Considering The Temperature Sensing Estimates Of Land Cover And Spatial Interpolation Data Improvements

For research about the air temperature information, we could only get the point scale instead of space continuous via meteorological stations which distribute everywhere distant. Of course we’re able to get our goal by spatial interpolation, however, the sparser the original data distribute, the more errors we’ll get. By remote sensing, we can easily get rid of the space scale problem, but when there’re no satellites above our research area, we get nothing but vacancies on time scale. In order to construct an air temperature field of continuous on both time and space scale, we build a model for the relation between air temperature and land temperature. By using this model, not only can we calculate air temperature under the conditions of continuous land temperature field, but also can estimate the land temperature with no satellites’data for long time range, so that contribute to the career on the laws of thermal environment’s changes on both time and space scale.The main contents of this article are as follows:1. Compare among sorts of interpolation methods based on100meteorological stations’ air temperature dataBased on100meteorological stations’air temperature data,we complete three kinds of interpolation methods:IDW,Kriking,Spline,and compare their individual results on statistical errors. The results says, the interval errors is too big for Spline to get a right scale of the air temperature and the mean value and the variance are neither satisfactory on statistical field. IDW will get exception regions because of the inhomogeneous distribution for distribution of original data. Kriging get the least relation with distribution of the original data. So we can conclude that Kriging is the most objective.2. The research about the distribution characteristics of land temperature field which affected by land cover based on remote sensing data (MODIS)Based on8days’synthetic data about land temperature(MODIS),we get the distribution of Peking’s land temperature of each month in2012.Then with ranking them by mean value-the variance, we get the distribution characteristics of land temperature field on both space and time scale. The result says that downtown area and southern area get a higher temperature and the trends is colder out-of-town than downtown. Seasonally, the summer get a larger span range than winter.Without considering the land cover kinds, we couldn’t get a satisfactory results by spatial interpolation when there’s inhomogeneous distribution for distribution of original data. So we use VIS model to get a inversion of impervious surface area and research the relation between impervious surface area and land temperature. Then we draw the conclusions that impervious surface area and land temperature has a positive correlation3. Improvement of air temperature’s spatial interpolation by considering land cover conditionIn order to get a excelsior result of air temperature’s, we build a model about estimating air temperature with statistical methods, rainfall, impervious surface area, seasonal factors considered. Then the result says that linearly dependent condition between land cover and land temperature varieties. Via precision test about the model and comparison with spatial interpolation results, we could fix the interpolation results to achieve our goal.