| Water is an essential and important material for human and all living things on the earth, and it is extremely valuable natural resources in the process of agricultural production and economic development. Soil moisture is one of the important parameters in water cycle, it is not only the link between precipitation and groundwater, but also determines the growth and development of vegetation. The blown-sand region of the Ordos is an arid and semi-arid desert region, which have little rainfall and strong evaporation. In recent years, surface soil moisture evaporation and loss leads to regional soil salinization, desertification and vegetation and land degradation. Not only has restricted the development of local agricultural economy, the ecological environment is also increasingly fragile. Therefore, it is significant to monitor soil moisture in drought and water shortage areas.Microwave remote sensing technology is the main means of soil moisture monitoring at present. Compared with the traditional measurement methods, microwave remote sensing has the advantage of full day, all-weather, and can realize large area and real-time monitoring. Soil moisture retrieval by microwave remote sensing is a very effective method, which has practical significance for the assessment of water environment and the rational use of water resources.Soil moisture retrieval by microwave remote sensing is mainly establishing the function relationship between the backscattering coefficient and soil moisture. However, surface roughness is an important parameter affecting the backscattering coefficient for bare area. Therefore, accurately describing the surface roughness and eliminating the influence of the roughness on the backscattering is the key to improve the soil moisture retrieval accuracy.In order to achieve soil moisture retrieval of blown-sand region, the relationship between backscattering coefficient, the roughness as well as backscattering coefficient and soil moisture were simulated based on AIEM(Advanced Integral Equation Model). The three models are established, which are the model based on combined roughness, the model based on effective correlation length and the model based on roughness calibration. Explore these three ways to describe surface roughness. By evaluating the accuracy of each model, the optimum soil moisture retrieval model is determined Soil moisture retrieval model was established based on RADARSAT-2 C-band quad polarization(VV, HH, VH, HV) data and make the soil water distribution map. And analyze the relationship between soil moisture and topography from different scales by using the DEM data and GF-1 remote sensing data of study area. The main research results are as follows:(1) The experimental results show that the model of soil moisture retrieval by using the Fresnel reflection coefficient based on AIEM is better than the one which is established by soil moisture content.(2) In all of models, precision of models based on combined roughness is high. Roughness calibration models are better. Effective correlation length models are not good. Furthermore, soil moisture inversion model based on new combined roughness parameter Zs=s3/l by using AIEM have a high accuracy, and R, MRE, MAE, IA and RMSE are 0.942, 26.4%, 4.95%, 0.885 and 6% respectively. The model can be applied to the retrieval of soil moisture in arid and semi-arid regions.(3) Soil moisture retrieval map show that soil moisture is considerately low, with the average of 12% in study area. Some areas where soil moisture is less than 10% almost are belonging to dunes, with an obviously character of sandy soil. In generally, soil moisture is distributed alternatively, shaped into strips in direction of northwest– southeast.(4) Through the typical profile analysis of the relationship between soil moisture, topography and geomorphology at different scale, it was found that soil moisture content distribution is in keeping with sand dunes and marsh regions, which means that the soil moisture of dune area is relatively low, below 12%, while the soil moisture of marsh area is relatively much higher, with the average of 21%. According to the topographic features, it can be seen that dunes with high altitude and marsh zone are distributed cross each other in the entire study area, and their shape is a strip of northwest-southeast. In addition, It can also been acquired that the surface of some sand dunes take on a rugged and scaly distribution because of the wind. Profiles of different scales display that soil moisture retrieval has a comprehensive fit with the terrain, which can verify the reliability of soil moisture retrieval further. |
PDF Full Text Request