Mathematical and computational techniques for sustainable agricultural development in Nepal

Abstract

Challenges facing sustainable agriculture in Nepal include limited resources, climate variability, and suboptimal land use. In this work, mathematics and computing are applied to increase agricultural productivity in a sustainable manner. A Linear Programming model was developed for land-water allocation that resulted in a 12.5% increase in crop yield. Machine learning models, especially Random Forest, achieved 92.8% accuracy in crop yield prediction and data-driven decision-making. The seasonal variations of soil moisture has been analyzed using the Soil Moisture Model. The study finally concludes that mathematics and computing serve as a resource optimization tool in building climate resiliency toward ensuring long-term food security in Nepal.