Monte Carlo Simulation for Rice Yield Risk Estimation Based on Weather and Soil Quality Factors

Abstract

This study applies Monte Carlo simulation to estimate rice yield risks in the Medan region during 2024 by incorporating key weather variables (temperature, rainfall, and humidity) and soil quality indicators (pH, water content, salinity, texture, and organic matter). Given the increasing impacts of climate change and land degradation on food security, a probabilistic approach is essential for quantifying uncertainties in crop production. Using 10,000 simulated scenarios based on historical and field-derived parameter distributions, the model estimates an average rice yield of approximately 4.2 tons per hectare with a standard deviation of 0.2 tons per hectare, indicating relatively stable production under normal conditions. However, 20% of the simulations produce yields below 3.9 tons per hectare, reflecting elevated risks of crop failure during adverse environmental situations. Sensitivity analysis identifies rainfall and soil pH as the most influential variables, where extreme deviations may reduce yields by up to 35%. These findings offer critical evidence for policymakers and farmers to develop adaptive management strategies aimed at safeguarding sustainable rice production in the region.