汉江上游流域蒸散量计算方法的比较及改进

Comparison and Modification of Evapotranspiration Estimation Method in Upper Hanjiang Basin

利用汉江上游流域13个气象站的气象观测数据,对SWAT2000模型内嵌的3种潜在蒸散量计算方法:①Penman Monteith(PM)法、 ②Pristley Tayler(PT)法和③Hargreaves法以及FAOPenman Monteith(FAOPM)法、DeBruin法进行了蒸散量(ET0)计算方法的分析和评价。以气象站实测小型蒸发数据经过折算得到的陆面潜在蒸散量为标准,对模拟蒸散量结果进行检验,根据相关系数、解释方差、确定性系数以及相关曲线斜率等4个指标的比较,详细讨论了各方法的适用性及模拟精度,认为5种方法中PM法和FAOPM法模拟精度最高。但是进一步分析发现FAOPM法存在部分中间气象参量计算错误以及存在与实测数据偏差较大的情况,会影响到水文过程中其它子过程的模拟。因此选择PM法作为汉江上游流域蒸散量计算的首选方法,并通过数据分析提出利用高程对模拟结果进行修正,提高了蒸散量模拟的精度。

Five different evapotranspiration (ET_0) estimation methods, namely: Penman-Monteith (PM), Pristley-Taylor (PT), Hargreaves, FAO Penman-Monteith (FAO PM) and De Bruin, have been used to evaluate the validity and precision of simulated ET_0 results with the observed meteorological data from 13 weather stations in upper Hanjiang basin. The objective of this article is to find the best method among them and made further improvement on it to get a higher simulation precision on ET_0 simulation in study area. Based on four analysis index study such as correlation coefficient, explanation variance (EV), Nash-Sutcliffe Criterion (NSC) and correlation curve slope for the simulated results and the real ones calculated by 20-cm evaporating dish values with evaporation pan coefficient, PM and FAO PM methods are thought of better methods to yield rather satisfied estimation among the five selected approaches for and FAO PM methods considered both energy balance and water vapor diffusive theory in ET_0 simulation model construction while other three methods only considered latter elements. But the results also showed that PM method was better in correlation coefficient and EV indexes while FAO PM method is better in NSC and correlation curve slope indexes. In order to find the most favorable method among them, detailed studies have been done and the results show that FAO PM method has some shortcomings in calculation of the middle-process parameters. For example the simulated saturated water vapor in FAO PM method is lower than observed real water vapor value and the calculated solar radiation is larger than observed one. These will bring errors for some sub-processes simulation in hydrological studies. Therefore, PM method is recommended as the best ET_0 estimation approach in upper Hanjiang basin. The final precision of ET_0 by PM approach is further improved by introducing elevation modification to the simulated results. The coefficient between correlation curve slope and station elevations has reached 0.966. After elevation modification the magnitude of simulated ET_0 can be very near to the observed pint-sized evapotranspiration value.