| As an important social and environmental problem disturbing the world today, the land desertificationis threatening human existence and development. Vegetation coverage and land utilization are twoimportant factors that impact the desertification, and easy to be observed. Mu Us Sand land is one of thefour major sand land in our country, and at the junction of southeast of Ordos Highlands, north of LoessPlateau in Shanxi and northeast of Ningxia, it is located at the transition belt of desert steppe-steppe-foreststeppe, most of this area belong to temperate zone and the west-frontiers of our country’s monsoon region,environment of which is very sensitive and week, it has been the typical area of sand land in our country.Since the20th century, the situation of land desertification is overall reserved and which is particularlytypical in Mu Us Sand land.Mu Us Sand land is selected as the research area in this paper. Firstly, the variation trend of vegetationcoverage is analyzed by calculating respectively MODIS-NDVI data of vegetation coverage from2000to2010using dimidiate pixel model. Secondly, TM/ETM+data of the three period that is2000,2005and2010are classified by using object-oriented classification technology to investigate the change of land-usetype in the area. Finally, spatial patterns of both the change of vegetation coverage and that of land-use typeare analyzed. This study can provide decision support for ecological protection and environmentalimprovement of Mu Us Sand land in the future. This study finds that:(1) The change characteristics of spatial patterns of vegetation coverage grade in Mu Us Sand land:the vegetation with low coverage grade decreases, while that with high coverage grade increases, thevegetation with medium coverage grade sometimes increase and sometimes decrease, but which decreaseoverall. The overall trend of vegetation coverage is that the vegetation with medium and low coveragegrade decrease, while the vegetation with high coverage grade increase gradually. The spatial distributionof vegetation coverage of Mu Us Sand land is that the vegetation with high coverage grade is mainlydistributed in most of the area of Ejin Horo Banner, Shenmu county and Jingbian county, the vegetationwith medium coverage grade account for a large proportion and is widely distributed, and that with lowcoverage grade is mainly scatter in Etuokeqi, Etuokeqianqi and Wushenqi. From the ratio of pixel, the ratioof the vegetation with high coverage grade is34.7%, that of the vegetation with medium coverage grade is34.7%and that of vegetation with low coverage grade is5.99%in2010.(2) The change characteristics of land use type: artificial surface increases290.47km2in2000-2010,which is mainly transformed from grassland, sand land, and farmland. As the policy of returning farmlandto forests, farmland decrease58.21km2while woodland increase30.28km2. During2000-2010, grassland,sand land and wetland are decreasing year by year and sand land reduce102.37km2. Despite the reduction is little, work of preventing and controlling desertification has played a role in making the sandy areadecrease gradually. The situation of land use of Mu Us Sand land in2010is that the area of grassland is30832.55km2,64.23%of the total area, sandy area is8447.05km2,17.6%of the total area, the area ofwoodland is4374.39km2,9.11%of the total area, the area of farmland is3760.82km2,7.84%of the totalarea, the area of artificial surface is617.62km2,1.29%of the total area, the area of wetland is257.29km2,0.54%of the total area.(3) The vegetation coverage change and land use change of integration analysis: usingARCGIS and subtract raster data in the layer superposition technology, analyses on the spatiallayout. Changes in2000-2005area mainly concentrated in the northwest and northeast;Changes in2005-2010area mainly concentrated in the northeast, east and central regions.In addition, land use map of2010can be obtained by using object-oriented classification techniques, andthat of2000and2005can also be obtained with similar principle. This technique is a less time consumingand high-precision method. Through field verification, the classification accuracy is more than85%. |
PDF Full Text Request