| Soil moisture is an important variable of the global water cycle system,controlling the process of water,energy and carbon exchange between the surface and atmospheric interfaces.It has an important impact on the climate system and its changes by affecting land surface evapotranspiration,water migration and the carbon cycle interacting with the Earth’s climate system.Accurate soil moisture information is the core input parameter for various climate models,such as hydrological model,ecological model,and land surface model.Compared with other remote sensing techniques,passive microwave remote sensing has the advantage of being sensitive to soil moisture changes under all-weather conditions.As the latest generation of soil moisture mission,SMAP is equipped with the most advanced L-band microwave radiometer to provide high-quality observation data.However,due to the low spatial resolution of passive microwave soil moisture products,it can only be applied to large-scale applications and cannot be applied to small-scale or regional scale applications.Tibetan Plateau is an early warning zone,sensitive zone and start-up zone for global climate change.Soil moisture is particularly important in the study of its climate and water cycle.However,due to lack of high spatial resolution soil moisture data severely limits the development of this area.Downscaling the SMAP passive microwave soil moisture to obtain high-spatial resolution soil moisture information over the Tibetan Plateau and exploring its application value,is of vital importance to the study of regional and even global hydrological and environmental changes.In the downscaling process,it is necessary to reasonably introduce high-resolution auxiliary information.Visible-thermal infrared remote sensing data is an ideal data source for providing auxiliary information with its rich observation,high revisit cycle and high resolution.However,there are still some problems that need to be solved in the process of visible-infrared remote sensing data assisted passive microwave soil moisture downscaling.The work of this thesis is carried out in this context.This paper takes the Tibetan Plateau as the research area and studies the core issues in SMAP soil moisture dowscaling,namely data input,algorithm development and application of results.The research content of the thesis mainly includes the following four aspects:(1)A comprehensive and detailed evaluation work of three SMAP operational soil moisture products was conducted over the Tibetan PlateauSMAP passive microwave soil moisture product is the processing object of this paper’s downscaling algorithm,and its accuracy directly determines the pros and cons of downscaling results.There are significant differences between the east and west across the climatic zone and the underlying surface in the Tibetan Plateau.It is vital to make reasonable use of ground observations to best represent the true level of soil moisture in the region,and to fully understand the accuracy of SMAP soil moisture products.To ensure the reliability and comprehensiveness of SMAP soil moisture product evaluation,this paper proposes a passive microwave soil moisture product validation strategy.In-situ soil moisture mearments from three networks,Ngari,Nagqu and Maqu which located in different areas of the Plateau were collected.The three soil moisture networks cover different climatic zones and vegetation cover conditions,representing the arid,semi-arid and humid soil moisture levels of the Tibet Plateau.Three soil moisture products SCA-V,SCA-H and DCA of SMAP were evaluated over the Tibet Plateau.In the comparison of SMAP soil moisture from differernt orbits,the soil moisture from ascending orbit in the Nagqu and Maqu areas is better than that from decending orbit.In the comparison of different products,SCA-V,SCA-H and DCA can better capture the change trend of soil moisture.From the comparison of accuracy metrics,SCA-V is superior to the other two products.ub RMSD of SCA-V basically meets the accuracy goal of SMAP less than 0.04 cm~3/cm~3.Therefore,SCA-V soil moisture products are recommended in later downcaling studies.(2)Develop an time-space continuous land surface temperature estimation method aiming at the problem of space discontinuity caused by cloud contaminationLand surface temperature is one of the core parameters of SMAP soil moisture downscaling algorithm.However,the cloudy weather in the Tibetan Plateau has caused the land surface temperature retrival from thermal infrared remote sensing to be seriously missing.Land surface temperature retrival from passive microwave technique can go through the clouds with the all-weather advantage.But it also suffers from the low retrival accuracy and low resolution.Thermal infrared data can retrive high-spatial resolution,high-accuracy land surface temperature,but it is vulnerable to data missing caused by cloud contamination.Therefore,the combination of thermal infrared and microwave data is expected to estimate the surface temperature under all-weather conditions and maintain the high resolution features.This dissertation combines the advantages of AMSR2 and MODIS to estimate land surface temperature,and develops an a time-space continuous land surface temperature estimation method.The method is divided into three steps:(1)uses the X,Ku and Ka-band brightness temperature information from AMSR2,which are sensitive to land surface temperature changes,use microwave polarization difference index to correct the influence of vegetation and soil moisture on brightness temperature,the AMSR2 surface temperature retrival model was established.Compared with MODIS land surface temperature products,AMSR2land surface temperature shows good agreement with MODIS surface temperature;(2)AMSR2 and MODIS land surface temperature fusion,retaining clear sky pixels of MODIS,replacing MODIS with AMSR2 land surface temperature;(3)time interpolation,using time interpolation to fill the missing data of AMSR2 orbital gap.Compared with the original MODIS land surface temperature,the land surface temperature obtained by this method is continuous and can effectively compensate for the data loss caused by cloud contamination(3)Develop a downcaling method to improve the low spatial resolution of SMAP passive microwave soil moistureRelying on the introduction of optical remote sensing data as high-resolution auxiliary information,downscaling passive microwave soil moisture is an effective way to improve its resolution.This paper presents a downscaling method that combines multiple soil moisture indicators.In addition to using classical soil moisture indicators such as surface temperature and NDVI,this method also makes full use of the MODIS red/near/shortwave wave spectrum feature space.Several indicators correlated to soil moisture are involved in the downscaling process,and the gradient decision tree model is used to express the relationship between SMAP passive microwave soil moisture and soil moisture indicators.The method was applied to the Tibet Plateau to obtain soil moisture at spatial resolution of 1 km.In-situ soil moisture measurements from Nagri,Naqu and Maqu were used to evaluate the downcaling results at time,site and regional scales.The vadition results show that the method can improve the spatial resolution of the passive microwave soil while maintaining high spatial accuracy,which is a good addition to the current downscaling algorithm.(4)The high-spatial resolution soil moisture were applied in agricultural drought monitoring,rainfall estimation and landslide susceptibility mappingDue to the lack of high-spatial resolution soil moisture,there is little research on its related applications at home and abroad.The high-spatial resolution soil moisture generated in this dissertation provides an opportunity for applications at regional scale.In the aspect of agricultural drought monitoring,the agricultural drought index SWDI was constructed by using all-weather and high-resolution soil moisture data.Through the comparative analysis with commonly used drought index SPEI,VHI,AWD and drought disaster records,the suitability of SWDI over the Tibet Plateau was comprehensively evaluated for agricultural drought monitoring.It is shown that SWDI can effectively capture the agricultural drought in the Tibet Plateau.In terms of rainfall estimation,the high-spatial resolution soil moisture data was used as the input parameter of the SM2RAIN model,and the rainfall distribution map of 1 km spatial resolution in the Tibet Plateau was obtained.By comparing with the rainfall data of the meteorological site,it is found that the time trend of the two is very consistent.Accuracy statistics also show good agreement between each other.In this paper,the SM2RAIN model is applied to the high-spatial resolution soil moisture for the first time to obtain the corresponding resolution rainfall data,which proves the feasibility of applying the SM2RAIN model to the regional scale.In the future,on the basis of obtaining national or global high-resolution soil moisture,the problem of scarcity of high-resolution rainfall data will be solved.In the landslide susceptibility mapping,the landslides susceptibility in the Tibet Plateau were evaluated by using high-resolution soil moisture as the input data of the infinite slope stability model.Based on the comparison with other landslide susceptibility evaluation methods,NASA landslide catalogue information and the record of China’s flood and drought disasters,the dissertation comprehensively evaluates the suitability of landslide landslide susceptibility mapping which combine soil moisture and infinite stability model.The results show that the proposed method can effectively reveal the spatial distribution of landslide susceptibility in the Tibet Plateau,which provides an important basis for the prevention and management of landslides. |
PDF Full Text Request