Changes And Driving Factors Analysis Of Grasslands In Nagqu Area

The grassland area of Tibet Autonomous Region is the largest one in China,which accounts for 20.5% of the country's total grassland area,in which the Nagqu grassland area takes up 41.7% in proportion of Tibet Plateau.So the function roles of the Nagqu grassland are not only the pillar of animal husbandry in Tibet economic development,but also play an important role in maintaining the ecological balance and climate change of the Tibetan Plateau.In this paper,the remote sensing,GIS(Gepgraphic Information System)and the statistical software were used to analyze the spatial and temporal characteristics of the Nagqu grassland NDVI(Normalized Difference Vegetation Index)in recent 34 years.Based on NDVI vegetation index data of GIMMS(Global Inventory Modelling and Mapping Studies)from 1982 to 2006,and MODIS(Moderate-resolution Imaging Spectroradionmeter)NDVI data from 1982 to 2014,the driving factors of grassland changes were studied,and the results were showed as follow:The data dependency between the GIMMS NDVI and MODIS NDVI was well correlated,and so both of them could be used to analyze the dynamics of grassland vegetation and climate change response in further.The vegetation types in Nagqu area can be divided into shrub grassland,alpine desert grassland,alpine vegetation area and alpine meadow steppe.In which the shrub grassland were distributed in Jiali,Biru,Suo and Baqing county areas;Nima,Shuanghu,Anduo and Bange counties were mainly alpine desert grassland;alpine vegetation area were relatively scattered;Alpine meadow steppe was mainly located in the north of Nagqu,the boundary area with Xinjiang,and some counties in the southeastern area.The rest areas were non-vegetation lakes,glaciers and bare land.The NDVI of Nagqu area showed a fluctuating trend in recent 34 years,which got to the maximum value of 0.2542 in 1981,and the minimum value of 0.2106 in 1985.The NDVI average in these years was 0.2325,and the standard deviation was 0.0106.The change trend of NDVI in annual scale was not significant.But the inter-decadal variation of vegetation was presented as a rising tendency.The vegetation area was shown stable from 1990 to 1997,but degraded seriously from 2002 to 2007.After 5 years of grassland ecological compensation policy,the vegetation had recovered,which reflected the effectiveness of the policy implementation.The influencing factors on the grassland growth and development were analyzed here and divided into two parts: the nature and human factors.The nature factors included the temperature,precipitation,permafrost,prairie rodents,melting glaciers,lake area changes,and so on.The human factors were mainly refers to over-grazing,uncontrolled development and other human factors.The NDVI were positively correlated with the average temperature,the two-years cumulative precipitation,the regional livestock production and the livestock annual slaughter rate,and the correlation coefficients were 0.4,0.56,0.78 and 0.7,respectively.On the other hand,the NDVI was negatively correlated with the number of livestock at the end of the year,and the correlation coefficient was-0.26.The principal component analysis results showed that the driving factors of grassland change were: water factor,livestock factor and thermodynamic factor.The BP(back propagation)neural network was used to predict the NDVI of pasture in the past 5 years.It was concluded that the prediction result based on principal component was better than all the impact factors.The correlation coefficient was improved by 0.31 and the prediction effect satisfied the actual situation better.