Abstract:
The accurate determination of the latent heat flux of farmland is of great significance to the development of precise irrigation scheme and high agricultural production. The latent heat flux is often estimated by the Penman-Monteith (PM) model. In this model, the parameter, canopy resistance, is important but difficult to measure directly. The aim of this study was to find out an effective method to determine the canopy resistance for the estimation of latent heat flux in fields of winter wheat and summer maize in the southern Jiangsu, China. The experiment field of this study was located in Changzhou city, Jiangsu province (31°41 'N, 119°40 'E) . The area belongs to subtropical monsoon climate. The mean annual precipitation may reach 1 000 mm. The soil texture is loam clay with field water-holding capacity of 25% and wilting point of 9.6%. The Bowen ratio energy balance observation system was installed in the center of the field to record the water and heat flux values during the year from 2018 to 2020. The net radiation was measured at 2.5 m above ground. The air temperature and relative humidity were recorded at heights of 1.5 and 2.5 m. During the experiment, the infrared thermometer was adjusted according to the height of the crop canopy to continuously measure the temperature of the crop canopy. The wind speed and direction were observed at a height of 2.5 m. Soil heat flux was measured at 2 cm depth and rainfall was also measured. The Katerji-Perrier (KP) and Todorovic (TD) models were applied to optimize the canopy resistance in the P-M model. The performances of the two models were validated using the measured latent heat flux by the Bowen Raito and Energy Balance (BREB) method. The most suitable model for the winter wheat and summer maize was suggested based on the evaluation of the hourly time scales performance of the two models. Besides, the reasons for the error of the KP and TD models were analyzed. The results showed that during the growth period of winter wheat, the main meteorological factors showed a similar trend of change, and the daily mean value of net radiation showed a fluctuating trend of rise. The fitted slopes of linear regression model between measured and simulated latent heat flux by the two models were close to 1 in the two growing seasons of winter wheat and one growing season of maize. Both models of canopy resistance parameters had good simulation effects on latent heat flux simulation of winter wheat, with R2 and Nash coefficient not less than 0.84 and 0.86 respectively, but the accuracy of KP model was slightly higher than that of TD model. The KP model overestimated the latent heat flux of winter wheat and summer maize, while TD model overestimated the latent heat flux of summer maize. Saturated vapor pressure difference was the main factor affecting the errors of KP and TD sub-models of canopy resistance parameters, and the greater saturated vapor pressure difference would lead to larger absolute errors. The study provides scientific basis for local agricultural water management.