Maize is a very important crop in the USA. However, maintaining a consistent high maize yield is a challenge for farmers as it varies due in part to the effects of spatial and temporal variability of weather conditions, spatial soil variability and other biotic and abiotic factors. Crop simulation models have the potential to help understand the impact of weather variability on plant growth and development as the models integrate the soil-plant-atmosphere complex.

The objectives of this study were to determine the impact of the inter-annual weather variability on maize growth and development and to determine the relation between the simulated drought stress, biomass production, and yield using a systems analysis approach. The CSM-CERES-Maize model was calibrated using data from experiments conducted in Chula, Georgia during the 2008, 2009 and 2010 growing seasons. The effect of drought stress for different growth stages (vegetative, flowering and grain filling) were calculated as the difference between simulated biomass for potential production and the actual simulated biomass divided by the cumulative stress index for a particular growth stage.

Long term simulations were conducted for five different planting dates (every 20 days starting in April 5^th) in each of 30 years using observed daily weather data from a nearby location (Tifton, GA), soil profile characteristics, and common agronomic practices used by farmers. The water stress coefficient related to growth, as determined by the CSM-CERES-Maize model, was analyzed to identify its relation with simulated biomass. The probability of yield reduction due to the effects of the inter-annual weather variability, especially low precipitation, for the five planting dates was estimated. Irrigation needs for potential yield were estimated for each planting date.

On average, the estimated effect of drought stress was 472, 351, and 307 kg of aboveground biomass ha^-1 per unit of drought stress index for the vegetative, flowering and grain filling periods, respectively. The results of the long-term simulations concurred with many physiological studies showing that the period from planting to anthesis shortened and the period from anthesis to physiological maturity increased as the planting date was delayed from April to the end of June. Both the average and standard deviation of the simulated yield increased as the planting date was delayed. The total cumulative drought stress and the irrigation needed decreased when delaying planting. However, the proportion of the simulated cumulative drought stress on grain filling increased from 33% to 60% as the planting dates was delayed to the end of June.

Being able to estimate the value of cumulative drought stress sometime prior to harvest could help with predicting final yield. This study showed how crop models could help characterize the impact of the inter-annual weather variability on drought stress index, on irrigation requirements, and the overall impact on maize yield using long-term weather data.