Analysis And Forecast For The Temperature Of Unconventional Observations Of Fuling In Chongqing

Fuling Meteorological Bureau started the unconventional observational projects of temperature for real atmospheric environment, cement ground and grass surface from2006. For further developing the unconventional observations of temperature forecast work, better meeting the needs of all walks of life for meteorological service to the actual demand, this article utilized the unconventional observation temperature datas and meteorological datas of2007and2008in fuling, analyzed the datas of cement ground, grass, real atmospheric temperature by statistical methods, colleced a large amount of factors which was in connection with predictand, analyzed the relationship between predictors and predictands, from which a number of meaningful candidate factors were pickde up, then choosed a set of optimal factors by the appropriate statistical methods,builded multivariate regression models. Finally, used the linear regression method to get the extreme temperature prediction equation, and returned to the test with datas of2009, from which finding the equations is of value,and last accorded to the inspection error to adjust the forecasting equation.Through a number of comparative analysis, no matter which kind of observations of temperature, they are closely related with the temperature which is rising and falling. The more the weather is fine, the observation of daily maximum temperature difference is bigger, but the differences of lowest temperature and the minimum temperature are not obvious, but in the precipitation the situation is on the contrary.In the equation of the return test, cement floor spring maximum temperature prediction can be plused1℃to the equation results appropriate. The prediction equation accuracy of the highest temperature of the grass surface is in50%, which is because that the grass surface maximum temperature affected by weather change is bigger, and need more data to adjust the equation. Considering the datas in recent years, the maximum temperature prediction of the grass surface can be adjusted±1～1.5℃according to the weather. Although the lowest temperature forecast accuracy of the grass surface in spring is over50%, but in order to more close to the truth,it can be reducted1～2℃.Because of the frost problem in winter, the minimum temperature prediction of the grass surface in winter can be less reported2℃. Owing to the high accuracy of real atmosphere, the temperature forecast equation is valuable.