Font Size: a A A

Changes Of Vegetation Cover And The Comparison Of Different Evaluation Methods On The Loess Plateau

Posted on:2021-02-06Degree:MasterType:Thesis
Country:ChinaCandidate:Z HeiFull Text:PDF
GTID:2370330626964379Subject:Environmental engineering
Abstract/Summary:PDF Full Text Request
Vegetation is one of the most important components in the terrestrial ecosystem,and is of great significance for improving the fragile ecological environment of the Loess Plateau.A systematic study of the characteristics of vegetation cover changes at the temporal and spatial scales of the Loess Plateau after returning farmland to forest?grass?will help to understand the effects of ecological engineering and will not provide a scientific basis for subsequent ecological restoration measures.Based on GIMMS NDVI and MODIS NDVI data,we use numerical comparison,change trend analysis and correlation analysis to study the characteristics of vegetation coverage change in different regions of the Loess Plateau from different spatial and temporal scales.The characteristics of cultivated land change in the Loess Plateau from 2000 to 2019 were analyzed based on the NDVI time-series of the main surface cover.We also compared the differences in vegetation parameter characteristics between 2 different evaluation methods in different regions of the Loess Plateau,the main results are presented as the following:?1?It is feasible to form long-term vegetation data for the Loess Plateau with MODIS NDVI data supplemented by GIMMS NDVI data for extension of its time series.GIMMS NDVI and MODIS NDVI have a strong correlation in vegetation spatial distribution and dynamic growth trend change trends.The regression coefficients of the two reaches 0.95 at the regional scale,and the proportion of the area of the area where the two were positively correlated at the pixel scale reached 87.69.The two types of datasets have good consistency in reflecting the changes of vegetation phenology.?2?The vegetation NDVI of the Loess Plateau for many years showed a spatial distribution of low northwest and high southeast.The vegetation NDVI on the time scale increased significantly from 1982 to 2019.The vegetation NDVI increase rate was 0.001 yr-1 from 1982 to 1999.There was no significant change in vegetation coverage in more than half of the area.The proportion of the area that increased significantly was about 14.04%?P<0.05?.The rate of vegetation NDVI increase from2000 to 2019 was 0.006yr-1,and the vegetation coverage in more than 90%of the regions showed an increased trend,the area ratio that increased significantly 71.56%?P<0.05?.The change trend of vegetation NDVI increased exponentially becoming faster after 2000.?3?By analyzing the characteristics of the NDVI time-series curve of the main surface cover in the study area,extracted the pixel distribution of cultivated land in the Loess Plateau area from 2000 to 2019,and the overall sampling accuracy reached85.55%,which proved that the method was effective and feasible,and the accuracy error of the valley plain area was less than 10%.The average distribution of cultivated land in the study area has obvious regional differences over the years,with more and denser cultivated land in the southeast,less and scattered cultivated land in the northwest,and a small fluctuation of cultivated land area in the region in the past 20years,the reduction of cultivated land is mainly induced by ecological engineering and the urbanization area.?4?The NDVI of vegetation on the Loess Plateau improved significantly from 2000to 2019?P<0.01?.In the study of vegetation restoration process,it was found that the vegetation index obtained by using the same data through different processing methods had different results in the evaluation of vegetation restoration process.The mean growth season average NDVI?M2?from the monthly maximum synthesis method is0.446,and the mean growth season average NDVI?M1?generated form the monthly average is 0.352.The increasing trend of vegetation NDVI for M2 and M1 is 0.007yr-1and 0.0057yr-1,respectively.The proportion of the area with increasing trend is assessed by M2 is larger than by M1.In the areas where the NDVI increasing with a trend from0 to 0.01,the increasing area judge by M1 and M2 accounting for 83.97%and 74.08%,respectively;in the area with a 0.01-0.02 if increasing trend,and the areas assessed by M1 and M2 accounting for 11.8%and 21.42%,respectively.The areas based on M1and M2 accounted for 0.06%and 0.44%in areas where the increasing trend within0.02-0.03,respectively.In the area where the increasing trend range greater than 0.03,the difference between 2 methods was very small.?5?The degree of vegetation restoration in the Loess Plateau of different regions is different.The growth trend of vegetation from high to low is:Loess Plateau region,Loess Plateau hilly region,Loess Plateau region excluding cropland,Loess soil excluding cropland Plateau hilly mountain area.In addition,after excluding cropland pixels,the rate of vegetation increase slowed,and the degree of vegetation restoration decreased.
Keywords/Search Tags:the Loess Plateau, vegetation change, remote sensing, NDVI, the Grain for Green Project, MVC
PDF Full Text Request
Related items