Estimation of evapotranspiration and sensitivity to climate and the underlying surface based on the Budyko Framework
ZHANGDan1,2,, LIANGKang3, NIERong4, GURenying5 1. State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering,Hohai University,Nanjing 210098,China2. Key Laboratory of Watershed Geographic Sciences,Nanjing Institute of Geography and Limnology,Chinese Academy of Sciences,Nanjing 210008,China3. Key Laboratory of Water Cycle and Related Land Surface Processes,Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences,Beijing 100101,China4. PLA University of Science and Technology,Nanjing 211101,China5. Ningbo Yinzhou Meteorological Bureau,Ningbo 315194,China 收稿日期:2016-01-4 修回日期:2016-05-10 网络出版日期:2016-06-20 版权声明:2016《资源科学》编辑部《资源科学》编辑部 基金资助:国家自然科学基金项目(41401032)江苏省自然科学基金项目(BK20141059)河海大学水文水资源与水利工程科学国家重点实验室开放研究基金项目(2014490811) 作者简介: -->作者简介:张丹,女,河南周口市人,助理研究员,研究方向为水热平衡及其对环境变化的响应。E-mail:nuistgiszd@163.com
关键词:Budyko假设;蒸散发;植被;气候;下垫面;敏感性;流域 Abstract Evapotranspiration is a key processes in water cycles and energy balance. In this study,evapotranspiration was estimated based on water-energy balance theory using hydrological and meteorological data from 71 typical catchments in China from 1960 to 2000. Sensitivities for evapotranspiration to precipitation,potential evapotranspiration and characterizations of the underlying surface (represented by parameter α)are further investigated in catchments classified by four groups:under all conditions without parameter α (group 1),under all conditions with one parameter α (group 2),under different climate zones (group 3)and different predominant land covers (group 4). The results show that α at group 2 is 2.202,and smaller in humid catchments than arid catchments in group 3. α at forest catchments in group 4 are the smallest,followed by grass catchments and mixed catchments. The NSE of evapotranspiration estimation by the original Budyko equation is only 0.64 at group 1,which can be improved effectively by adding parameter α. The NSE of evapotranspiration estimation are 0.80,0.81 and 0.83 in group 2,group 3 and group 4,respectively. In particular, the NSE at forest catchments is 0.88. Evapotranspiration in group 3 is most sensitive to precipitation in arid catchments,followed by α and potential evapotranspiration,while it is most sensitive to α in humid catchments,followed by potential evapotranspiration and precipitation. In group 4,evapotranspiration is most sensitive to α in forest and mixed catchments,while it is most sensitive to precipitation in grass catchments. The results are helpful for water-energy balance modeling,hydrological predictions in ungauged basins and water management decision-making.
Keywords:Budyko framework;evapotranspiration;vegetation;climate;underlying surface;sensitivity;basin -->0 PDF (2356KB)元数据多维度评价相关文章收藏文章 本文引用格式导出EndNoteRisBibtex收藏本文--> 张丹, 梁康, 聂茸, 顾人颖. 基于Budyko假设的流域蒸散发估算及其对气候与下垫面的敏感性分析[J]. , 2016, 38(6): 1140-1148 https://doi.org/10.18402/resci.2016.06.13 ZHANGDan, LIANGKang, NIERong, GURenying. Estimation of evapotranspiration and sensitivity to climate and the underlying surface based on the Budyko Framework[J]. 资源科学, 2016, 38(6): 1140-1148 https://doi.org/10.18402/resci.2016.06.13
图1给出了71个典型流域的蒸散发-降水-潜在蒸散发的相关关系曲线,可以看出所有的点均分布在Budyko水热耦合假设边界条件之内。所有流域的森林覆盖率平均值为42.6%(表1)。在方案2中(Fu_1参,对所有流域不进行分组),α取值为2.202。在方案3中(Fu_2参,基于气候指标对流域进行分组),干旱流域个数占所有流域数量的53.5%,其森林覆盖率平均值为31.3%,α取值为2.211;湿润流域个数占所有流域数量的46.5%,其森林覆盖率平均值为55.7%,α取值略小于干旱流域,为2.188。在方案4中(Fu_3参,基于植被类型对流域进行分组),森林流域个数占所有流域的39.4%,森林覆盖率平均值为66.4%,α取值为2.032;草地流域个数占所有流域的23.9%,森林覆盖率平均值仅为14.2%,α取值为2.252;混合流域个数占所有流域的36.6%,森林覆盖率平均值为35.6%,α取值为2.295。总的来说,森林流域的α最小,其次为草地流域,混合流域的α值最大,这说明参数α有效地反映了流域的下垫面特征。 显示原图|下载原图ZIP|生成PPT 图171个典型流域蒸散发-降水-潜在蒸散发关系曲线 -->Figure 1Relationships among evapotranspiration-precipitation-potential evapotranspiration at the 71 typical catchments -->
3.2 不同方案模拟结果分析
表2和图2给出了4种方案对蒸散发的模拟结果。总的来说,4种方案的蒸散发估算相对误差均在14%以内,说明本文选择Budyko模型和傅抱璞模型对蒸散发进行估算是合理的。其中,Budyko模型的模拟效果最差,效率系数仅为0.64,决定性系数、绝对误差和相对误差分别为0.83、59.2mm和13.8%,其模拟值较实测值普遍偏高。加入反映下垫面特征的参数α后(Fu_1参),模拟效果较Budyko模型得到了较大的改进,效率系数为0.80,绝对误差和相对误差分别为45.8mm和10.4%。对气候指标进行分类后(Fu_2参),模型的效率系数提高到0.81,绝对误差和相对误差分别为45.5mm和10.4%,且湿润流域的模拟效果明显优于干旱流域。对植被类型进行分类后(Fu_3参),模型的效率系数在4种方案中最高(0.83),绝对误差和相对误差分别为41.7mm和9.5%,其中森林流域的模拟效果最好,效率系数达0.88,相对误差仅为7.6%;草地流域的模拟效果最差,效率系数仅为0.52,相对误差为11.6%,混合流域的模拟效果介于森林流域和草地流域之间。结果表明,对研究区的气候和植被类型进行分组,估算不同气候和植被类型条件下的参数α,可以有效地提高模型的模拟精度。 显示原图|下载原图ZIP|生成PPT 图2基于Budyko假设的不同方案蒸散发观测值与估算值相关关系 -->Figure 2Relationships between observed and estimated evapotranspiration based on Budyko framework -->
Table 1 表1 表1不同气候和植被条件下模型参数对比 Table 1Comparison of parameter α under different climate and vegetation cover conditions
方案
分组
森林覆盖率 /%
流域个数所占百分比/%
参数 α
Budyko
-
42.6
100.0
-
Fu_1参
-
42.6
100.0
2.202
Fu_2参
干旱
31.3
53.5
2.211
湿润
55.7
46.5
2.188
Fu_3参
森林
66.4
39.4
2.032
草地
14.2
23.9
2.252
混合
35.6
36.6
2.295
新窗口打开 Table 2 表2 表2不同方案蒸散发模拟精度统计 Table 2Accuracies of estimated evapotranspiration by different categories
敏感性分析是在确定性分析的基础上,分析模型输入项的变化对模型可能产生的影响。以降水、潜在蒸散发和反映下垫面特征的参数α分别增加10%和减少10%为例,表3给出了4种方案中蒸散发对输入项变化的敏感性。 在Budyko模型中,蒸散发对降水的变化最为敏感,降水增加10%蒸散发增加5.3%,降水减少10%蒸散发减少5.9%,而对潜在蒸散发的变化敏感性相对较小,潜在蒸散发增加10%蒸散发增加4.2%,而潜在蒸散发减小10%蒸散发减少4.8%。加入反映下垫面特征的参数α后(Fu_1参),蒸散发对α的变化最为敏感,α增加10%,蒸散发增加5.4%,而α减少10%,蒸散发减少7.1%;蒸散发其次对降水最为敏感,对潜在蒸散发的敏感性最小。 在对气候指标进行分类后(Fu_2参),蒸散发对α的敏感性增强,α增加10%,蒸散发增加6.6%,尤其是当α减少10%时,蒸散发减少8.7%。蒸散发对降水和潜在蒸散发的敏感程度相当。值得注意的是,在干旱流域和湿润流域蒸散发对输入要素的敏感性有明显的差异。在干旱流域,蒸散发对降水的变化最为敏感,其次是α,其对潜在蒸散发的敏感性最小;而对于湿润流域而言,蒸散发对α和潜在蒸散发的敏感性明显大于降水。图3a和图3b分别以71个流域中最干旱的流域(干燥指数3.7)和最湿润的流域(干燥指数0.5)为例,给出了降水和潜在蒸散发同时变化的情况下蒸散发的变化。可以看出,干旱流域的蒸散发变化率等值线呈现显著的东西向变化,说明蒸散发变化对降水的变化最为敏感;而在最湿润流域的蒸散发变化率等值线呈现显著的西南-东北向变化,说明蒸散发对降水和潜在蒸散发的敏感性大致相当,其中对潜在蒸散发的敏感性略大于降水。 在对植被类型进行分类后(Fu_3参),蒸散发对α的变化最为敏感,α增加10%,蒸散发增加5.7%,α减少10%,蒸散发减少7.4%;蒸散发对降水的敏感性次之,降水增加10%,蒸散发增加5.5%,降水减少10%,蒸散发减少6%;蒸散发对潜在蒸散发的敏感性最小,潜在蒸散发增加10%,蒸散发仅增加4%,潜在蒸散发减少10%,蒸散发仅减少4.6%。对比3种植被类型的流域,森林流域和混合流域蒸散发对降水的减少最为敏感,降水减少10%,蒸散发分别减少7.6%和6.6%,草地流域蒸散发对降水的增加最为敏感,降水增加10%,蒸散发增加7.2%,且3种植被类型的流域均对潜在蒸散发的敏感程度最小。图4a和图4b分别以71个流域中森林覆盖率最高(88.6%)和草地覆盖率最高(73.9%)的两个流域为例,给出了降水和潜在蒸散发同时变化的情况下蒸散发的变化。可以看出,森林流域蒸散发对降水和潜在蒸散发的敏感性相似,而草地流域蒸散发对降水的敏感性明显大于潜在蒸散发。 Table 3 表3 表3蒸散发对降水、潜在蒸散发和下垫面参数α的敏感性分析 Table 3Sensitivies of evapotranspiration to precipitation,potential evapotranspiration and parameter α (%)
方案
分组
降水+10%
降水-10%
潜在蒸散发+10%
潜在蒸散发-10%
α+10%
α-10%
Budyko
-
5.3
-5.9
4.2
-4.8
-
-
Fu_1参
-
5.4
-5.9
4.1
-4.7
5.4
-7.1
Fu_2参
干旱
7.2
-7.7
2.3
-3.0
4.9
-6.6
湿润
4.0
-4.6
5.5
-6.1
5.8
-7.5
全部
5.4
-4.8
5.2
-5.7
6.6
-8.7
Fu_3参
森林
4.3
-7.6
2.4
-2.9
4.6
-6.2
草地
7.2
-6.6
3.5
-4.0
5.0
-6.5
混合
6.0
-6.6
3.5
-4.0
5.0
-6.5
全部
5.5
-6.0
4.0
-4.6
5.7
-7.4
新窗口打开 显示原图|下载原图ZIP|生成PPT 图3干旱流域和湿润流域蒸散发对降水和潜在蒸散发变化的敏感性 -->Figure 3Sensitivities of evapotranspiration to precipitation and potential evapotranspiration at arid catchment and humid catchment -->
显示原图|下载原图ZIP|生成PPT 图4草地流域和森林流域蒸散发对降水和潜在蒸散发变化的敏感性 -->Figure 4Sensitivities of evapotranspiration to precipitation and potential evapotranspiration at grass catchment and forest catchment -->
3.4 讨论
表4给出了研究区所有流域1980-2000年间土地利用类型的平均变化状况。可以看出,1980-2000年期间,研究区各土地利用类型变化情况不大,因此认为本文选用2000年的土地利用状况来代替流域多年平均特征是合理的。 值得注意的是,本研究基于降水、潜在蒸散发和反映下垫面特征的参数α三者之间的相互独立性,分析了蒸散发对三者的敏感性。但事实上三者之间并不是完全独立的。比如降水增加会改变局地的湿润状况,进而影响其温度、感热等,从而对蒸发能力(潜在蒸散发)产生影响;局地湿润状态的改变,对其植被类型的分布也会产生一定的影响,从而影响下垫面特征参数α。对比图3a和图4a、图3b和图4b可以发现,所选择的干旱流域和草地流域、湿润流域和森林流域蒸散发对降水和潜在蒸散发的敏感性有很高的相似性,这也反映了在水分限制(E0>P)的区域,气候较为干旱,植被的生长有限;而在能量限制(E0≤P)的区域,气候较为湿润,有利于植被的生长,植被和气候互相影响、互相反馈。尽管如此,这种非完全独立性并不会从实质上影响本研究的结论。此外,蒸散发模型考虑下垫面特征参数后,并不是所有下垫面的蒸散发模拟结果都能够得到显著的提高,如干旱流域、草地流域、混合流域等,下一步工作将进一步细化分组方案,以加强这些地区蒸散发影响因素的研究。 Table 4 表4 表41980年和2000年研究区各类土地利用类型所占百分率统计 Table 4Ratios of different types of landuse in the study area in 1980 and 2000(%)
本文基于Budyko水热耦合假设,采用全国71个典型流域的水文和气象数据,研究了气候和植被对蒸散发的影响。结果表明: (1)傅抱璞模型中下垫面参数α可以有效地反映流域的下垫面特征,其中湿润流域和森林流域的α相对较小,分别为2.188和2.032; (2)加入反映下垫面特征参数的蒸散发模型能有效提高蒸散发的估算精度,其中以湿润流域和森林流域的蒸散发模拟精度较高,其效率系数NSE分别为0.72和0.88; (3)总的来说,当加入反映下垫面特征的参数后,蒸散发对下垫面参数α的变化最为敏感,其次是降水,蒸散发对潜在蒸散发的敏感性最小。 本研究表明划分不同的气候和植被类型区域可以有效的提高流域蒸散发的估算精度,这为蒸散发的区域性特征分析及稀缺资料地区水文研究提供了有意义的参考。下一步将考虑选择更多的典型流域,通过对同一气候区的不同植被类型进行分组来估算流域蒸散发,为蒸散发区域特征的系统研究提供理论依据;同时本研究仅限于多年平均蒸散发的估算,年尺度甚至日尺度蒸散发的研究有待于进一步开展。 The authors have declared that no competing interests exist.
Allan RG,Pereira LS,RaesD,et al.Crop Evapotranspiration-Guidelines for Computing Crop Water Requirements-FAO Irrigation and Drainage Paper 56[R]. Rome:FAO,1998. [本文引用: 2]
[Liu CM,ZhangD.Temporal and spatial change analysis of the sensitivity of potential evapotranspiration to meteorological influencing factors in China [J]. Acta Geographica Sinica,2011,66(5):579-588.] [本文引用: 1]
[LiuY,Zhou MC,Chen ZJ,et al.Sensitivity of the potential evapotranspiration to climate and vegetation in Hanjiang River basin based on S-W Model [J]. Transactions of the Chinese Society of Agricultural Engineering,2013,29(10):92-100.] [本文引用: 1]
[4]
Priestley C HB,Taylor RJ. On the assessment of surface heat flux and evaporation using large-scale parameters [J]. Monthly Weather Review,2009,100:81-92. [本文引用: 1]
[Wang GQ,Zhang JY,He RM,et al.Trends of temperature change in middle of Yellow River and its impact to the evaporation potential [J]. Journal of Water Resources & Water Engineering,2007,18(4):32-36.] [本文引用: 1]
[Qiang XM,Cai HJ,WangJ.Comparative study of crop evapotranspiration measured by Bowen ratio and lysimeter [J]. Transactions of the CSAE,2009,25(2):12-17.] [本文引用: 1]
[Li SE,Kang SZ,Zhu ZL,et al.Research progress of measurement of land surface evapotranspiration based on eddy covariance technology [J]. Scientia Agricultura Sinica,2008,41(9):2720-2726.] [本文引用: 1]
[Zhang RH,Du JP,SunR.Review of estimation and validation of regional evapotranspiration based on remote sensing [J]. Advances in Earth Science,2012,27(2):1295-1307.] [本文引用: 1]
[10]
CouraultD,SeguinB,OliosoA.Review on estimation of evapo-transpiration from remote sensing data:From empirical to numerical modeling approaches [J]. Irrigation & Drainage Sys-tems,2005,19(3-4):223-249.
[Qiu QF,ZengY,Miao QL,et al.Estimation of annual actual evapotranspiration from nonsaturated land surfaces with conventional meteorological data [J]. Science in China,2003,33(3):281-288.] [本文引用: 1]
[Mo XG,Liu SX,Yu HN,et al.Seasonal variation of energy budget and evapotranspiration partitioning in wheat field [J]. Acta Geographica Sinica,1997,52(6):536-542.]
[Sun FB,Yang DW,Liu ZY,et al.Validation of coupled water-energy balance in the Haihe River basin and inland river basins [J]. Journal of China Hydrology,2007,27(2):7-10.] [本文引用: 4]
[17]
Han SJ,Hu HP,Yang DW,et al.Irrigation impact on annual water balance of the oases in Tarim Basin,Northwest China [J]. Hydrological Processes,2011,25(2):167-174. [本文引用: 1]
[Fu BP.On the calculation of the evaporation from land surface [J]. Scientia Atmospherica Sinica,1981,5(1):23-31.] [本文引用: 1]
[19]
Budyko MI.The effect of solar radiation variations on the climate of the Earth [J]. Tellus Series A-dynamic Meteorology & Oceano-graphy,1969,21(5):611-619. [本文引用: 1]
[Sun FB,Yang DW,Liu ZY,et al.Study on coupled water-energy balance in Yellow River basin based on Budyko Hypothesis [J]. Journal of Hydraulic Engineering,2007,38(4):409-416.] [本文引用: 5]
[21]
ChenX,NeginA,WangD.Modeling interannual variability of seasonal evaporation and storage change based on the extended Budyko framework [J]. Water Resources Research,2013,49(9):6067-6078.
[22]
ZhangY,Chiew F H S. Estimation of mean annual runoff across southeast Australia by incorporating vegetation types into Budyko framework,Australian [J]. Journal of Water Resources,2012,15(2):109-120. [本文引用: 3]
[LiB,Li LJ,Qin YC,et al.Impact of climate variability on streamflow in the upper and middle reaches of the Taoer River based on the Budyko hypothesis [J]. Resources Science,2011,33(1):70-76.] [本文引用: 1]
[24]
Sun SL,Chen HS,Ju WM,et al.Effects of climate change on annual streamflow using climate elasticity in Poyang Lake Basin,China [J]. Theoretical & Applied Climatology,2013,112(1-2):169-183. [本文引用: 1]
[25]
ZhangL,PotterN,HickelK,et al.Water balance modeling over variable time scales based on the Budyko framework-Model development and testing [J]. Journal of Hydrology,2008,360(1-4):117-131. [本文引用: 1]