A GIS-based Approach for Estimating Spatial Distribution of Seasonal Temperature in Zhejiang Province, China

Li Jun,Huang Jing-feng,Wang Xiu-zhen

DOI: https://doi.org/10.1631/jzus.2006.a0647

2006-01-01

Abstract:This paper presents a Zhejiang Province southeastern China seasonal temperature model based on GIS techniques. Terrain variables derived from the 1 km resolution DEM are used as predictors of seasonal temperature, using a regression-based approach. Variables used for modelling include: longitude, latitude, elevation, distance from the nearest coast, direction to the nearest coast, slope, aspect, and the ratio of land to sea within given radii. Seasonal temperature data, for the observation period 1971 to 2000, were obtained from 59 meteorological stations. Temperature data from 52 meteorological stations were used to construct the regression model. Data from the other 7 stations were retained for model validation. Seasonal temperature surfaces were constructed using the regression equations, and refined by kriging the residuals from the regression model and subtracting the result from the predicted surface. Latitude, elevation and distance from the sea are found to be the most important predictors of local seasonal temperature. Validation determined that regression plus kriging predicts seasonal temperature with a coefficient of determination ( R 2 ), between the estimated and observed values, of 0.757 (autumn) and 0.935 (winter). A simple regression model without kriging yields less accurate results in all seasons except for the autumn temperature.