Suitability assessment of P-M model by measuring ET0 of turfs in Beijing, China

Abstract: Reference evapotranspiration (ET0) is an important parameter for agricultural irrigation scheme. Penman-Monteith (P-M) model is used extensively to simulate ET0 for its accuracy in Europe and America area. But it's rarely evaluated by lysimeter measurement in China. This study aimed to: 1) reveal ET0 characteristics in Beijing area; 2) evaluate the accuracy of P-M model under different weathers and at different scales; 3) research the effect of plant species on ET0 measurement and P-M model evaluation. The evapotranspiration of the cool-season grass Festuca arundinacea, the warm-season grass Buchloe dactyloides, and the native species Carex leucochlora during the growing seasons of 2012-2014 was measured by lysimeters. Meteorological parameters of the solar radiation, the temperature, the air relative humidity, and the wind speed were measured using a weather station and were put into P-M model to obtain ET0 values. Comparison was conducted between the P-M calculated ET0 and the measured values under different weather conditions and at different scales. The consistency was analyzed using the statistical parameters of the slope and the coefficient of determination (R2) of the linear regression, the root mean square error (RMSE), and the concordance index (d). The results indicated that the calculated ET0 showed consistent trends with the measured values at all the daily, weekly, and monthly scales. The peak of the monthly ET0 came in May, with the values of 4.18±0.27 (P-M), 4.43±0.98 (F. arundinacea), 3.96±0.23 (C. leucochlora), 3.53±0.25 mm/d (B. dactyloides). October had the lowest value. The monthly calculated ET0 showed a significant linear relationship with the solar radiation, the average air temperature, and the highest air temperature (P<0.01). The regression between ET0 and the solar radiation had the highest R2 of 0.885. The ratio of the calculated ET0 to the measured value varied with different weathers. It increased with the solar radiation decreasing from the sunny to the rainy day. The P-M model overestimated ET0 under overcast and rainy weathers. The RMSE and d between the P-M calculated ET0 and the measured values increased with the assessment scale decreasing. The measured ET0 differed significantly, sorted by F. arundinacea > C. leucochlora > B. dactyloides. The calculated ET0 was the highest consistent with the measured value of F. arundinacea, with the regression slope of 0.99-1.03, the lowest RMSE of 0.62-1.05 mm/d, and the highest d of 0.89-0.90. It's concluded that P-M model was applicable for Beijing area. But it could overestimate ET0 under overcast and rainy weathers or in cold spring.