Space Model Of Precipitation And Temperature Factors Of Advanced High-resolution In Different Region In Gansu Province

The results of ecology research has showed that meteorological factors not only determine the space distribution, structure and function of the meadow resource type,but also is basic data resource of scientific research for Ecology, Geography, Agronomy and the whole world changing, especially play very important role in studying on simulation and management of regional and global ecosystem changing. According to Ren JiZhou's comprehensive order classification of grassland, grassland of different regions in the world can be classificed 48 types from local precipitation, accumulated temperature and the both interaction, however, the wide application of the comprehensive order law of grassland would be limited for the meteorological website is dispersed and discontinuous in fact. So it is necessary to treatment of spacing for dispersed statistics.Based on 2003-2004 year meteorological observation data, digital elevation model (DEM), MODIS remote sensing NDVI and EVI data of image, the relationship of precipitation and temperature factor with geographical space position, elevation, sunshine time and vegetation index (NDVI, EVI, i.e. the lower cushion character) was analyzed by utilizing Geographical information system (GIS), remote sensing (RS), mathematics statistical method and Kriging space interpolating value technology. The key factor of influence on precipitation and temperature was found by corelations and multivariate linear regressions analysis method, the space regression model of rectified average annuals precipitation, average annuals temperatures and >℃ accumulated temperatures in Gansu Province, Gannan-Qilian plateau region, southern mountain area of Gansu, Loess Plateau district, Hexi Corridor district based on second error and topographic factor(gradient, direction of slope)was finished. In paper, on one hand, sunshine time was considered in the space modeling for the role in influencing precipitation and temperatures.On the other hand, the data of influencing remote sensing were introduced in space research, quantitative expression of the remote sensing data (vegetation index) for lower cushion was used in the space model.In research, 500m X 500m high-resolution space database (GRID) of average annuals precipitation, average annuals temperatures and >℃ accumulated temperatures was produced, and space technological route of precipitation and quantity of temperature was formed by utilizing the method above.The results showed,1. the vital factor in influencing precipitation varied in different character of time and space. In Gansu Provinces, the key factor of influencing precipitation was latitude and altitude in turn, the key factor of influencing temperature was altitude, latitude and sunshine time in sequence. In Gannan-Qilian plateau district, the key factor of influencing precipitation was vegetation index, longitude and altitude in turn, the key factor ofinfluencing temperature was altitude, latitude in sequence. In south mountains of Gansu, there were not significantly influencing factors for precipitation, altitude and latitude were key factor for affecting temperature. In losses plateau region, the key factor of influencing precipitation was vegetation index, longitude and latitude in turn;the key factor of influencing temperature was altitude, latitude. In Hexi corridor district, the key factor of influencing precipitation was longitude, altitude and latitude;the key factor of influencing temperature was altitude, sunshine time and longitude.2. the sunshine time was significantly correlation with precipitation.3. the altitude played key role in determining quantity of average annals temperature and > *Caccumulated temperatures in Gansu.4. the accurate degree of spacing was increased by rectify of second error, Square of the correlation coefficient between measure value and predicted value in precipitation average annals temperature > °C accumulated temperatures was 0.9999,0.9999,0.93 respectively;average absolute error was 1.197mm,0.443°C,203.791°C respectively;the maximum error was -0.107mm,0.001 °C ,40.176 °C respectively;RMSE was 1.685mm,0.014 'C ,289.651 °C respectively;the minimum error was 0.051mm,0.000 °C ,6.806 °C respectively.5. the spacing results by multivariate linear regressions analysis method can well reflect character and rule of space distribution of precipitation and temperature.