ANNUAL VARIATIONS OF TEMPERATURE ON FOUR URBAN UNDERLYING SURFACES AND FITTING ANALYSIS

Based on the observed 2-year temperature data for four kinds of typical urban underlying surfaces,including asphalt, cement, bare land and grass land, the annual variations and influencing factors of landsurface temperature are analyzed. Then fitting equations for surface temperature are established. It is shownthat the annual variation of daily average, maximum and minimum temperature and daily temperature rangeon the four urban underlying surfaces is consistent with the change in air temperature. The difference oftemperature on different underlying surfaces in the summer half year (May to October) is much moreevident than that in the winter half year (December to the following April). The daily average and maximumtemperatures of asphalt, cement, bare land and grass land are higher than air temperature due to theatmospheric heating in the daytime, with that of asphalt being the highest, followed in turn by cement, bareland and grass land. Moreover, the daily average, maximum and minimum temperature on the four urbanunderlying surfaces are strongly impacted by total cloud amount, daily average relative humidity andsunshine hours. The land surface can be cooled (warmed) by increased total cloud amount (relativehumidity). The changes in temperature on bare land and grass land are influenced by both the total cloudamount and the daily average relative humidity. The temperature parameters of the four land surfaces aresignificantly correlated with daily average, maximum and minimum temperature, sunshine hours, dailyaverage relative humidity and total cloud amount, respectively. The analysis also indicates that the range offitting parameter of a linear regression equation between the surface temperature of the four kinds of typicalland surface and the air temperature is from 0.809 to 0.971, passing the F-test with a confidence level of 0.99.