Study On Interpolation Of Temperature On The Tibet Plateau With Bayesian Maximum Entropy

The terrain and geomorphology of the Tibet Plateau are complex,including alpine,subalpine,hilly,basin,plateau and other geomorphological types.The elevation is high in the west and low in the east,which is the main factor restricting the spatial distribution of temperature on the Tibet Plateau,leading to obvious spatial heterogeneity of temperature.Meteorological stations are sparse and irregularly distributed over the Tibet Plateau,so it is of great significance to estimate the spatial/temporal distribution of air temperature by means of spatial/temporal interpolation method with limited data.The traditional temperature interpolation method is mainly based on the measured data of meteorological stations,combined with auxiliary variables such as precipitation and altitude,and the interpolation accuracy is high for the region with flat terrain,dense meteorological stations and reasonable distribution,but for the region with sparse stations and complex terrain,the interpolation accuracy is difficult to guarantee.On the other hand,the rapid development of remote sensing technology provides a wide range of spatial and temporal continuous meteorological information,which can make up for the lack of data caused by the scarcity and uneven distribution of meteorological stations to a certain extent.This paper first summarizes and classifies the main spatiotemporal interpolation methods.Then the principle and calculation process of Bayesian Maximum Entropy(BME)method are introduced.Next,the BME method is used to interpolate the monthly mean temperature over the Tibet Plateau by taking the monthly mean temperature data of meteorological stations as hard data and the monthly mean temperature retrieved by remote sensing as soft data.Finally,the results of BME interpolation are compared with those of inverse distance weighting method,Kriging method,spline function method and Thiessen method using cross validation method.The results show that the BME interpolation results have high estimation accuracy and can reflect the spatial and temporal distribution of monthly mean temperature over the Tibetan Plateau more accurately.In this study,the BME method was used to interpolate the monthly mean temperature over the Tibet Plateau,providing a reference for the spatial-temporal interpolation of temperature or other meteorological factors in the region with rare and irregular meteorological stations and complicated topography.BME method is a successful application of multi-source data fusion method in meteorological data interpolation.