Study On Calculation Of Surface Water Vapor Pressure Using MODIS Near Infrared Data

Surface water vapor pressure is a basic data in atmospheric science and hydrology research,so it is very important to obtain the high-resolution surface water vapor pressure.The terrain of the study area Guizhou Province is complex,and its spatial distribution of water vapor pressure is quite different.There are limited number of conventional ground meteorological stations on the ground because of topographical,socio-economic and other problems.There is a significant uncertainty with the surface water vapor pressure distribution map of the study area based on ground station observation data space interpolation.In this paper,we use MODIS remote sensing data to estimate the surface vapor pressure of clear sky in Guizhou Province.According to the different ways of using MODIS data,the paper introduces the remote sensing inversion method of surface water vapor pressure by indirect method and direct method.Firstly,the indirect method principle of estimating surface vapor pressure based on water vapor content is elaborated.Then,Using principle of near-infrared water vapor inversion for reference,the direct models are used to estimate surface water vapor pressure.There are four direct models,exponential model,polynomial model,channels linear model and integrated model containing elevation,which are developed.We can choose one from three absorption channels(17,18 and 19),and choose one from two differential transmittance(reflectance and radiance)when we use MODIS Near Infrared data,so several equations are built to calculate surface vapor pressure in each direct model.The MODIS remote sensing data(MOD021km,MYD021km,MOD05,MYD05)of Guizhou Province from June 2017 to February 2019 and the conventional observation data of ground meteorological stations are collected.Through the cloud detection of remote sensing images,eight typical remote sensing images with less cloud cover were selected,and the direct method,indirect method and spatial interpolation method were used to estimate the instantaneous surface water vapor pressure at the imaging time,and the fitting effects and errors of each algorithm were compared.According to the preferable estimating model of instantaneous surface water vapor pressure,the average surface water vapor pressure of spring,summer,autumn and winter from June 2017 to May 2018 was calculated,and the average surface water vapor pressure maps of four seasons with 1km~2 resolution in the study area was obtained,then the temporal-spatial characteristics and causes of these maps are analyzed.After the above research,the following conclusions are obtained.(1)Based on MODIS remote sensing data,the surface water vapor pressure distribution map with 1km~2 resolution can be generated,which has spatial advantages and can better reflect the local characteristics of water vapor pressure distribution in complex terrain region.(2)In this paper,the MODIS near-infrared reflectance image is used to directly calculate the surface water vapor pressure,which avoids the acquisition or calculation of intermediate data(Precipitable Water Vapor products),and the direct models has higher precision.(3)Because the three absorption channels have different sensitivities to water vapor,the absorption channels used in preferable estimating model of four seasons are different.In the summer and autumn models,the 17~thh channel has a higher coefficient of determination and a smaller error,while in spring and winter,it is the 18~thh channel and the 19~thh channel respectively.(4)The spatial and temporal differences of average surface water vapor pressure in Guizhou Province are significant,and the monsoon and terrain have crucial impacts on the distribution of water vapor.The high value areas in each season are mainly located in the southwestern part of Guizhou Povince and the western part of Zunyi Prefecture.In summer,Southeast Guizhou is also the high value area of surface water vapor pressure.The low value areas are mainly distributed in the Bijie and Liupanshui Prefecture,which are in the northwest of Guizhou Province.