Monitoring Vegetation Dynamics in Xilinguole League, Inner Mongolia, China by Using Spot/vegetation Data

A new method for grassland vegetation dynamics monitoring was developed based on the linear time trends (LTT) estimation and change vector analysis (CVA) using SPOT-VGT imagery. Firstly, seasonally integrated normalized difference vegetation index (SINDVI) was calculated by using 10-day NDVI. Then by evaluating spatially averaged SINDVI of vegetated areas, the apparent change periods were identified. Thirdly, the areas where SINDVI changed significantly were extracted during the apparent periods. And finally, the magnitude of NDVI change in the above regions was calculated with CVA. The method was used in the grassland dynamics detection in Xilinguole League (Inner Mongolia, China) to verify its performance. The results indicated that the new method is effective for long-term vegetation dynamics monitoring. There are two periods with obvious decrease of NDVI: 1998-2001 and 2003-2007. From 1998 to 2001, the area where SINDVI showed significant decrease trend (ASSDT) accounts for 82.29% of the whole vegetated area; and in this area obviously decreased region was mainly located in the middle part of the study area. From 2003 to 2007, ASSDT accounts for 48.73% of the whole vegetated area; and the obviously decreased region was mainly located in the northeast of the study area.