Font Size: a A A

Spatio-temporal Variations And Its Driving Factors Of The Grassland Sandy Desertification In The Horqin Sand Land Based On Remote Sensing

Posted on:2015-01-14Degree:DoctorType:Dissertation
Country:ChinaCandidate:J Y LiFull Text:PDF
GTID:1263330431463388Subject:Agricultural remote sensing
Abstract/Summary:PDF Full Text Request
Desertification, land degradation in arid, semi-arid and dry sub-humid areas resulting from climaticvariations and/or human activities, is treated as one of the most critical environmental hazards, since theincreasing rate of it implies a clear manifestation of human interaction and climatic-change processes onthe environment. As one of the seriously affected countries with a long-term and large-scaledesertification problem, China is not immune to this environmental hazard. Since the direct economicloss of desertification in China is about128.14billion RMB each year in the late20th century,desertification is often termed as the main and most serious form of the degradation. With a populationof over1.3billion and limited farmland, it is a survival issue for China to tap and manage the grasslandefficiently and sustainable, since it has the second-largest grassland area of about400million hectaresof which accounts for41.41%of the country’s territory. Facing the worsening degradation of grasslands,state and local governments in China have implemented a series of ecological protection projects, theeffect of which has become contentious at all levels of government and a topic of public concern.Grassland desertification is the most direct indicator of the effect of the grazing ban. Monitoring andaccurate assessment of the status and changing trend of grassland desertification is instrumental indeveloping effective policies to combat grassland degradation, for understanding the changing trend ofgrassland degradation.We selected18counties and cities in Horqin Sandy Land as the study area. Based on a series ofLandsat TM/ETM+/OLI images, field observations and expert review, a Horqin grassland sandydesertification classification and grading system was constructed. Then, spectral mixture analysis (SMA)and a decision-tree method were used to interpret images of the study area from four years:1985,1992,2001and2013. The following results were obtained:1.Based on relevant desertification classification indicator studies, China’s national standard andfield inspection and verification, combined with the land use, grassland type and basic desertificationcharacteristics of the study area, we classified the land cover of the study area into nine categories.Finally, we conducted seminars and demonstration projects for experts to substantiate the scientificvalidity, operability and interpretability of the remote sensing images. Combining the sample plots andquadrat data, the grassland type and general environment of the study area, we further determined theappropriate gradation system for remote sensing of grassland desertification in Ningxia2.Before producing unmixed pixels, a mask was made for farmland, human settlements, water,forest and other land uses based on visual interpretation, the MNDWI, NDVI, TC and the NDII, amongother methods.3.Comparison and selection of methods: Linear spectral mixture model retrivals the occupation ofbare sand, vegetation index retrivals the vegetation coverage. We find that the desertification level inhigh soil coverage area often be overrated while be evaluated by vegetation coverage. However,evaluation based on bare sand area can effectively avoid this kind of problem with high interpretation accuracy.4.Spatial distribution characteristics:The moderately desertified grassland and severely desertifiedgrassland mainly distributed in the southwest of Horqin Sandy Land, where involves Wengniute Banner,Naiman Banner, Kulun Banner and Aohan Banner in administrative division. And the severelydesertified grassland especially located in the tringle region that forms by Wufendi-Wulan in WengniuteBanner, Xar Moron River and Jiaolai River.5.Area statistical characteristics:The landcover in the study area is mainly covered by grassland,which accounts for70%of its territory. Among the grassland, non-, slightly, moderately and severelydesertified grassland accounts for55%,9%,4%and3%of its territory, separately.6.Spatio-temporal variation characteristics: From1985to1992, the grassland desertification degree“reversed” and “obviously reversed” mainly distributed in the central part of Bairin Left Banner and thenorthern part of Bairin Right Banner, while “developed” and “seriously developed” mainly distributedin the northeast part of Wengniute Banner, northwest part of Naiman Banner and west of Kailu County.From1992to2001,“reversed” and “obviously reversed” mainly interspersed among Naiman Baner, thenorthern part of Aohan Banner and the eastern part of Ar Horqin Banner, while “developed” and“seriously developed” mainly distributed in the middle-east part of Wengniute Banner, northeast part ofKulun Banner et al. From2001to2013, the grassland sandy desertification in Horqin Sandy Land wasmainly “reversed” and “obviously reversed”, except some “developed” and “seriously developed” inKeerqinzuoyihou Banner,central part of Kulun Banner and Bairin Right Banner.7.Driving factors: During the research period in this paper, the warm desiccation tendency wasquite obvious. The coupling relation between wind and precipitation not benefit the reversion of thegrassland sandy desertification. Furthermore, the growing number of population, livestock and farmlandarea also blocked the reversion trend of grassland sandy desertification. However, the results in thispaper suggested that the grassland sandy desertification turned in a reversion trend since1992, andespecially clear reversed between2001and2013. These results show that the grazing ban, together withother ecological engineering measures, has helped reverse desertification and promote the restoration ofgrassland vegetation.
Keywords/Search Tags:Horqin, grassland sandy desertification, LSMM, spectral mixture analysis, driving forceanalysis
PDF Full Text Request
Related items