[Dynamic assessment of grassland degradation in Naqu of northern Tibet].

Based on the past years vegetation cover, annual maximal grass yield and June-September mean modified soil-adjusted vegetation index (MSAVI) that were inversely deduced from the 10-day composite data of NOAA/AVHRR channels 1 and 2 and NDVI in 1982-2000, this paper analyzed the recent 20 years dynamics of grassland desertification in Naqu of northern Tibet. The results showed that in recent 20 years, the area of degraded grassland in Naqu was averagely 43.1% of the total land area. It was decreased in the former ten years while increased in the latter ten years, but overall, had a decreasing trend. The degraded area was larger in west part of Naqu than in its other regions. Among the eight climatic factors including temperature, precipitation, potential evapotranspiration, vapor pressure, wind velocity, sunshine hour, ratio of precipitation to evapotranspiration, and ratio of temperature to precipitation, the most remarkable factor affecting the dynamic of grassland degradation was the potential evapotranspiration.

[Classification and dynamic changes of grasslands in northern Tibet based on recent 20 years satellite data].

Based on the NOAA/AVHRR 10-day composite NDVI data from 1982 to 2000 and by the methods of principal component analysis and unsupervised classification, the vegetations in Naqu District of northern Tibet were classified, and the annual and inter-annual variations of NDVI on the pixels selected from different grassland types were analyzed. The regions where the mean NDVI values during the main growth season of grass were equal to or great than 0.1 were defined as vegetation regions, while those where the mean NDVI values were less than 0.1 were defined as non-vegetation regions. The spatiotemporal changes of NDVI in each pixel in vegetation regions were analyzed, and the results showed that the grasslands in northern Tibet could be classified into four types, i. e., high-cold meadow, high-cold meadow steppe, high-cold steppe, and high-cold desert. This classification was accorded with practical status. The annual changes of the NDVI in the four grassland types showed one peak value, with the maximum in August. In recent 20 years, the mean NDVI from July to August in vegetation regions decreased gradually from southeast to northwest, ranging from 0.1 to 0.6, and the variation coefficient changed between 0.05 and 0.40, which was smaller in the place of high NDVI than in the place of low NDVI. The annual variability of NDVI changed between -0.005 and 0.008. In recent 20 years, the vegetations had no observable change, but the vegetation development in 20% of vegetation regions, which mainly located in Nima in western Naqu and in Jiali, Biru, Suoxian and Baqing in eastern Naqu, was weakened.

[Regional simulation of rice yield change under two emission scenarios of greenhouse gases].

Based on the newest emission scenarios of SO2 and greenhouse gases, i. e., the Intergovernmental Panel on Climate Change (IPCC) Special Report on Emission Scenarios (SRES) A2 and B2 scenarios, and by using RCM (Regional Climate Model)-PRECIS and CERES-Rice model, this paper simulated the rice yield change in 2080 at 50 x 50 km scale. The results showed that there was a great range of yield change across whole China. The yield would increase along the Changjiang River and in South China, and decrease in North and Northeast China. Because of the direct effect of CO2 on rice growth, the SRES A2 scenario would be more positive to the increase of rice yield than B2. In 2080, the total rice yield in whole China would increase under A2 emission scenario, while decrease under B2 emission scenario.