Predictive models for reference evapotranspiration in Gaoyou irrigation district

Abstract

【Objective】 Gaoyou irrigation district is located at the low reaches of the Yangtze River in northern Jiangsu province, China. To improve its irrigation management, we compared four models for predicting the reference evapotranspiration (ET0) in the district. 【Method】 The analysis was based on meteorological data measured from 2003 to 2017 and the temperature forecasted from 2016 to 2017 in the district, from which we calculated the ET0 using the Penman-Monteith (PM) formula recommended by FAO-56. Using these calculated ET0, we forecasted its change using the Blaney-Criddle (BC), Hargreaves-Samani (HS), McCloud (MC) and reduced PM (PMT) model, respectively. 【Result】 For forecasting ET0 1 to 7 days in advance, the average root mean square error of the BC, HS, MC and PMP model was 1.07, 1.00, 1.16, 0.99 mm/d, respectively; their associated average mean absolute error was 0.85, 0.74, 0.94, 0.75 mm/d, respectively; their associated average correlation coefficient with measured data was 0.79, 0.81, 0.76, 0.81, respectively. Overall, the results of HS and PMT model are comparable and both models are superior to other models for forecasting the ET0 up to 7 days in advance. 【Conclusion】 Among the four models we compared, the HS and PMT models are more accurate for predicting ET0 change 1-7 days in advance for the Gaoyou irrigation district.

Keywords

• reference evapotranspiration
• temperature forecast
• irrigation