Interpolation optimization of climatic factors and its correlation analysis withgrassland NPP under regional scale

With the climatic data from 42 stations and measured data of grassland aboveground biomass from 154 sites located in and around Gansu Province, the spatial distribution of annual average and seasonal temperature, annual and seasonal precipitation from 1996 to 2005 was simulated, and its spatial distribution pattern and correlation with grassland NPP were analyzed based on improved AMMRR（Analytic Method based on Multiple Regression and Residues） interpolation method (considering slope and aspect). The results showed that：1) comparing with other interpolation methods, improved AMMRR method could completely embody the microcosmic variation of annual average temperature and annual precipitation response to the effect of topography, which was more suitable to simulate realistic situation. And 2) there was significant negative correlation between grassland NPP and temperature in each period except winter, but opposite to precipitation. Moreover, the correlation between grassland NPP and precipitation was higher than temperature.