APPLICATION ARTICLE Assessing the predictability of different kinds of models in estimating impacts of climatic factors on food grain availability in India Manojit Chattopadhyay 1 • Subrata Kumar Mitra 2 Accepted: 19 April 2017 Ó Operational Research Society of India 2017 Abstract The study explored the relationship of the climatic predictor variables such as seasonal temperature and rainfall pattern and non-climatic variable such as area under cultivation with the predictand per capita food grain production. We applied a linear method ‘‘Generalized Linear Model’’ and two non-linear methods ‘‘Multivariate Adaptive Regression Spline’’ and ‘‘Generalized Additive Model’’ to Indian data and assessed the data on basis of their performance in predicting food grain production. It was found that an adaptive version of generalized additive model has yielded the lowest predictive error in terms of lower root mean squared error. Better predictability of food grain production based on climatic factors may necessarily help to anticipate the nation’s food grain availability. The forecasts would facilitate scientists, farmers, policy makers, business organizations and the government to formulate appropriate adaptable strategies to cope with the climatic variability influence on food availability. Keywords Food grain Á Seasonal temperature Á Seasonal rainfall Á Climatic influence Á Area under cultivation Á Prediction & Manojit Chattopadhyay mchattopadhyay@iimraipur.ac.in Subrata Kumar Mitra mitrask@gmail.com 1 Information Technology Area, Indian Institute of Management Raipur, GEC Campus, Sejbahar, Raipur 492015, India 2 Finance Area, Indian Institute of Management Raipur, GEC Campus, Sejbahar, Raipur 492015, India 123 OPSEARCH DOI 10.1007/s12597-017-0314-9