EGI OpenIR  > 研究系统  > 荒漠环境研究室
基于通用陆面模式 CoLM 的新疆蒸散发模拟研究
袁秀亮
Subtype博士
Thesis Advisor李龙辉 ; Philippe De Maeyer
2018-06-05
Degree Grantor中国科学院大学
Place of Conferral新疆乌鲁木齐
Degree Discipline理学博士
Keyword陆面模式 蒸散发 绿洲农田生态系统 膜下滴灌 Land Surface Model Evapotrnspiration Oasis Agricultural Ecosystem Drip Irrigation under Plastic Mulch
Abstract新疆深居亚欧大陆内部,是世界上典型的干旱与半干旱区域。 水资源是制约当地经济发展与生态稳定的最重要因素。蒸散发作为连接陆地与大气之间的重要水汽通量,是本地区水循环过程中水分耗散的主要途径。近几十年来,新疆人类活动加剧,耕地面积不断扩张,气候变化剧烈,显著影响了本地区蒸散发的时空变化特征。因此,准确估算新疆蒸散发,探究其对气候变化和土地覆被变化的响应机制, 对于客观认识新疆水资源状况及制定合理的水资源分配方案具有积极的作用。基于机理过程的陆面模式是解决以上问题的一种有效手段,但由于缺乏反映绿洲农田生态系统膜下滴灌改变地表特征及其影响地表水热传输过程的相关模块, 因此, 利用陆面模式模拟新疆蒸散发的研究,仍然存在很大的不确定性。基于此,本研究以陆面模式 Common Land Model (CoLM) 为研究平台,首先对 CoLM 进行改进和完善, 然后结合气象数据及土地覆被数据,通过情景模拟实验来分析不同气候因子和土地覆被变化对实际蒸散发的相对贡献及影响机制,探究不同土地覆被类型蒸散耗水的变化特征。获得的主要结论如下:1) 首先针对绿洲农田生态系统,基于能量平衡理论, 将膜下滴灌的使用所产生的地表水热传输过程嵌入到 CoLM 陆面模式中,结合乌兰乌苏农业站涡度相关观测数据,开展 CoLM 陆面模式的站点验证及关键参数的敏感性分析研究,旨在提高其模拟水汽通量的精度。结果显示,模型模拟的蒸散发对灌溉参数 (灌溉发生时间及灌溉用水量) 具有很强的敏感性。在未考虑覆膜的情况下,灌溉模块的添加提高了蒸散发的总量,然而其模拟的蒸散发日变化波动剧烈,并未显示出明显的季节性变化特征,精度未明显提高,均方根误差从 76 W/m2 增加到了87 W/m2,线性拟合的斜率从 0.04 提高到 0.72,相关系数从 0.2 增加到 0.4。地膜模块的添加提高了模型模拟蒸散发的精度,均方根误差从 87 W/m2下降到了 44W/m2,线性拟合的斜率从 0.72 增加到了 0.92,相关系数从 0.4 增加到 0.75。地膜模块的添加降低了蒸散发日变化的波动性,并且表现出明显的季节性特征,与观测值具有很好的吻合性。2) 以 PRINCETON 气象数据作为驱动数据,以中国科学院新疆生态与地理研究所产生的土地利用及土地覆被数据作为地表参数化数据,模拟了 1970 到2014 年间新疆蒸散的时空变化特征。结果显示:新疆年均蒸散发量大约为 120.04(± 20.13) mm/年。研究时段内,总体上蒸散发在增加,其变化趋势通过了显著性检验 (p = 0.02),增加的速率为 0.53 mm/年。在空间上,蒸散发的变化在不同区域差异比较大,大部分区域蒸散发在增加 (69.9%的像元)。蒸散发增加相对较大的区域主要出现在新疆中部天山区域,增加的速率最高达到 1.98 mm/年。有 30.1%的区域显示蒸散发在减少。蒸散发减少的区域主要出现在新疆西南部边缘区域。通过情景模拟实验显示:新疆蒸散发年际变化主要受气候变化的影响,而土地覆被变化对蒸散发影响相对较小。1970 到 2014 年间,蒸散发总共增加了 13.8 mm,其中,气候变化导致新疆蒸散发增加了 11.6 mm,土地覆被变化导致新疆蒸散发增加了 2.2 mm。3) 在 1970 年到 2014 年,几乎所有的植被类型蒸散发都出现了显著增长的趋势 (p < 0.05),其增长速率在 0.46 ~ 0.72 mm/年之间。所有土地覆被类型的蒸散耗水总量的时间变化特征基本相似,即从 1970 年到 21 世纪初不断增加,并且在 21 世纪初达到最高,在之后的时间段内 (2001 - 2014),蒸散耗水总量呈减少趋势。作物类型蒸散耗水总量在新疆总的蒸散量中占有的比例在整个研究时段内呈递增趋势,从上世纪 70 年代初的 3.96%到 21 世纪初的 5.62%,而其它土地覆被类型占有的比例没有表现出明显的时间变化特征。本研究从陆面模式在干旱区应用过程中出现潜热模拟不准确的问题着手, 完善了陆面模式在模拟绿洲农田生态系统水汽通量中的膜下滴灌功能模块,提高了其模拟精度。本研究对陆面模式的完善部分,将为后续陆面模式在干旱区的研究工作提供重要的借鉴意义。另外,新疆蒸散发模拟工作以及蒸散发对气候变化和土地覆被变化的响应机制的研究,将为探究新疆实际蒸散发在区域水循环中的作用,维持新疆水资源的可持续发展提供科学支撑。
Other AbstractThe Xinjiang Uyghur Autonomous Region of China is located deep inside theEurasian continent, and is one of the typical arid and semi-arid regions in the world.Water resources are the most important factor that restricts the development of localeconomy and ecological stability. Evapotranspiration is an important water vapor fluxconnecting the land and the atmosphere, and is also the main way of water dissipationin the water cycle in the region. In recent decades, human activities with theexpansion of cultivated land, as well as the drastically changed climate, affectedsignificantly the temporal and spatial variation of evapotranspiration in the region.Therefore, accurately estimating evapotranspiration in Xinjiang and exploring itsresponse to climate and land cover change have an active role in objectivelyunderstanding the status of water resources and formulating a reasonable waterresource allocation strategy. The process-based land surface model is an effective toolto solve above issues. However, due to the lack of land-surface processes that reflectthe hydrothermal transfer induced by the use of drip irrigation under plastic mulch,there is still a great deal of uncertainty in estimating evapotranspiration by using landsurface models.Based the above background, in this dissertation, we selected the Common LandModel (CoLM) as a platform, and we first modified CoLM. Then, we analyzed theresponse of evapotranspiration to climate and land cover change, and analyzed thechanges of evapotranspiration for different land cover types, combining withreanalysis climate data. The main results are presented as follows:1) Firstly, for the oasis agro-ecosystem, based on the theory of energy balance,the surface water heat transfer process induced by using drip irrigation under plasticmulch is incorporated into the CoLM. Combining with the eddy covarianceobservations of Wulanwusu agricultural station, the validation of CoLM andsensitivity analysis of key parameters were carried out to improve the performance ofCoLM for simulating water vapor flux. Our results show that the model-simulated evapotranspiration has strong sensitivity to irrigation parameters (irrigation start timeand irrigation demand water). Without considering plastic mulch effects, the additionof irrigation module increased the total amount of evapotranspiration whereas thesimulated diurnal variation of evapotranspiration was poor, and failed to captureseasonal patterns of evapotranspiration: the root mean square error increased from 76W m-2 to 87 W m-2; the slope of the linear fit increased from 0.04 to 0.72; and thecorrelation coefficient increased from 0.2 to 0.4. The addition of plastic mulchincreased the accuracy of simulated evapotranspiration: the root mean square errordecreased from 87 W m-2 to 44 W m-2; the slope of the linear fit increased from 0.72to 0.92; and the correlation coefficient increased from 0.4 to 0.75. Moreover, theaddition of plastic film module reduced the variability of diurnal variation ofevapotranspiration and showed obvious seasonal characteristics, which was in goodagreement with the observed values.2) We analyzed the spatio-temporal changes of evapotranspiration in Xinjiangfrom 1970 to 2014, using PRINCETON meteorological data as the driven data andland use/cover dataset produced by Xinjiang Institute of Eocology and Geography,Chinese Academy of Sciences, as the surface parametric data. Our results show thatthe annual average evapotranspiration in Xinjiang is about 120.04 (± 20.13) mm/year.During the study period, the evapotranspiration increased significantly (p = 0.02) atthe speed of 0.53 mm/year. Spatially, the magnitude of evapotranspiration change isquite different in different regions, and evapotranspiration is increased in most regions(about 69.9% of pixels). The areas with increased evapotranspiration mostly occur inthe central regions in Xinjiang with a increase rate of up to 1.98 mm/year. About 30.1%of the area shows that evapotranspiration decreased. The area whereevapotranspiration decreased mainly occurred in the marginal area of southwesternXinjiang. Based on scenario experiments, our results show that the interannualvariation of evapotranspiration in Xinjiang is mainly affected by climate change,while the change of land cover has less influence on evapotranspiration. Between1970 and 2014, climate change lead to a increase of 11.6 mm in evapotranspirationand a 2.2 mm increase in evapotranspiration affected by land cover change in Xinjiang.3) During the study period between 1970 and 2014, the evapotranspiration showsan significant increase tendency (p < 0.05) for all the land cover types, and themagnitudes range from 0.46 to 0.72 mm/year. The changes of evapotranspiration overtime for all types of land cover show a similar temporal pattern. They increased from1970 to the beginning of the 21st century, and then reached the highest level. In thefollowing 10 years, the total amount of evapotranspiration turned to decrease. Theproportion of evapotranspiration of crop types in Xinjiang's total evapotranspirationhas been increasing during the study period from 3.96% at the beginning of the 1970sto 5.62% at the beginning of the 21st century, while the proportions of other landcover types did not show significant change characteristics.This dissertation improved CoLM in simulating the water vapor fluxes in oasisagro- ecosystem. The improved part of CoLM in this study will provide importantreference for the research work in arid and semi-arid region. In addition, the study ofthe response of evapotranspiration to climate and land cover change will providescientific support for exploring the role of evapotranspiration in regional water cycleand for maintaining the sustainable development of Xinjiang water resources.
Subject Area地图学与地理信息系统
Language中文
Document Type学位论文
Identifierhttp://ir.xjlas.org/handle/365004/14943
Collection研究系统_荒漠环境研究室
Affiliation1.中国科学院新疆生态与地理研究所;
2.根特大学
Recommended Citation
GB/T 7714
袁秀亮. 基于通用陆面模式 CoLM 的新疆蒸散发模拟研究[D]. 新疆乌鲁木齐. 中国科学院大学,2018.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[袁秀亮]'s Articles
Baidu academic
Similar articles in Baidu academic
[袁秀亮]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[袁秀亮]'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.