Estimation of grassland condition in mongolia from field survey and satellite images

Mongolian grassland is faced with the problem of degradation. We should understand grassland condition which is seemed to be affected from land-use by using remote sensing. We couldn't classify grassland by condition whether coarse or dense vegetation from the visual image. So we chose the Normalized Difference Vegetation Index (NDVI) which is the most popular vegetation index as to understand grassland condition. We calculated NDVI value on the LANDSAT/ETM+ image. We could find the heavily degradation area on image, where NDVI values are low. We confirmed that it was sand roads over grassland by field survey. Then we selected 5 sites considering land-use type. LT1 is meadow in settled area, LT2 is abundant cropland, LT3 is grassland in tourist camp, LT4 is meadow in nomadism area, and C is conservation grassland in national park selected as control. We conducted the field survey to investigate grassland condition such as vegetation coverage, height of plant and species at 5 sites. The result of field survey showed that grassland condition was related to land-use type. We compare the NDVI values to vegetation information at 5sites we selected. We couldn't show the relationship between NDVI and vegetation information. NDVI can be useful for grasping trend of seriously degraded grassland distribution but is inadequate for more detailed analysis of grassland condition. We should try another approach to estimate grassland condition.