The Simulating Of Land Surface Temperature With CLM Based On MODIS LAI

As a key variable in land surface process, surface temperature plays a vital role in surface-atmosphere interactions. It is important to strengthen the understanding of the land surface physical process and surface-atmosphere interactions through accurate estimation of surface temperature in spatial and temporal distribution. Station observation, remote sensing and simulation of land surface process are unable to offer continuous and high precise surface temperature information separately. By combining remote sensing and land surface model, this paper have applied remotely sensed data into land surface model and researched surface temperature simulation, sensitivity analysis of LAI and comparison of different MODIS LAI application using different MODIS LAI from2006to2008. The conclusions are summarized as follows:(1) The technology about different MODIS LAI of high spatial and temporal resolution applied to CLM simulations was developed based on analysis of different MODIS LAI data structure and CLM program structure. It included the methods such as coordinate transformation, sampling, weighting calculation and data format conversion. It also needs weighting calculation about PFT in order to get monthly average LAI which matches with CLM.(2) Sensitivity analysis shows that surface temperature performs a trend of decrease with LAI increasing. In the same error range of LAI, the simulation results are more closed to the reference as LAI is small. CLM is more sensitive to LAI when it is getting bigger gradually. In spring and summer, the influence of LAI is bigger and it is smaller in autumn and winter. The influence of LAI in the middle-lower reaches of the Yangtze River is bigger than that in the northeast area in winter. The sensitivity of LAI with season variety is more apparent in the northeast area than that in the middle-lower reaches of the Yangtze River. The relative error of surface temperature is generally bigger in the northeast area than that in the middle-lower reaches of the Yangtze River in the condition of different LAI error.(3)At the scale of site or regional, the improvement of simulation precision of different MODIS LAI is obvious and the simulation precision of improved MODIS LAI is higher. At site scale for the year average, compared to original data of CLM, the deviation-relative error and root mean square error of simulation results based on improved MODIS LAI is reduced by9.7%、10.1%、^2.2%, respectively, and they are reduced by2.7%、3.9%、1.4%, respectively, for MODIS LAI. At site scale for the season average, the errors of three simulation results are generally bigger in spring and winter, autumn following, and smallest in summer. Compared to original data of CLM, the accuracies of the simulation based on two kinds of MODIS LAI in spring and summer are higher than that in autumn and winter, improvement effect based on improved MODIS LAI in summer is obvious. At site scale for monthly average, compared to original data of CLM, the deviation、 relative error and root mean square error of simulation based on improved MODIS LAI is reduced by23%、21.7%、2.5%, respectively, and they are is reduced by9.8%、4.5%、1.9%, respectively, for MODIS LAI. At regional scale, such as the improvement of simulation precision of different MODIS LAI is obvious in northeast area、the middle-lower reaches of the Yangtze River、 southwest area and the southern area of North China, and the simulation precision of improved MODIS LAI is higher. In the seven observation area, the three simulation deviation is the biggest with the observed value in winter, spring following, and it is the smallest in summer and autumn.