The Application Research Of Soil Moisture Retrieval Based On The VI-LST Feature Space

Soil moisture is the soil water content of a certain depth, which plays a crucial role in the energy and mass exchange between the atmosphere and land. And it is often used as an environment factor and process parameter in meteorology researches while the automatic observation is an important source of soil moisture data. With the global warming, meteorological disaster especially drought show increased activity. Yunnan Province in the Southeast China, drought lasted longer and caused more serious consequence in the past few years in Yunnan. So the research of soil moisture is urgent and significant.Now automatic meteorological site and satellite remote sensing is the main source of soil moisture data. However due to the diversity of climatological conditions covered by the stations and differences in measurement setup, the quality of the measurements is highly variable, which may have a significant impact on the accuracy of the studies to the soil moisture. Satellite remote sensing is not restricted by the surface condition, and has a well continuity in space, but also short in temporal continuity. So to combine the automatic meteorological data and satellite remote sensing data may provide more soil moisture information.The paper calculated 3 vegetation index based on the MODIS albedo production, compared with in situ data, analyzed the correlation. Then built VI-LST feather space with MODIS land surface temperature production, obtained the soil moisture inversion equation. Made a further discuss of the applicability of 3 VI-LST feather space in Yunnan Province.Based on the soil moisture data of thirty-seven stations of Yunnan province in 2010-2014, the spectrum-based approaches were used in this paper to detect three kind of abnormal data. The constant, spike and noisy, caused by saturation of the signal and unresponsive sensors, were screen out by procedures analysis of the shape of the soil moisture time series. The result showed that, the existence of absent data are general among the stations, especially in some stations in the northwest of Yunnan Province, where the instrument set up later than other stations. And the missing had similar concentrations in the few months after completion of these automatic soil observation stations, which may related to the instability of instrument in the beginning. Constant accounted for the vast majority of abnormal value, the proportion of constant is up to 97% in all of the stations. The distribution of stations contains more spike showed less distinct pattern.When compared 3 vegetation index with soil moisture, EVI shows better correlation and stability, while MSAVI and NDVI have higher correlation coefficient in wet land situation, and NDVI is most closely related to soil moisture in Building lands and cities. However, the result of 3 VI-LST feather space revealed difference. NDVI-LST is superior to other 2 space, the matching of MSAVI-LST present a seasonal variation, and the EVI-LST can still be improved.Compared to the related meteorological data, diversities were found in the response of the soil moisture to the precipitation among different soil type. The stations of sand soil show quicker reaction to the precipitation, and the change of soil moisture is more significant as well. As the rainfall intensity increasing, the rise of the moisture increasing. The soil moisture in the stations of loam soil show continuous rise in the few hours after the rainfall, but the variation is rather weaker, and the deeper soil moisture change relatively smooth. The precipitation and soil moisture in the stations of clayey soil have a weak relationship, the soil moisture rise little after rain fall, and would descend more quickly in the deeper layer than other soil type, with the most probably reason the clayey soil has a larger density and compatibility. And we found the the rainfall intensity and the soil moisture rise are not simply positive correlated, the soil moisture rise less after downpour on the contrary, more detailed would be discuss in the further research. The soil moisture in Yunnan Province shows a overall trend of north higher the south, and is concerned with its water resource distribution. Summer soil moisture is much drier than that in winter, and this agrees with the summer drought happen in Yunnan in 2012.