Analysis On Vegetation Recovery In The Lower Reaches Of Tarim River Based On Low Resolution Remote Sensing Images

Taking field survey data, several satellite images in 2000-2006 as well as monitoring data of surface water level and underground water level after water transportation as data source, taking recovery ecosystem and landscape ecosystem as theory basis, taking main vegetation distribution regions(Yingsu-Alagan) where water was transported in the lower reaches of Tarim river as studied objects, applying mathematics analysis, "3S" technology and field work investigation, this paper exploited relationship between water transportation and vegetation ecology reaction. The main results showed:The MODIS-NDVI in the growing period from 2000a-2006a showed that vegetation in the monitoring sections (Yingsu, Kaerdayi, Alagan) recovered evidently though the recovery speed was different and the recovery extent in 2005a was evidently from view of the whole research region; The vegetation's growing kurtosis period was July and vigorous period was from June to August; MODIS-NDVI predictive model was established according to MODIS-NDVI change law after water transportation and its accuracy could be reached to 94.4%; The accuracy of feedback to this model by MODIS-NDVI vegetation coverage could be reached 82.6%.Analysis on the ETM/ASTER-NDVI showed that the differences between NDVI of monitoring sections were little before water transportation but extremely obvious after water transportation. The vegetation recovery speed and recovery level showed decreasing trend with water transportation distance increasing. The accuracy of the ETM/ASTER-NDVI model built upon the prediction mentioned was 80.11% and the accuracy of feedback to vegetation coverage by ETM/ASTER-NDVI could be reached 92.4%. The landscape indexes showed that the area vegetation covered expanded fast and took on a union trend. The sand spot increased in amount but decreased in area collective level decreased evidently. In the research area, spot amount decreased, spot density and spot area average increased, landscape spot shatter level decreased, extend index decreased with the process of time. The water transportation had inhibited the desertization in the research area.Analyzing with the Markov Model and showing that:under the continual water supply to the reseacher areas,the vegetions in these areas will transfer to a good direction in the next 30 years;Analyzing the vegetiong cover pattern, it can predict the field class and vegetion cover change with Markov Model ,adjust the interfer factors upon the imitation results and make the fields in different classes change to non-malignant direction.Analyzing the NDVI relativity,MODIS-NDVI and ETM/ASTER-NDVI have a harmonious correlation;NDVI has a negative correlation trend with undergroung water depth;basing on the ETM/ASTER-NDVI interval in the main field classes,it can locate the NDVI range of the main field classes.The vegetiong recovery level shows that: the vegetion recovery level in Yingsu profile is 21.9% after 7 years since the water supply project compared with the same distance to the river in Daxihaizi in NDVI which is a non water interception section and Kaerdayi and Alagan follow with 15.7% and 11.6%,and the integrated vegetion recovery is 16.4%.Through the study of vegetion recovery with MODIS and ETM/ASTER, the change felected by both of them is the same trend, but both os them have their own advantage.