Abstract
Abstract. The development of kharif rice yield prediction models was attempted through Machine Learning approaches such as Artificial Neural Network and Random Forest for the 42 blocks covering 13,141 sq km upland rainfed area of Purulia and Bankura district, West Bengal. Models were developed integrating monthly NDVI with weather and non-weather variables at block-level for the period 2006 to 2015. The model correlation obtained was 0.702 with MSE 0.01. Though the weather variables vs NDVI models are quite satisfactory, NDVI vs kharif rice yield models however, show relatively less correlation, about 0.6 revealing the requirement of varied additional farmer-controlled inputs. Development of NDVI vs crop yield models for different crop growth stages or fortnightly over a larger data set with selective adding of weather and non-weather variables to NDVI would be the most appropriate.
Highlights
In India, rice is one of the most important crops and contributes significantly to national economy
This paper focus on developing kharif rice yield prediction models for rainfed upland area through machine learning techniques integrating block level Normalized Difference Vegetation Indices (NDVI) with weather and non-weather variables
Yield prediction models were developed for the blocks of Purulia and Bankura districts for the period 1992 to 2015 (24 yrs) using 5 monthly weather variables only The Purulia model correlation is 0.669
Summary
In India, rice is one of the most important crops and contributes significantly to national economy It is grown in about 4.34 lakh sq km area spread in almost all the States and constitutes about 41% (2015-16) of the total food grains output. The long-term climate change and the complexities of seasonal weather variability have induced an inevitable uncertainty in yield. It is more so for kharif rice yield in the rainfed area which is most vulnerable to southwest monsoon rainfall- the main source of water. In India, such rainfed area occupies about 75 million hectares, i.e., about half the net sown area of 139.9 million hectares In this regard crop-yield prediction before harvest is required by planners as well as farmers. The satellite data-based crop prediction is being done in India by Mahalanobis National Crop Forecasting Centre (MNCFC), New Delhi
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
