Sustainable Agriculture Through Data-Driven Land And Soil Management For Rural Farming Communities

This thesis aims to investigate sustainable farming practices by utilizing a database containing soil, weather, and water data. Additionally, it evaluates cropping strategies using a system that incorporates nutrient dynamics and environmental values. The objective is to monitor, develop predictive models, and assess the impact of both controlled and uncontrolled nutrient variables. This is achieved through the development of an algorithm that provides information on soil health based on datasets and land use, with a particular focus on residual values after harvesting for deeper analytical insights and interpretation. This study focuses on key nutrients, including pH levels, water availability, and temperature forecasts. Utilizing the available dataset, machine learning algorithms are applied to evaluate and compare two main scenarios: pure cropping and intercropping. Additionally, the proposed model assesses the selection process for these features using three methodologies, which guide the identification of either a primary crop or a companion crop, depending on the scenario. This approach optimizes rural farming operations by providing farmers with predictive insights into soil health and nutrient availability, thereby promoting the long-term sustainability impact for both farmers and their communities.