Simulation Study On Dynamics Of Soil Moisture In Degraded Wetland

Wetland, together with forest and ocean, makes up the three ecosystems of the whole world. Wetland, which is called "the kidney of the Earth", is one of the richest ecological landscapes abound in biodiversity and one of the most important habitats for mankind. Studying the characteristics of aquatic ecosystem is one of the most important targets for wetland science research. Humidity, determined by soil moisture, may have a direct impact on spatial distribution patterns of wetland vegetation communities. Research on soil moisture is important because of its close and complex exchange relationship with the atmosphere, surface water, groundwater, plants and other elements, and it is also highly related to the characteristic of plant environment gradients.Using quantitative numerical simulation approaches to analyze the dynamic patterns of soil moisture has become an important means for soil moisture research, especially after1980s. The HYDRUS-3D model (Simunek et al.,2008), based on Richards equation, is a numerical model, which simulate water, heat, and solute transport in variably saturated porous media, and it is developed by the US Salinity Laboratory and International Ground Water Modeling Center (IGWMC). Although the mechanism and methodology of wetland models are familiar to researchers, but detailed data still be required for calibration and evaluation when analyzing soil moisture dynamics using these models. The complicated wetland environment with low accessibility and difficulty in monitoring and accessing data make it more difficult to describe the characteristics of wetland vegetation and the transitional zone between terrestrial and aquatic ecosystems eco-hydrological patterns. Therefore, most of the studies on soil moisture dynamics with HYDRUS software were focus on agriculture ecosystem, thus few studies have been done on soil moisture under different natural wetland plants.The Honghe National Natural Reserve (HNNR) is one of the typical inland wetlands in Northeast Plain. Based on the characteristics of eco-hydrology environment gradient in this area, we selected a typical wetland vegetations gradient area in HNNR as our experiment simulation area. The objective of this study is to analyze the different dynamic patterns of soil moisture estimated by HYDRUS and examine the different eco-hydrological process between different wetland plants. Data used in this study were meteorological data from automatic weather station in HNNR, daily soil moisture date from soil moisture recorder and soil particle size analysis results from laboratory experiment. Model validation processes were done by comparing the simulated soil moisture results with the observed soil moisture. And then, we analyze the HYDRUS-3D simulation results, including the soil moisture dynamic patterns of different wetland plants, transpiration rate and deep drainage of different wetland plant, to study the eco-hydrological process difference between different wetland plants. In addition, we discussed the uncertainty of simulation results. In this study, the following conclusions were obtained:(1) Soil moisture contents of the forest and meadow simulated by HYDRUS-3D model show good agreement with measured data, and the values of RMSE ranged between0.0291and0.0457cm3/cm3, R2varies from0.44to0.69.(2) The standard deviation and coefficient of variation and other statistical index were used in this study to describe the dynamic patterns of soil moisture at different soil depth. We found the dynamic pattern of surface soil moisture is different with that of deep soil moisture. The weather and surface plant cover condition have a great influence on surface soil moisture, thereby the standard deviation and coefficient of variation of the surface soil layer are higher, and the variation of the surface soil moisture is significant; by contrast, the deep soil moisture stay stable because it is not affected by weather condition, as a result, the standard deviation and coefficient of variation of the deep soil moisture are lower.(3) The comparison of dynamic patterns of soil moisture between island and meadow results show that, the volumetric soil water contents of meadow increase while coefficients of variation of soil moisture decrease with the increase of soil depths (0-40cm). By contrast, the volumetric soil water contents of forest increase between20cm and30cm soil depth but then decrease in the depth of30-40cm. Influenced by soil texture, the soil moisture of the island forest below20cm depth have similar response to rainfall, while the soil moisture of meadow above30cm depth have similar response.(4) Wetland water balance analysis result indicates that, meadow, the typical and important wetland plant type in the study area, has a greater water deficit than forest during the growing season, therefore is more likely to suffer from water shortage. Because of that, the most significant water supply time for meadow is between late July and early August. And this is very important for maintaining meadow habitant water requirements.(5) HYDRUS simulation results uncertainty resulted from data uncertainty, model structures uncertainty and parameters uncertainty. By comparing the observed soil moisture of island forest and that of island forest and brush transition zone plant between the same period in2010and2011, we found although installing soil moisture recorder will not influence the soil water movement in the soil profile where probes installed, difference between observed soil moisture and actual soil moisture still might existed. This difference resulted in the uncertainty of validation data, which belongs to the uncertainty of data.Based on the basic principles of eco-hydrology and soil water dynamics, we used finite element model HYDRUS-3D to simulate and analyze the dynamic patterns of soil moisture under different typical wetland plants in Honghe Nation Natural Reserve. This study will led to new understanding of mechanism of interactions between surface water-soil water-wetland plants; besides, the study results will provide scientific basis for further study on mechanisms of eco-hydrology interaction among different wetland plants at larger scales, such as catchment scale and basin scale. Because few studies have been done on the characteristics of wetland plants and their habitat, especially for the natural wetland plants, therefore, study on soil moisture dynamic patterns of different natural wetland plants will has more practical value and will provides critical information for protection and recovery of wetland ecosystems.