Validation of APSIM and CERES-Sorghum model for prediction of rabi sorghum yields in Solapur region of Maharashtra

Abstract

APSIM and CERES – sorghum models were calibrated by using the data collected from the field experiments conducted at Centre for Rabi sorghum, Solapur, Maharashtra and from the farmer’s fields of Solapur district. For APSIM, rabi sorghum data of M35-1 (popularly grown cultivar of the region) under different dates of sowing in different years (2001 – 2004) was used. For CERES data of rabi 2005-06 and rabi 2006-07 on phenology and yield were used. The APSIM model has predicted the grain yield with deviation of 1.3 % to 7.0 % and -0.3% to -1.1% during rabi 2001 and 2002 at CRS, Solapur. Because of late receipt of rains during the years 2003 and 2004, the sowings were delayed and the model over predicted the yields with a deviation of 10.4 % to 23.1 %. CERES model simulation results revealed that for normal sowing, the model predicted the yield with deviation of 6.2 % to 26.7 % and 7.3 % to 9.3 % for the years 2005 and 2006, respectively. CERES model under predicted the flowering of M 35-1 with deviation of -8.8% to 11.8%. Whereas, the maturity was predicted with deviation of -7.1% to 2.7%. Key Words: APSIM, CERES-Sorghum model, validation, calibration.