| Water resource is one of the core limiting factors of ecosystem on the Loess Plateau. Soil moisture is not only a crucial part of soil but also a direct water resource for dry land framing and vegetation growth. Studies on soil moisture change and its spatial distribution are a prerequisite to understand the climate change and vegetation-soil relationship, and also the important foundation to scientifically and reasonably restore sustainable vegetation.In the past 20 years, especially after implement of the Grain for Green project(GGP) a large scale of vegetation restoration generated significant impacts on the regional ecosystem and environment of the Loess Plateau. In the meanwhile, the soil and water processes and distribution have been changed. In order to understand the temporal and spatial changes of surface soil moisture and its response to the re-vegetation at a large scale, the soil moisture data extracted from the European Remote Sensing Satellite scatterometer(ERS) and normalized difference vegetation index(NDVI) were analysed and combined with the field observed data, land use and climate information using the classical statistics, geostatistics, and time series analysis theory and methods. This study verified the practicability of extracted data of microwave remote sensing for surface soil moisture and.validated the extraction model and related factors suitable for the Loess Plateau. Based the above research results,the regional surface soil moisture changes were investigated and the temporarily and spatially correspondence with re-vegetation and precipitation were analysed, as a result, a preliminary understanding on the response range and degree of soil moisture changes to vegetation restoration at large scales was achieved.The main results and achievements of this study listed as follows:(1) Through the comparison of the ERS scatterometer data and field measurements, combined with the feature of the loess plateau soil, soil moisture inversion model for microwave remote sensing data and its parameters were determined with consideration of the feature of soils on the Loess Plateau. Soil volumetric water content(Wswi) corresponding to surface soil moisture data(soil moisture index, SWI) was calculated. The results showed that there was an extremely significant correlation between Wswi and field measured soil moisture at the 0-10 cm deep, and the correlation decreased with increased depth, demonstrating that the microwave soil moisture data could well represent the actual soil moisture conditions, and its spatial distribution. This work suggested that the ERS inverted soil moisture data is applicable over the Loess Plateau, and can be used to explore the spatio-temporal change of the surface soil moisture content.(2) It was demonstrated that there was a high correlation between the ERS inverted Wswi and the field measurements over the flat areas, where croplands were widespread, a low correlation was detected in areas with a combination of agriculture, forest and grazing etc. For the rain-fed agriculture and irrigation agriculture meteorological sites, the intra-annual and inter annual changes of field measurements and microwave remote sensing inverted soil moisture were consistent with those of precipitation, but in a relatively dry season, farmland water determined by satellite was generally lower than field measurements. This implied that during the application of the satellite data to estimates of the farmland soil moisture, in addition to consider the land-form, soil, vegetation, climate and other natural factors, care should also be taken to take into account of the effects of agricultural measures(e.g., irrigation) on soil moisture. Application results showed that the parameterization of the inversion model was usually completed for particular regions rather than the globe. The values of the parameters should be adjusted when these models are applied to wider regions, and some new models are also needed which are able to address the specific environmental conditions.(3) On the basis of the comparison with the traditional regional soil moisture survey data, the microwave remote sensing inversion of 0 ~ 100 cm surface soil moisture data better reflected the spatial and temporal change of soil moisture condition over the loess plateau area from 1992 to 2013. In general, soil moisture content was higher in the south and east, lower in the north and west. Seasonal soil water value reasonably captured seasonal dynamics of soil water throughout a year: loss of soil water in spring, increase of soil water in summer, a fairly slow loss of soil water in autumn and winter. It gradually decreased from southeast to northwest. There was an apparent gradient for the vegetation, precipitation and soil moisture distribution along the selected sample belts. The results also showed that 1998-2000 were a turning point for the NDVI and soil moisture, strongly suggesting that the GPP significantly affected the vegetation and soil moisture condition across the Loess Plateau.(4) Through coupled analysis on spatio-temporal patterns of vegetation restoration and surface soil moisture content between the beginning of the GGP and 10 years later, it was found that NDVI significantly increased over 80.99% of the Loess Plateau, but surface soil moisture reduced for 72.64 of the region. Based on the overlay analysis of changes in vegetation cover and soil moisture, 57.65% of the Loess Plateau underwent increased vegetation cover but decreased soil moisture, and precipitation in 32.80% of these areas increased. There was 23.34% of the Loess Plateau with increased vegetation cover and soil moisture and 14.99% of the Loess Plateau with declined vegetation cover and soil moisture content.The findings of this study demonstrated that in arid and semi-arid regions the vegetation restoration at large scales could lead to soil scarcity. Adjustments to vegetation restoration strategies are required in terms of soil moisture change in order to achieve sustainable development of regional ecosystems. |
PDF Full Text Request