Penman-Monteith 模型估算潜在蒸散发的不确定性及其对径流模拟的影响
Alternative TitleUncertainty of estimating potential evapotranspiration by Penman-Monteith model and its influence on runoff simulation
Thesis Advisor郝兴明
Degree Grantor中国科学院大学
Place of Conferral北京
Degree Discipline理学硕士
Keyword潜在蒸散发 参数校正 SWAT 模型 径流模拟 喀什噶尔河流域 Potential Evapotranspiration Parameter Correction SWAT model Runoff Simulation Kashgar River Basin
Abstract蒸散发作为水量平衡和水循环的重要环节,该项参数的输入对 SWAT 分布式水 文 模 型 的 模 拟 精度 具 有 较 大 的 影 响 。 国 际 粮 农 组 织 ( FAO ) 推 荐的Penman-Monteith 方法估算潜在蒸散发在世界各地得到了广泛的应用和验证,但最新有研究表明该方法在干旱区的应用过程中会导致不可忽视的误差,主要是由于在干旱半干旱地区,植被覆盖率较低,远达不到 FAO 假设的完全被植被覆盖的稳定下垫面条件,这导致地表温度与空气温度差异较大,现在通用的 PM 公式仅通过空气气温拟合饱和水汽压曲线斜率和默认地表反照率为固定值 0.23 的做法必然会导致计算偏差。因此本研究利用 MODIS 遥感产品获取地表温度和地表反照率数据对这两项参数校正,探究了两项参数对 Penman-Monteith 方法的影响,以及该方法在干旱区应用过程中的不确定性。同时,潜在蒸散发作为关键的导入参数,它的准确计算可以进一步提高SWAT模型在干旱区的应用精度。虽然SWAT模型在参数率定的过程中会部分过滤掉参数的敏感性,但对于蒸发强烈的干旱区来说,这项数据的准确与否必然会影响模型的模拟效果。因此本研究基于参数修正前后 PM 公式计算的潜在蒸散发分别输入模型,并进行了对比分析,主要结论如下:1. 在干旱地区,通过空气气温一项参数拟合温度饱和水汽压曲线会导致 PM在计算过程中过度高估当地的潜在蒸散发,而设定地表反照率为 0.23 的做法会导致低估,通过同时对温度饱和水汽压曲线斜率的和地表反照率校正后,Penman-Monteith 的高估误差仍然存在。在西北干旱区 84 个国家标准气象站点,50%的站点表现为年均值都呈现高估状态,超过 30%以上的站点年均高估值都达到 100mm 以上。高估量最大站点(七角井站)年均高估为 681.52mm,全区年均高估 156.70mm。同时,也有部分站点表现为低估,年最大、最小和平均低估幅度分别为 323.28mm, 3.28mm 和 101.61mm,这些站点主要集中于平原植被覆盖较好的地区。在蒸发强烈的 7-8 月, FAO-56 Penman-Monteith 公式对 ET0 的高估更为明显,高估的站点数超过70%,高估幅度最高达到233mm,平均为45.26mm。就整个区域平均状况而言,年均高估量为 33.69mm,空间尺度上看植被覆盖度低的地区,这种高估更为显著。时间尺度上看主要集中与 7 至 8 月,平均高估29.18mm2.选取干旱区内典型区喀什噶尔河流域作为研究对象,设置恰其嘎站和克勒克站两个控制点,建立了 SWAT 分布式水文模型。设定 2003-2004 年为预热期,2005-2009 年为率定期, 2010-2011 作为验证期, 分别将基于传统 Penman-Monteith模型和参数校正后的 Penman-Monteith 计算的潜在蒸散发输入到模型,对两个控制站的月径流量进行模拟分析,利用 R2 和 NS 系数对模拟结果进行评价。研究结果表明:无论在率定期还是验证期,潜在蒸散发的准确计算显著提高了 SWAT模型在干旱区径流的模拟精度,因此,提高潜在蒸散发输入精度是减少 SWAT分布式水文模型不确定性的重要途经。
Other AbstractPotential evapotranspiration as an important link in water balance and water cycle, inputof this parameter has a great influence on the simulation accuracy of SWAT distributedhydrological model. The Penman-Monteith method recommended by the International Foodand Agriculture Organization of the United Nations (FAO) has been widely used andvalidated around the world, but the latest research has shown that the method can lead to anunignorable error in the application of the arid area, mainly due to the low vegetationcoverage in the arid area, It is far from the stable underlying condition of the broad grass thatthe FAO assumes. The difference between surface temperature and air temperature is large.Only by fitting the slope of saturated water vapor pressure curve by air temperature and thedefault surface albedo is a fixed value of 0.23 will inevitably lead to the calculation deviation.Therefore, this study uses MODIS remote sensing products to obtain surface temperature andsurface albedo data to correct these two parameters, and explores the influence of the twoparameters on the Penman-Monteith method and the uncertainty of the application of themethod in arid areas. At the same time, the potential evapotranspiration as a key importparameter, its accurate calculation can further improve the application accuracy of the SWATmodel in arid region. Although SWAT model partially filters out the sensitivity of parametersduring parameter rate determination, the accuracy of this data will inevitably affect thesimulation effect of the model for arid regions with strong evaporation. As a result, thepotential evapotranspiration PM calculated by the formula before and after parametercorrection is inputted into the model, and the main conclusions are as follows:1.The fitting of saturated water vapor pressure curves by a parameter of air temperaturein arid areas can cause PM to overestimate the local potential evapotranspiration in thecalculation process, while the setting of surface albedo of 0.23 can lead to underestimation.By simultaneously correcting the slope of saturated water vapor pressure curve and surfacealbedo, Penman-Monteith overestimation error still exists. Among the 84 national standardmeteorological stations in the northwest arid zone ,50% of the stations showed an overestimation of the average annual value, and more than 30% of the stations had an averageannual high valuation of more than 100 mm. The average annual overestimation of the largestsite (Qijiao jing) is 681.52 mm, and the average annual overestimation of the whole region is156.70 mm. Meanwhile, there are some sites that show underestimation, with annualmaximum, minimum and average underestimation of 323.28 mm,3.28mm and 101.61 mm,which are mainly concentrated in areas with better vegetation cover in the plain. Theoverestimation of the ET0 by the FAO-56Penman-Monteith formula was more pronounced inJuly-August, when the number of overvalued sites exceeded 70 per cent, with the highestoverestimation reaching 233 mm, averaging 45.26 mm. As far as the overall regional averageis concerned, the average annual overestimation is 33.69 mm, on a spatial scale, it is moresignificant in areas with low vegetation cover. At the time scale, the main focus was from Julyto August, with an average overestimation of 29.18 mm2. A SWAT distributed hydrological model was established by selecting the KashgarRiver basin in the typical arid region as the research object and setting up two control pointsof Qiaqiga station and Kleke station. Set 2003-2004 as preheating period ,2005-2009 ascalibration period ,2010-2011 as validation period, input potential evapotranspiration basedon traditional Penman-Monteith model and parameter-corrected Penman-Monteith calculationto model, simulate the monthly runoff of two control stations, and evaluate the simulationresults using R2 and NS coefficients. According to the results, the accurate calculation ofpotential evapotranspiration improves the simulation accuracy of SWAT model runoff in aridzone. So, improving the input accuracy of potential evapotranspiration is an important way toreduce the uncertainty of SWAT distributed hydrological models.
Subject Area自然地理学
Document Type学位论文
First Author Affilication中国科学院新疆生态与地理研究所
Recommended Citation
GB/T 7714
花顶. Penman-Monteith 模型估算潜在蒸散发的不确定性及其对径流模拟的影响[D]. 北京. 中国科学院大学,2020.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[花顶]'s Articles
Baidu academic
Similar articles in Baidu academic
[花顶]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[花顶]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.