Bias Correction and AREVIA Model for Paddy Crop Production in Selected Districts of Karnataka

摘要

Forecasting is an essential tool in agriculture to study the area and production of various crops. Estimates of Paddy Production are taken from crop cutting experiments and through satellite. The data obtained from the crop cutting experiment estimates are always more accurate than that of the satellite estimates. Crop cutting experiments have its own limitations mainly in the coverage of area in a limited period. Remote sensing data collected covers larger area. In view of its large resolution, biasness could be observed in the data collected. To reduce this, different bias corrective methods such as Difference Method (DM) and Modified Difference Method (MDM) were used. Based on the Normalized Mean Square Error (NRMSE) value the best method for each data set is identified. The study resulted that MDM was the best method for bias correction showing the least value of NRMSE and it can be used for smoothening of the modeled data. In the presence of auto correlation, model fitting was done by ARIMA models for the crop production data of all selected districts of Karnataka. ARIMA (0,1,0) was found to be the best fit for production of Paddy crop for all the selected districts.