Weather Based Support System for Farmers

Agriculture remains the backbone of rural economy, yet crop productivity continues to be impacted by unpredictable environmental conditions. Farmers struggle to select the right crop per season due to insufficient insights regarding weather, pesticide usage and crop-specific sensitivity. This paper presents a comprehensive Machine-Learning-based Crop Recommendation and Yield Prediction System that forecasts yield for multiple crops in a village and performs simulation-based sensitivity analysis. The proposed system evaluates the effects of ±20% change in pesticide usage and ±10% change in weather (temperature and rainfall) on the final yield. Furthermore, high explanatory value is achieved by identifying top positive and negative growth factors for every crop. The system also presents harvest quarter information, pesticide cost estimation and interactive “what-if” simulation for farmers. The model was implemented using Python, Flask, and Chart.js visualization, and real world dataset containing crop-wise quarterly weather and pesticide usage details. Results indicate higher decision accuracy and financial advantage using data-driven cultivation planning. The solution contributes to precision agriculture and technology-assisted farming.