西安灞河水源地地下水位埋深的变化及RBF模型预测  

Change of Groundwater Depth and Its Prediction with RBF Model in Water Source Region of the Bahe River Basin in Xi'an

作　　者：董艳慧[1] 周维博[2] 马艳 卜卿 闫少芳 

机构地区：[1]西安工业大学建筑工程学院,陕西西安710032 [2]长安大学环境科学与工程学院,陕西西安710054 [3]西安市水资源利用技术服务中心,陕西西安710007 [4]河北建投任丘热电有限责任公司,河北任丘062550

出　　处：《干旱区研究》2015年第1期23-28,共6页Arid Zone Research

基　　金：国家自然科学基金项目(51009008);西安工业大学校长科研基金(XAGDXJJ1123)资助

摘　　要：为了探究减少地下水开采量对地下水位回升的修复效果,对西安市灞河水源地1984—2005年的地下水开采量及水位变化进行了分析,并对减少开采量后,潜水和承压水位埋深进行了模拟和预测。以蒸发、降雨、径流和人工开采量作为输入变量,利用RBF神经网络模型,预测2006—2023年的潜水和承压水位埋深。1984—2001年的数据用于模型的训练,2002—2005年的数据用于模型的验证。结果表明:灞河水源地潜水和承压水位有不同程度的上升。2023年潜水和承压水位较2005年最大上升幅度分别为1.5 m和1.24 m,仍比1984年水位分别低11.28m和13.34 m,回升速度较慢,2023年承压水位降落漏斗依然存在。In order to reveal the restoration of groundwater level after reducing the volume of groundwater exploitation,the change of groundwater exploitation and groundwater level in water source field of the Bahe River Basin in Xi＇an during the period from 1984 to 2005 was analyzed. Furthermore,the depths of phreatic water and confined water after reducing the groundwater exploitation were simulated and predicted. The evaporation,precipitation,runoff volume and groundwater exploitation were regarded as the input variables,and the Radial Basis Function neural network model was used to predict the depths of phreatic water and confined water during the period from^2006 to 2023. The data during the period from 1984 to 2001 were used to train the model,and those during the period from 2002 to 2005 were used to validate the model. The results showed that the phreatic water and confined water levels in 2023 would be raised by 1. 5 m and 1. 24 m in maximum compared with those in 2005. They,however,were still deeper than 11. 28 m and 13. 34 m respectively compared with those in 1984. The depression cone of confined water would still exist in 2023.

关 键 词：水源地 地下水位埋深 预测 潜水 承压水 霸河 

分 类 号：P641[天文地球—地质矿产勘探]