Application of Normalized Difference Vegetation Index in Predicting Maize Yields: A Spaceborne Photography Solution to Agricultural Food Insecurity in Kenya

Abstract High levels of acute food insecurity affects approximately 2.8 million people in Kenya as of early march 2025 and as per the Global Hunger Index, kenya ranked 100th out of 127 countries, with agriculture highly vulnerable to climate change. Accurate crop yield prediction is critical to mitigating food insecurity. This study aimed to assess the effectiveness of the Normalized Difference Vegetation Index (NDVI) in predicting maize yield using a simple linear regression model. A retrospective observational study was conducted using historical maize yield data from the Kenyan Ministry of Agriculture and NDVI measurements from the Kenya Space Agency. NDVI was used as an independent variable in a linear regression model to estimate maize yield. Statistical analysis was performed using R software to evaluate the relationship between NDVI and yield. The study included multiple regions and seasons; no specific control group was applicable due to the observational design. A strong positive correlation (r > 0.86) was found between NDVI and maize yield. The regression model results were statistically significant (p < 0.05), indicating that NDVI reliably predicts maize yield. The model demonstrated consistent patterns across different datasets, confirming its robustness in yield estimation. NDVI is a strong predictor of maize yield and can support early yield forecasting and agricultural monitoring systems in Kenya. This approach may contribute to more effective strategies for managing food insecurity in the face of climate variability