Application of Agrometeorological Models for Maize Yield Prediction in Kaduna State

Crop yield prediction models combining meteorological inputs with crop growth parameters offer significant operational value for food security planning and agricultural insurance programs. This study applies agrometeorological yield prediction models to estimate maize production in Kaduna State using seasonal meteorological data from NiMet stations at Kaduna and Kafanchan for the period 2016 to 2022. The FAO-AQUACROP model was parameterized for Nigerian maize varieties based on field calibration data from the Institute for Agricultural Research, Samaru, and driven by observed temperature, rainfall, and evapotranspiration inputs. Simulated yields were validated against maize production statistics from the Kaduna State Agricultural Development Agency. Results demonstrate that AQUACROP explains 79 percent of the variance in observed district-level maize yields, with RMSE of 0.41 tonnes per hectare. Sensitivity analysis reveals that rainfall during the tasseling and grain-fill stages of August to September is the most critical meteorological driver of yield variability, with a 20 percent rainfall deficit in this period producing an average 31 percent yield reduction. Operational application of the model to NiMet seasonal outlooks for 2023 produced a district-level yield prediction within 8 percent of subsequently observed outputs. Recommendations include formal adoption of AQUACROP-based yield prediction within Kaduna State's agricultural early warning system and training of agricultural development program staff in model operation. Keywords: AQUACROP, maize, Kaduna, yield prediction, agrometeorological.