Dynamics of natural vegetation on the T ibetan P lateau from past to future using a comprehensive and sequential classification system and remote sensing data

Long‐term series of vegetation change is one of the key study contents for evaluating terrestrial ecosystems and plays an important role in global change study. The CSCS ( C omprehensive and S equential C lassification S ystem) model was used in the paper to analyze the change of P otential N atural V egetation ( PNV ) types during the period 1951–2000 and for the three future years 2020, 2050 and 2080. The growth conditions for each PNV type were analyzed based on MODIS Normalized Difference Vegetation Index ( NDVI ) datasets on the T ibetan P lateau ( TP ) from 2001 to 2010. The results showed that: (i) The mean annual precipitation ( MAP ) and mean annual temperature ( MAT ) had increasing trends (1.72 mm year −1 and 0.025°C year −1 , respectively) over the past 50 years from 1951 to 2000. (ii) Based on the three general circulation models ( CCCMA , CSRIO and Had CM 3), the MAT and MAP on TP would increase in the future, with magnitudes 3.54 to 6.08°C and 28.71 mm to 126.84 mm from 2000 to 2080, respectively; it is also predicted that two new vegetation categories: warm desert and savanna would appear in the future and that the area of forest would increase significantly, while the area of grassland would have an obvious decreasing trend. And (iii) Overall, the ecological environment of vegetation in the recent 10 years has been improved in most of the areas in central and southern regions of Q inghai P rovince, but degraded in some areas of S ichuan P rovince, and the southwest and south of T ibet.