Development of yield forecast model using multiple regression analysis and impact of climatic parameters on spring wheat

Understanding the impacts of climate change in agriculture is important to ensure optimal and continuous crop production. The agricultural sector plays a significant role in the economy of Upper Midwestern states in the USA, especially that of North Dakota (ND). Spring wheat contributes most of the wheat production in ND, which is a major producer of wheat in the USA. This study focuses on assessing possible impacts of three climate variables on spring wheat yield in ND by building a regression model. Eighty-five years of field data were collected and the trend of average minimum temperature along with average maximum temperature, average precipitation, and spring wheat yield was analyzed using Mann-Kendall test. The study area was divided into 9 divisions based on physical locations. The minimum temperature plays an important role in the region as it impacts the physiological development of the crops. Increasing trend was noticed for 6 divisions for average minimum temperature and average precipitation during growing season. Northeast and Southeast division showed the strongest increasing trend for average minimum temperature and average precipitation, respectively. East-central division had the most decreasing trend for average maximum temperature. A significant relationship was established between spring wheat yield and climatic parameters as the p-value is lower than 0.05 level which rejects the null hypothesis. The regression model was tested for forecasting accuracy. The percentage deviation of error for the model is approximately ±30% in most of the years.