摘要
[ Objectivel The research aimed to study prediction model for spatial distribution of the average temperature based on GIS. [ Method l Average temperature over the years as research object, based on Ordinary Kriging （OK）, Inverse Distance Weight （ IDW）, SPLINE and Mixed In- terpolation （MLR）, monthly temperature data from 1979 to 2008 at 18 long-term meteorological observation stations in Hainan Island were conduc- ted spatial grid treatment. Via contrasts and analyses on different interpolation methods, the optimum interpolation method for average temperature over the years in Hainan Island was selected. [ Resuitl By error analyses of the four interpolation methods for average temperature in recent 30 years in Hainan Island, it was found that accuracy was MLR 〉 IDW 〉 OK 〉 SPLINE. Spatial interpolation effect of MLR was the best for average temperature in Hainan Island. Spatial distribution of the average temperature in Halnan Island had obvious south-high-north-low latitudinal zonality and vertical zonality of gradually declining as altitude rise. In addition, temperature along coast was slightly higher than that in inland. Lapse rate of the temperature in each month in Hainan Island was 0.38 -0.85℃/100 m, and lapse rate of the annual average temperature was about 0.74 ℃/ 100 m. In different areas, lapse rate of the temperature as altitude was different at different time. [ Condusion] The research provided basis for ob- taining continuous distribution situation of the agricultural meteorological factor and establishing accurate prediction model of the spatial distribution in Hainan Island.
[ Objectivel The research aimed to study prediction model for spatial distribution of the average temperature based on GIS. [ Method l Average temperature over the years as research object, based on Ordinary Kriging （OK）, Inverse Distance Weight （ IDW）, SPLINE and Mixed In- terpolation （MLR）, monthly temperature data from 1979 to 2008 at 18 long-term meteorological observation stations in Hainan Island were conduc- ted spatial grid treatment. Via contrasts and analyses on different interpolation methods, the optimum interpolation method for average temperature over the years in Hainan Island was selected. [ Resuitl By error analyses of the four interpolation methods for average temperature in recent 30 years in Hainan Island, it was found that accuracy was MLR 〉 IDW 〉 OK 〉 SPLINE. Spatial interpolation effect of MLR was the best for average temperature in Hainan Island. Spatial distribution of the average temperature in Halnan Island had obvious south-high-north-low latitudinal zonality and vertical zonality of gradually declining as altitude rise. In addition, temperature along coast was slightly higher than that in inland. Lapse rate of the temperature in each month in Hainan Island was 0.38 -0.85℃/100 m, and lapse rate of the annual average temperature was about 0.74 ℃/ 100 m. In different areas, lapse rate of the temperature as altitude was different at different time. [ Condusion] The research provided basis for ob- taining continuous distribution situation of the agricultural meteorological factor and establishing accurate prediction model of the spatial distribution in Hainan Island.
基金
Supported by "Project 211" Construction Item,Hainan University
Basic Science Research Business Expense,Rubber Research Institute ,CATAS[YWFZX09-03(N)]
Special Item of the Modern Agricultural Industrial Technology System Construction(CARS-34)
