Improved Surface Soil Moisture Estimation Methods And Their Applications Based On AMSR Radiometers

Soil moisture is an important natural resource for the existences of earth lives,and it is significant at controlling the water-energy interactive progresses between land surface and atmosphere as well as maintaining the global climate stability.Soil moisture is one of the core input parameters for all kinds of research models on land surface,ecology,hydrology and crop growth,which means that the dynamic variation of this parameter can potentially lead to unnegligible impacts on land surface energy cycles,physical and chemical properties of soil compositions,and global climate change progress.Therefore,accurate and quantitative estimation of global-scale land surface soil moisture is of great significance on maintaining the global ecological system balance and controlling the climate change situation.Besides,soil moisture is also an important indicator for evaluating the suitability of wetness degree in agricultural field,and therefore plays a crucial role on monitoring agricultural disasters such as drought and waterlogging.As a consequence,the accurate knowledge of soil moisture distribution on a large-scale agricultural field is a necessary premise for decent govermental policies on alleviating the natural disasters,so as to guarantee the national food safety for maintaining a harmonious society.In the current time,satellite-based remote sensing technique is one of the most effective ways to obtain spatial and temporal information of soil moisture on a large-scale area.However,the spatio-temporal resolutions of soil moisture observations acquired by difference kinds of satellite sensors are not consistent.Compared with optical remote sensing,passive microwave remote sensing is more sensitive to the variation of surface soil moisture,and compared with active remote sensing,passive microwave observations can produce more stable and reliable soil moisture estimation results.But passive microwave technology also has several disadvantages.First of all,the spatial resolutions of passive microwave observations are generally poorer than the other remote sensing sources.This also indicates that the soil moisture estimation accuracy can be negatively influenced when the static open water fraction within a passive microwave pixel goes higher.Therefore,the further improvement of the soil moisture retrieval algorithm under such condition is much needed.Second,although the spatial resolution of soil moisture products generated by passive microwave technique could be enhanced by downscaling them with optical remote sensing data,optical remote sensing images easily suffer from pixel loss under cloudy or rainy weather,which can seriously reduce the coverage of the downscaled soil moisture images.The above mentioned disadvantages also restrict further applications of passive microwave soil moisture on agriculture and other industries.In this study therefore,we chose the three provinces of Hubei,Anhui.and Jiangsu in the Yangtze River Basin and the Huaihe River Basin of China,as our primary study area.As this area is characterized by a large number of lakes and streams and thus a generally wet soil surface,we employed AMSR-E/AMSR-2 passive microwave brightness temperatures for improving a typical passive microwave soil moisture retrieval algorithm,especially for pixels with higher fractions of static water surface(FSWS).In the subsequent step,we developed a new soil moisture downscaling algorithm which is effective for the cloudy weather conditions.based on this improved passive microwave soil moisture product and MODIS optical remote sensing datasets.And finally,we also investigated some novel and potential applications of the improved soil moisture products generated by this study in the agricultural production industry.The specific research contents of this study are summarized as four key points as in the following:?.Improving passive microwave soil moisture retrieval algorithm for pixels with higher FSWSIn this study we employed the typical passive microwave soil moisture retrieval algorithm of "Land Parameter Retrieval Model(LPRM)" to retrieve surface soil moisture based on AMSR-E/AMSR-2 observations.Then we investigated the influence of different FSWS ranges on the retrieval accuracy of soil moisture,and proposed an improved version of the retrieval algorithm for pixels with relatively higher FSWS ranges.The results show that the improved algorithm can effectively promote the retrieval performance when the FSWS of a pixel ranges between 0.01 and 0.4.?.Downscaling passive microwave soil moisture with optical remote sensing data over cloudy areasAs optical remote sensing data can suffer from pixel loss easily over cloudy areas,we proposed a multi-temporal interpolation method for optical land surface temperature(LST)dataset.Based on the interpolated LST data,we further improved a traditional passive microwave soil moisture downscaling method from its current version.The results show that the improved downscaling method has a better consistency with the in situ(station-observed)soil moisture measurements.As far as the improved method is conerned,the soil moisture data downscaled by interpolated LST data inputs show a close RMSE value to the results downscaled by real satellite observed LST data under cloud free condition(in most validation sites,the difference of their RMSEs is below 0.02 cm3/cm3).Such results suggest that this improved method is capable to obtain soil moisture maps with relatively high accuracies,high spatio-temporal resolutions at full cover in a large area?.Mapping paddy rice planting area based on passive microwave soil moisture dataThis study proposed a novel methodology which employs passive microwave soil moisture data to map paddy rice planting area on a large scale basis.The methodology was applied with the terrestrial area of the entire China(except for Taiwan Province,Hong Kong and Marco)for identifying paddy rice planting area of the country in different previous years.The mapping results show a rather good consistency with historical records from the yearbooks of Chinese government,with a correlation coefficient higher than 0.9.In another validation scheme,the passive-microwave-data-based results also show high correlation coefficients(?0.64)with paddy rice monitoring results based on optical data of higher spatial resolution in all validation areas selected over the country.The former even outperformed the latter in areas of more sporadical paddy rice agricultureIV.Mapping and monitoring of waterlogging disaster on overwintering crops based on downscaled soil moisture dataBased on the downscaled high-resolution soil moisture datasets,we established an indicator system for monitoring waterlogging disasters of overwintering crops.And with the established indicators,we simulated the waterlogging damage degrees of different historical years' crop growth seasons in our primary study area.As with the results,it shows that the spatial distributions of waterlogging disasters are generally consistent with the relevant results obtained based on already-existing meteorological indicators and that obtained based on in situ survey data by previous researchers.This result suggests that the downscaled soil moisture dataset we obtained in this study is generally applicable for monitoring waterlogging disaster of overwintering crops in a large area and has the potential to further improve the efficiency of the waterlogging monitoring and warning system which currently serves the administrative government but depends on the point-scale meteorolocial indicators.