Performance of the Priestley–Taylor equation in the semiarid climate of North China

Abstract

The reference crop evapotranspiration is the most important component in the determination of agricultural water use. Since too many different methods are available in computing this value, the evaluation of these methods is the first step before using them. The objective of this paper is to assess the applicability of the Priestley–Taylor equation, widely used in many crop simulation models, by comparing its value with that of the Penman utilizing 46–50 years data series from six weather stations in the semiarid north china. The FAO Penman 1979 version is adopted because of its wide use in China. The input data includes monthly average sunshine duration, air temperature, relative humidity, air pressure and wind speed. Daily rate of monthly average is first obtained, and then multiplied by the days of each month to get the monthly total. During the calculation, the intermediate net radiation is estimated, and the soil heat flux is omitted. The required data, subjected to strict quality control before delivering and can make sure the accuracy of this study, is provided by the Climate Data Center, National Meteorological Bureau. The six locations are all national standard meteorological stations with long data record and are almost evenly distributed in the region. The evaluation was conducted at three time scales, annually, monthly and daily. The Priestley–Taylor equation underestimated the reference crop evapotranspiration significantly at yearly time scale, ranging from 128.6 to 364.1 mm, or 14–31% in terms of percentage. Spatially, it performed decreasingly according to Taiyuan > Shijiazhuang > Zhengzhou > Tianjin > Beijing > Jinan. At monthly time scale, the performance in most locations depends largely on month. In the humid months of July and August, it is comparable to the Penman equation. But when the wind speed is above certain level, the agreement decreased. In other months, its performance is unacceptable. At daily time scale in July and August in the best-performed year and the worst performed year, the Priestley–Taylor equation also performed well. At the latter two time scales, the spatial performing order varies from that at yearly time scale, and varies for different months and years. At all time scales considered, the performance of the Priestley–Taylor equation in the semiarid region is inversely affected by the ratio of aerodynamic term to radiation. This was supported by the correlations between this ratio and the ratio of the Priestley–Taylor value to the Penman value for annual, monthly and daily series. In addition, precipitation and wind speed also play a part in the performance. Years, months or days with higher precipitation or lower wind speed usually result in better performance, or vice versa. In summary, care should be taken when applying the Priestley–Taylor equation in the semiarid climate in north China. Temporally, it can be used in July and August and at daily time scale in these two months, but unsatisfactorily in other months and at yearly time scale. Spatially, except locations with wind speed higher than 2.61 and 2.36 m s −1 as Jinan in July and August, it is applicable in other locations in these two months. With the widespread use of this equation in crop simulation modeling, the possible modification to the coefficient α in the region deserves further investigation.