Includes bibliographical references.

A field experiment was carried out in the kharif season of 2019 at the Agricultural Research Farm of College of Post Graduate Studies in Agricultural Sciences, Umiam, Central Agricultural University (Imphal). The experiment was conducted using Two Factorial Randomized Block Design with 3 replications for 3 rice cultivars (Ranjit, CAU R1 and Shahsarang1) under 4 nutrient levels (control, 100% RDF, 100% FYM and 50% RDF+50% FYM). Agronomic data were recorded throughout the growth period of the crop at an interval of 30 days after transplanting (DAT). Daily weather data related to maximum and minimum temperatures, solar radiation and total rainfall over the crop growth period was collected from the North Eastern Space Application Centre (NESAC), Umiam, Meghalaya. The study was undertaken to calibrate the DSSAT CERES-Rice model for popular rice varieties of Meghalaya and to simulate the performance of DSSAT CERES-Rice model for organic and inorganic inputs in lowland rice.
The results established that growth parameters viz., plant height, number of tillers per hill, leaf area, leaf area index (LAI), and Crop Growth Rate (CGR) improved with the succession of the crop growing period and reached maximum at 60 DAT and declined thereafter. Growth indices like Relative Growth Rate (RGR) and Net Assimilation Rate (NAR) decreased with the succession in crop growth period. Dry matter production was maximum at harvest. Among the cultivars Ranjit recorded significantly higher number of tillers per hill and Shahsarang1 recorded significantly superior plant height, leaf area, LAI, dry matter production and yield. CAU R1 recorded significantly higher test weight and Ranjit recorded significantly longer panicles. Among the nutrient levels, significantly superior performance of the above parameters was observed in case of 50%RDF+50%FYM followed by 100%RDF, 100%FYM and Control. The cultivar Shahsarang1 cultivated under 50%RDF+50%FYM performed better in terms of growth and yield as compared to other treatments.
The model predicted grain yield and tops weight which were in good agreement among the observed and simulated values but physiological maturity was underestimated for CAU R1.The simulated grain yield, tops weight and days taken to attain physiological maturity were in agreement with the observed values for Shahsarang1. Model calibration for Ranjit variety could not be done due to its low yield and disease infestation.The CERES rice model failed to predict yield under control (0% nitrogen) plot which may be attributed to the medium concentration of N present in the field. The CERES Rice model underestimated the grain yield, biological yield and physiological maturity under 100% FYM and 50% RDF+50%FYM for both CAUR1 and Shahsarang1 rice varieties with RMSE values of 1382.07 kg ha-1 (grain yield), 4495.73 kg ha-1 (biological yield) and 11.81(physiological maturity); and 1539.04 kg ha-1 (grain yield), 5193.17 kg ha-1 (biological yield) and 11.20 (physiological maturity) respectively. Simulated values for CAUR1 and Shahsarang1 with respect to the grain yield, biological yield and physiological maturity showed low index of agreement (dstat) with the observed values.