摘要
采用大型蒸渗仪研究不同地下水埋深条件下微咸水（咸水）灌溉对土壤水盐的影响,在此基础上,建立了BP神经网络水盐耦合模型。研究结果表明：土壤中盐分含量随地下水埋深的增加而减小,随灌溉水盐分水平的增加而增大。在地下水埋深为2 m,3 m,4 m的地区,采用矿化度小于4 g/L的水进行灌溉,在夏玉米整个生长周期的0～100 cm土层内不会形成积盐现象;拓扑结构为8∶2∶2的BP网络模型,能模拟不同地下水埋深条件下微咸水（咸水）灌溉的水盐运移,且精度较高。
In this paper,the large lysimeter was used to study the effect of brackish water irrigation on soil water and salt under the conditions of different groundwater depth.The results show that soil salinity decreased with groundwater depth increasing and salinity level of irrigation water decreasing.Salinity accumulation does not appear in 0～100 cm soil layer in the whole growth stage of summer maize when water salinity level is less than 4 g/L and groundwater depth is 2m,3m and 4m.The BP network model that the topology structure is 8∶2∶2 can simulate soil water and salt by brackish water irrigation under the condition of different groundwater depth,and the accuracy is higher.
出处
《水土保持学报》
CSCD
北大核心
2007年第6期7-10,15,共5页
Journal of Soil and Water Conservation
基金
国家科技支撑计划项目(2006BAD17B02)
国际合作项目(2006DFA72190)
