The usefulness of medium range weather forecast in improving the quality of output from CROPGRO-Soybean model

Abstract

 CROPGRO-Soybean model calibrated for local conditions of Raipur has been used to evaluate the relevance of medium range weather forecast relative to the soybean crop growth period. A procedure that makes use of historical weather data, medium range weather forecast (mrwf) and current weather data in conjunction with the CROPGRO-Soybean model was developed to arrive at a probable distribution of predicted yield. A series of perfect mrwf for 5 days were assumed for assessing the sensitivity of the crop management system to forecast information. The relative importance by time of year was taken as a reduction in variance due to a perfect 5-day mrwf.

The results of the study, conducted for two reference years 1986 (low-production year) and 1993 (high-production year) at Raipur, showed that the yield estimation can be done 20 days in advance before the physiological maturity for low-production year (1986) and 15 days in advance before physiological maturity for high-production year (1993). For both the years mrwf during reproductive phases are more valuable. It has also been concluded that the longer forecast periods are responding earlier in the growing season with higher values too.