Improved yield estimation technique for rice and wheat in Uttar Pradesh, Madhya Pradesh and Maharashtra States in India

Real time and location specific accurate estimation of crop yields is essential for advance planning of post harvesting agricultural activities. Weather parameters affect final crop production during crop life span. For getting accurate yield forecast, remote sensing can provide the location specific information of actual vegetation. Previously, weather variables viz., rainfall, temperature and relative humidity were incorporated in Correlation Weighted Regression Models for estimating crop yield. Moreover, VCI (Vegetation Condition Index) retrieved from satellite remote sensing was also adopted to forecast the winter wheat yields. Usually, crop performance and yield are predicted by different methods like weather parameters based models, Remote sensing (VCI) based models and also through crop cutting experiments (by State Agriculture Departments). However, in this paper, effort has been made to estimate yield more precisely by using combination of weather parameters and VCI data to get real time location specific yield forecasting predictors. To generate Rice (Oryza saliva L.) and Wheat (Triticum aestivum L.) yield forecast for selected districts of Uttar Pradesh, Madhya Pradesh and Maharashtra, statistical technique has been adopted during Kharif and Rabi seasons in years 2015, 2016 and 2017 with the help of historical data i.e., actual production data from 2004 to 2014. Yield forecasts generated by proposed model (combination of weather and remote sensing) were validated with actual production data provided by State Agriculture Departments and positively coupled comparison results imply that it can be applied in practice for both Rabi and Kharif season crops to generate location specific yield forecast more accurately.