Measuring the impact of climate change on potato production in Bangladesh using Bayesian Hierarchical Spatial-temporal modeling.

Abstract

Potato is a staple food and a main crop of Bangladesh. Climate plays an important role in different crop production all over the world. Potato production is influenced by climate change, which is occurring at a rapid pace according to time and space. The main objective of this research is to observe the variation in potato production based on the discrepancy of the variability in the spatial and temporal domains. The research is based on secondary data on potato production from different parts of Bangladesh and five major climate variables for the last 17 years ending with 2020. Bayesian Spatial-temporal modelling for linear, analysis of variance (ANOVA), and auto-Regressive models were used to find the best-fitted model compared with the independent Error Bayesian model. The Watanabe-Akaike information criterion (WAIC) and Deviance Information Criterion (DIC) were used as the model choice criteria and the Markov Chain Monte Carlo (MCMC) method was implemented to generate information about the prior and posterior realizations. Findings revealed that the ANOVA model under the Spatial-temporal framework was the best model for all model choice and validation criteria. Results depict that there is a significant impact of spatial and temporal variation on potato yield rate. Besides, the windspeed does not show any influence on potato production, however, temperature, humidity, rainfall, and sunshine are important components of potato yield rate in Bangladesh. It is evident that there is a potential impact of climate change on potato production in Bangladesh. Therefore, the authors believed that the findings will be helpful to the policymakers or farmers in developing potato varieties that are resilient to climate change to ensure the United Nations Sustainable Development Goal of zero hunger.