The Grid Prediction Of Daily Minimum Temperature In Real Geographical Environment

Under the background of global climate warming,extreme meteorological disasters,especially the frost had serious influence on crop growth and development.The occurrence of disasters are fleeting so that the previous study of the distributed simulation for minimum temperature which is locaus on month or longer time scale cannot meet the needs of disaster risk zoning and so on.Therefore,it is of great significance to explore the grid method of daily minimum temperature.A kind of spatial method for daily minimum temperature is provided in this paper.which is decomposed the daily minimum temperature in real geographical environment into the ground term(the daily minimum temperature on the horizontal plane)and the mountain impact term according to the menthod of extrapolation and the idea of separation and combination.The main results are as follows:(1)Considering the three elements affecting temperaturefully:the global radiation,altitude and long wave radiation,we realized the distributed simulation of the monthly mean global radiation and minimum temperature on the horizontal plane by establishing the empirical relationship model and the statistical model of physical experience.Based on this,the spatial distribution of the daily minimum temperature on the horizontal plane is calculated.At the same time,the error analysis of results is also made in this paper.The results show that the fitting error of simulated sites for global radiation on the horizontal plane is small.The average NRMSE and MABE of each month are only 8.707%and 0.847 MJ m-2.The error value of different seasons is:summer>spring or autumn>winter.The averavg MABE of 12 months of simulated sites for the month minimum temperature on the horizontal plane are only 0.403?.The averavg MABE of the minimum temperature on the horizontal plane on the 15th day of each month in validation sites is 1.051?,and the error is bigger than monthly results,however,both of them shows the smallest error is in summer,followed by spring and autumn,and the largest in winter.(2)Considering the effect of sky factors,atmospheric conditions and terrain factors(slope,aspect)on radiation and temperature comprehensivly,we realized the distributed simulation of the monthly mean global radiation and minimum temperature on the complex terrain,and carried out the error verfication.The results show that the simulation results on the complex terrain and horizontal plane have similar macroscopic distribution,but the results on the complex terrain can reflect the influence of local topographythat is the radiation and temperature in south slope is always higher than north slope.The average NRMSE and MABE of each month for global radiation on the complex terrain are 8.823%and 0.878 MJ m-2 which are slightly bigger than horizontal plane.The largest error is still in summer,the smallest in winter.The validation error of monthly mean minimum temperature on the complex terrain displayes the average MABE of each month is only 0.807?,and the simulation effect is good.The performance in summer is better than spring and autumn,obviously better than winter.(3)Based on the idea of separation and combination,we realized the grid calculation of the daily minimum temperature on the complex terrain combining with the daily minimum temperature on the horizontal plane and mountain impact items.The results can reflect the macroscopic distribution of temperature and the local detail characteristics.In addition,the validation error showed that the average MABE and RMSE of minimum termperture on the 15th day of every month are 1.128?and 1.626? respectively,which are slightly bigger than horizontal plane.The largest error is in summer and the smallest in winter.The results were slightly underestimated on the 15th day of each month.(4)The local laws of simulation results on the complex terrain show that the monthly mean global radiation,the monthly mean minimum temperature and the minimum temperature on 15th day of each month all reached the maximum in the south slope,and the minimum in the north slope.The effect of slope on them had seasonal law which is January>October>April>July.With the increase of slope,altitude and latitude,the effect of slope on them in January showed an increasing trend.