IMHE OpenIR  > 数字山地与遥感应用中心
若尔盖高原生态系统变化遥感精细表征研究
Alternative TitleA research on quantifying ecosystem fine changes of the Zoige Plateau based on remote sensing
Language中文
夏浩铭
Thesis Advisor李爱农
2016
Degree Grantor中国科学院大学
Place of Conferral北京
Degree Name博士
Degree Discipline自然地理学
Keyword若尔盖高原 遥感精细表征 过水面 覆盖度 蒸散发 物候期
Other Abstract

随着全球气候变化与人类活动的共同作用,陆地生态系统的变化日益突出,呈现出生态系统稳定性降低、生态系统提供的生态服务功能减弱等问题,严重影响到陆地生态系统的可持续发展,并进而威胁到人类社会的生存与发展。因此,精确获取陆地生态系统的动态信息,实现陆地生态系统的有效监测是当前迫切需要解决的问题。近年来遥感对地观测技术的快速发展,为实现陆地生态环境的大范围、高时效与连续监测提供了非常有效的手段,使其逐渐成为当前陆地生态系统大范围动态监测的重要途径。然而,在当前生态系统遥感监测方法研究中,由于单一遥感观测数据在时空表达方面的局限性,导致已有方法无法精确且有效地表征陆地生态系统的动态特征,限制了遥感技术在该领域的进一步发展。为突出解决这一关键问题,本文选择地处我国青藏高原东缘、且其生态系统较为脆弱的若尔盖高原为研究区。近年来,在气候变化和人类活动干扰的共同作用下,若尔盖高原生态系统面临着湿地萎缩、草地退化和水资源量减少等生态环境问题。针对该区域面临的生态问题,本文重点探寻以其为代表的高寒湿地草地生态系统遥感精细表征方法研究。具体研究工作从基于遥感技术的生态系统变化空间精细表征和时间精细表征着手。在空间尺度的表征方法研究方面,针对研究区水陆过渡带和稀疏草地区难以监测的具体问题,本文将超高空间分辨率无人机数据与中高空间分辨率遥感数据协同,开展若尔盖高原湿地草地遥感亚像元尺度的湿地过水面积百分比与草地覆盖度精细表征方法研究;在时间尺度的表征方法研究方面,针对研究区草地生长过程与水热条件动态监测问题,本文以具有长时间序列且高频观测的中空间分辨率遥感数据(MODIS)为依托,基于遥感物候监测与地表蒸散发估算方法,开展气候变化背景下的若尔盖高原草地生长过程监测与地表蒸散发的时间序列遥感精细表征方法研究。通过以上研究工作,论文取得如下结论:(1)在水陆过渡带遥感精细表征研究方面,本文选择湿地过水面这一湿地生态系统变化的综合指示器,基于无人机(UAV)和Landsat-8 OLI数据,开展湿地过水面亚像元精细表征方法研究。结果显示:首先,超高分辨率的UAV影像(0.16m)能够精确地捕捉湿地过水面信息。其次,基于2014年UAV和Landsat-8 OLI数据,开展线性光谱分解模型、人工神经网络模型和回归树模型三种典型亚像元湿地过水面制图方法的建模研究,经验证和对比分析可知回归树模型所获得精度最高(R2=0.933,RMSE=8.73%),而且该方法应用于2013年湿地过水面监测时同样获得较好的预测效果,通过UAV数据验证得到该模型的预测精度较高(R2=0.986,RMSE=8.84%)。研究结果证明了该部分研究建立的回归树模型可有效地应用于时间序列湿地过水面的制图,能够实现面向湿地过水面的遥感精细表征。(2)在稀疏草地区遥感精细表征研究方面,本文选择草地生态系统变化的指示器草地覆盖度这一重要指标,基于UAV和Landsat-8 OLI数据,开展草地覆盖度亚像元精细表征方法研究。结果显示:首先,构建了一种基于UAV可见光波段差异植被指数(VDVI)的像元二分法模型,该方法能够有效地估算UAV可见光影像的亚像元覆盖度信息。其次,基于2014年UAV覆盖度和Landsat-8 OLI数据,选取三类典型的亚像元覆盖度估算模型(经验模型、像元二分模型和三波段梯度差模型)进行建模和评价。结果显示本研究所构建的经验模型估算精度最高(R2=0.901,RMSE=6.55%),能够用于若尔盖高原草地覆盖度的精细制图。该方法应用于2013年草地覆盖度监测时同样获得较好的预测效果,通过UAV数据验证所得预测结果精度较高(R2=0.736,RMSE=8.48%)。研究结果表明了该部分研究所建立的经验模型可以有效地监测草地覆盖度时间序列变化,能够实现面向草地退化的遥感精细表征。(3)在草地生长过程遥感精细表征研究方面,本研究通过动态阈值法的遥感物候监测方法,基于高时间分辨率的MODIS影像,对若尔盖高原草地物候期时间精细表征进行了研究。研究结果表明,该研究所采用的草地生长物候遥感监测方法能够较为准确地反映若尔盖高原草地的生长过程以及年际动态特征,实现研究区时间尺度上的草地生长过程遥感精细表征。本研究通过对研究区各像元的草地生长始期(SOG)、末期(EOG)和生长季长度(LOG)等物候期监测发现,各物候期的发生特征均与研究区的水热状况密切吻合,从南向北、从低海拔向高海拔和地表水分从高向低,生长始期逐渐变晚、末期逐渐变早、生长季长度逐渐变短。而物候期的年际变化特征也反映出其与研究区气温、降水等关键气候因子密切相关。(4)在水热条件遥感精细表征研究方面,本研究选取对水热平衡至关重要的参数——地表蒸散发,采用蒸散发估算的经典模型(SEBS模型),基于高时间分辨率的MODIS影像,对若尔盖高原地表蒸散发时间精细表征方法进行了研究。研究结果表明,该研究采用的蒸散发遥感监测方法能够有效地反映研究区年内蒸散发时空分布特征,实现研究区时间尺度上蒸散发的遥感精细表征。本研究通过对研究区各像元蒸散发的时空分布监测发现,蒸散发大小与地表的水热条件密切相关。在高原西南部和东部的干热河谷区蒸散发最大,在高原中北部的河流和湖泊附近蒸散发次之,在草地沙化、退化区和海拔较高的山区蒸散发最小。月蒸散发最大值和最小值分别出现在7月和2月。总之,基于以上四个方面本文在若尔盖高原开展了生态系统变化遥感精细表征的研究,获得了若尔盖高原水陆过渡带、稀疏草地区、草地生长过程和水热条件变化的遥感精细表征方法,建立了一套适合高原生态系统变化的监测体系,能够有效监测若尔盖高原面临的生态环境问题,期望为若尔盖高原生态系统的保护和生态恢复提供科学依据。

;

With the interaction of global climate change and human activities, the change of terrestrial ecosystem is becoming more and more prominent. The terrestrial ecosystem present some characteristics such as decreasing in stability of ecosystem and diminishing of ecosystem services ability, which have seriously affected the sustainable development of terrestrial ecosystem and finally threaten the survival and further development of human society. Therefore, obtaining accurate dynamic information and achieving the effective monitor of terrestrial ecosystem is an urgent problem to be solved. In recent years, with the rapid development of remote sensing technology for Earth Observation,it has become an important way to achieve the real-time, consecution, dynamic and efficiency monitoring of the terrestrial ecosystem on a large scale. However, in present research on the method of remote sensing monitoring ecosystem, there are some limitations of single remote sensing data in the temporal and spatial expression. As a consequence, the dynamic characteristics of the terrestrial ecosystem could not be accurately and effectively expressed by the existing methods, which restricted the further development of remote sensing technology.In order to solve this main problem, the Zoige Plateau, which is located in the eastern of the Tibetan Plateau, China, was selected as a study area. The Zoige ecosystem has changed dramatically with interference from global climate change and human activities in recent years. Some severe ecological problems have emerged in the Zoige Plateau, such as marshlands shrinkage, grassland degradation and decrease of water resource. For these problems, this thesis focuses on exploring fine characterization methods with remote sensing technology for alpine wetland-grassland ecosystem in the Zoige Plateau. The specific work starts from researching on ecosystem changes with remote sensing technology on spatial fine characterization and temporal fine characterization. In study of characterization methods of spatial scale, according to the difficulty in monitoring the land-water transition zone and sparse grassland, we will carry out remote sensing sub-pixel scale research of fine characterization of subpixel inundation percentage and grassland cover in Zoige Plateau based on collaboration of the ultra-high spatial resolution unmanned aerial vehicle (UAV) data with mid and high spatial resolution remote sensing data. In study of characterization methods of temporal scale, according to monitoring the growth process of grassland and change of water resources, the phenological monitoring and evapotranspiration estimation method with remote sensing were carried out based on MODIS which has long time series and high frequency observation. Finally, we will study fine characterization with remote sensing on monitor the growth process of grassland and time series of evapotranspiration (ET) which is the key parameter of water resources under the background of climate change. Though the above research work, the conclusions are as follows:(1) In study of fine characterization of land-water transition zone with remote sensing, Sub-pixel inundation percentage (SIP) which is a comprehensive indicator of the wetland ecosystem changes was selected. We will study on the method of fine characterization on wetland SIP mapping based on UAV and Landsat-8 OLI data in this paper. The Results were as follows: Firstly, ultra-high resolution of the UAV image (0.16 m) could capture the information of wetland SIP accurately. Secondly, three typical models (linear spectral decomposition model, artificial neural network model and regression tree model) were applied to modeling wetland SIP mapping based on UAV and Landsat-8 data in 2014. Regression tree models showed the highest precision (R2 = 0.933, RMSE = 8.73%) among three methods according to the validation and comparative analysis. And this method also achieve successfully prediction when it was applied to map wetland SIP of 2013, the prediction accuracy which is high (R2 = 0.986, RMSE = 8.84%) was validated by UAV data of 2013. The results show that the regression tree model which was established in this study could be applied effectively to map wetland SIP in the long time series and to realize fine characterization of wetland SIP with remote sensing.(2) In study of fine characterization of sparse grassland with remote sensing, the grassland coverage which is an important index for monitoring changes of grassland ecosystem was selected in this paper. We will carry out the research of sub-pixel fine characterization methods of grassland coverage based on UAV and Landsat-8 OLI data. The results showed that: Firstly, the pixel dichotomy model was constructed based on the visible difference vegetation index (VDVI) of UAV, which can effectively estimate the subpixel grassland coverage of the visible light image of UAV. Secondly, three categories of typical subpixel coverage estimation models (empirical model, pixel dichotomy model and three-band gradient difference model) were selected for modeling and evaluation based on the 2014 UAV coverage and Landsat-8 OLI data. The results show that the empirical model constructed by this research has the highest estimation accuracy (R2=0.901, RMSE=6.55%), which can be used for the fine mapping of grassland coverage in Zoige Plateau. When this method was applied to monitor the grassland coverage in 2013, the prediction accuracy is high (R2=0.736, RMSE=8.48%) through the validation of UAV data. The results show that the empirical model established in this part of research can effectively monitor the change of grassland coverage in long time series and can be applied for fine characterization of grassland degradation with remote sensing.(3) In study of fine characterization of grassland growth process with remote sensing, the dynamic threshold method was adopted to monitor phenology of Zoige Plateau with remote sensing based on high temporal resolution of MODIS images. The results show that phenological monitoring method used in this study can accurately reflect the growth process of grassland and characteristics of annual dynamic in Zoige Plateau. We can ultimately realize the fine characterization of grassland growth process on the temporal scale with remote sensing. The results show each phenophase characteristics were anastomosis with the water heat conditions. Accompanying the deterioration in heat and water conditions from low altitude to high altitude, south to north and surface water from high to low, SOG (start of growing season) was delayed gradually, EOG(end of growing season) advanced slowly and LOG (long of growing season) shortened gradually. Water played an important role in the regional differentiation of phenology. The inter-annual variation of phenology also reflects the close relationship between it and key climate factors such as temperature, and precipitation.(4) In study of fine characterization of hydrothermal conditions change with remote sensing, ET which is an important parameter for water resources balance was chose. The classical ET model (SEBS) was adopted to study fine characterization of surface evapotranspiration in Zoige Plateau based on the high temporal resolution of MODIS images. Study results show that the remote sensing monitoring method can effectively reflect the distribution characteristics of ET in spatial and temporal scale. We realized the fine characterization with remote sensing on temporal scale for the study area. By monitoring the temporal and spatial distribution of ET in the study area, it was found that the ET is closely related to the surface water and heat condition. Evapotranspiration is highest in the dry hot river valley area in the southwest and eastern part of the Zoige Plateau. ET is higher along the rivers and lakes in the northern of Zoige Plateau and the minimum ET appears in the area of desertification or degraded grassland, and higher elevations mountainous. The monthly maximum and the minimum ET appeared in July and February respectively.In a word, based on fine characterization research of the Zoige Plateau ecosystem change with remote sensing, we obtained the fine characterization methods with remote sensing in land-water transition zone, sparse grassland, the grassland growth process and the change of hydrothermal conditions on the Zoige Plateau, and set up a monitoring system for plateau ecosystem changes which could effectively monitoring the ecological environmental problems faced by the Zoige Plateau. Hope to provide a scientific basis for the protection and restoration of the ecological system in the Zoige Plateau. 

Document Type学位论文
Identifierhttp://ir.imde.ac.cn/handle/131551/18898
Collection数字山地与遥感应用中心
Affiliation中国科学院成都山地灾害与环境研究所
Recommended Citation
GB/T 7714
夏浩铭. 若尔盖高原生态系统变化遥感精细表征研究[D]. 北京. 中国科学院大学,2016.
Files in This Item:
File Name/Size DocType Version Access License
若尔盖高原生态系统变化遥感精细表征研究.(12503KB)学位论文 开放获取CC BY-NC-SAView Application Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[夏浩铭]'s Articles
Baidu academic
Similar articles in Baidu academic
[夏浩铭]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[夏浩铭]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: 若尔盖高原生态系统变化遥感精细表征研究.pdf
Format: Adobe PDF
All comments (0)
No comment.
 

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