Global sensitivity analysis of model parameters of dual crop coefficient in subsurface drip irrigated maize field

Abstract: The purpose of this research was to study parameters of double crop coefficient model that were sensitive to evapotranspiration of the maize field under shallow buried drip irrigation condition. The global sensitivity analysis of the model parameters was carried out by the extended Fourier amplitude sensitivity test (EFAST) method, and the sensitive parameters were selected to improve the efficiency and precision of the model by adjusting these sensitive parameters. The field experiments were carried out in Institute of Agricultural Science of Tongliao, Inner Mongolia, China (43.74°N, 122.55°E, 164m, elevation above mean sea level) in 2017. The results showed that when the values of model parameters were fluctuated within ±10%, the maximum soil evaporation was 18.72% higher than the minimum value, the maximum crop transpiration was 25.37% higher than the minimum value, and the maximum evapotranspiration was 19.9% higher than the minimum value during the whole growing season of maize. Soil evaporation was a process of the consumption of surface soil water, thus the dynamic difference of daily soil water storage in 1 m soil layer was small. Crop transpiration was a process of the consumption of soil water in the root layer of maize, thus the change of total crop transpiration resulted in a greater impact on the water consumption of 1m soil layer. When the total soil evaporation in the whole growing season of maize was simulated, the sensitive parameters were evaporable water on soil surface and basic crop coefficient of middle growth period, and their corresponding global sensitivity indexes were 0.662 and 0.321, respectively. Their global sensitivity indexes were 33.6-69.4 times higher than the mean of insensitive parameters. When the total crop transpiration during the whole growing season was simulated, the sensitive parameters were the threshold of soil water storage of roots free from water stress, basic crop coefficient of middle growth period and field water capacity, and their corresponding global sensitivity indexes were 0.569, 0.485 and 0.455, respectively. Their global sensitivity indexes were 34.5-43 times higher than the mean of insensitive parameters. The relationship between sensitive parameters and evapotranspiration were investigated and the results showed that after the topsoil was completely wet, the amount of soil water evaporated determined the amount of soil evaporation during the drying process, and total soil evaporation increased due to the increase of evaporable water on soil surface. The basic crop coefficient of middle growth period changed soil evaporation by affecting the soil evaporation coefficient. The total soil evaporation decreased due to the increase of basic crop coefficient of middle growth period. The higher threshold of soil water storage of roots free from water stress would result in the narrower interval of the available soil water in the root zone of maize, thus limiting crop transpiration. The total crop transpiration was inversely related to it. The basic crop coefficient of middle growth period and the total crop transpiration were positively correlated. The effect of basic crop coefficient of middle growth period on the total crop transpiration was much higher than the basic crop coefficient of initial and end period. The soil with high field water capacity could store more water for crop transpiration when irrigation and rainfall were high, and the total crop transpiration increased with the increase of field water capacity. The results provide theoretical support for exploring efficient water management methods of maize cultivation under shallow buried drip irrigation.