An Experimental Investigation on Accessible Machine Learning Models for Examining the Possibility of Environmental Changes On Land for Farming Relevance

This research explores the pressing global challenge of food security in the context of climate change, with a focus on its impact on agricultural land suitability. The primary goal is to forecast potential risks linked to declining land suitability and shifting irrigation trends, which have direct implications for global food availability. Aligned with the United Nations’ Sustainable Development Goals to combat hunger and malnutrition, this study concentrates on Central Eurasia—a region marked by distinct socio- economic vulnerabilities. Using interpretable machine learning models, we examine how various carbon emission scenarios influence land suitability for agriculture. Our predictive model achieves an accuracy of 86% and a mean average precision of 72% in a multi-class classification of land suitability. By highlighting the most at-risk zones in Eastern Europe and Northern Asia, the findings offer valuable guidance for policymakers. These insights support data-driven decisions related to the efficient distribution of essential resources such as water and fertilizers, helping to avert future food crises. Overall, the study confirms the potential of machine learning in anticipating and mitigating the adverse effects of climate change on food production systems.