainfall on some occasions only at Ludhiana but the performance, in terms of seventy-two hours around the forecasted time, gave better results. The observed bright sunshine hours showed poor predictability (less than 30%). Forecasted grain yields for all levels of soil moisture availability treatments at both the locations were.