摘要
To better understand soil moisture dynamics in the Yangtze River Estuary （YRE） and predict its variation in a simple way, a field monitoring experiment was carried out along the north branch of the Yangtze River, where seawater intrusion was strong and salt-water variation is one of the limiting factors of local agriculture. In present paper, relation between antecedent precipitation index （API） and soil water content is studied, and effects of groundwater depth on soil water content was analyzed. A relatively accurate prediction result of soil water content was reached using a neural network model. The impact analysis result showed that the variation of the API was consistent with soil water content and it displayed significant correlations with soil water content in both 20 and 50 cm soil layer, and higher correlation was observed in the layer of 20 cm. Groundwater impact analysis suggested that soil moisture was affected by the depth of groundwater, and was affected more greatly by groundwater at depth of 50 cm than that at 20 cm layer. By introducing API, groundwater depth and temperature together, a BP artificial network model was established to predict soil water content and an acceptable agreement was achieved. The model can be used for supplementing monitoring data of soil water content and predicting soil water content in shallow groundwater areas, and can provide favorable support for the research of water and salt transport in estuary area.
To better understand soil moisture dynamics in the Yangtze River Estuary （YRE） and predict its variation in a simple way, a field monitoring experiment was carried out along the north branch of the Yangtze River, where seawater intrusion was strong and salt-water variation is one of the limiting factors of local agriculture. In present paper, relation between antecedent precipitation index （API） and soil water content is studied, and effects of groundwater depth on soil water content was analyzed. A relatively accurate prediction result of soil water content was reached using a neural network model. The impact analysis result showed that the variation of the API was consistent with soil water content and it displayed significant correlations with soil water content in both 20 and 50 cm soil layer, and higher correlation was observed in the layer of 20 cm. Groundwater impact analysis suggested that soil moisture was affected by the depth of groundwater, and was affected more greatly by groundwater at depth of 50 cm than that at 20 cm layer. By introducing API, groundwater depth and temperature together, a BP artificial network model was established to predict soil water content and an acceptable agreement was achieved. The model can be used for supplementing monitoring data of soil water content and predicting soil water content in shallow groundwater areas, and can provide favorable support for the research of water and salt transport in estuary area.
基金
financially supported by the Ecological and Environmental Monitoring Project (JJ[2011]-017)
funded by the Executive Office of the Three Gorges Project Construction Committee of the State Council of China
the National Non-Profit Research Program of China (200903001)
the National Basic Research Program of China(2010CB429001)
作者简介
XIEWen—ping．Tel：＋86—25—86881231．E—mail：wpxie@issas．ac．unYANGJing—song，Tel：＋86—25-86881222，Fax：＋86—25—86881000，E—mail：jsyang@issas．ac．cn
