首页> 外文会议>Specialty Symposium on Integrated Surface and Ground Water Management, May 20-24, 2001, Orlando, Florida >A Neural Network System for Predicting Ground-Water Elevations at User-Specified Sites Based on Regional Surface-Water Data
【24h】

A Neural Network System for Predicting Ground-Water Elevations at User-Specified Sites Based on Regional Surface-Water Data

机译:基于区域地表水数据的用户指定地点地下水高程预测的神经网络系统

获取原文
获取原文并翻译 | 示例

摘要

Fairbanks, Alaska soils are highly transmissive with a very shallow water table. Consequently, the interaction of ground-water and surface-water can be linked or correlated. A neural network model was developed to predict the ground-water elevation at user specified sites of interest based on the discharge data of a nearby river, the Chena river. The neural network model was trained using the samples provided by a field-calibrated, finite element model of ground-water flow for the Fort Wainwright region, adjacent to Fairbanks. The United States Geological Survey has maintained a discharge stage gage for the Chena river since 1948. Once trained with available field data, the neural network model can predict the ground-water elevation at sites within the study domain throughout this time period. This strategy and implementation can reduce ground-water monitoring costs and labor requirements substantially. This system also provided the necessary underpinnings to model contaminant fate and transport situations over extended periods of time.
机译:阿拉斯加费尔班克斯的土壤具有很高的透光性,地下水位很浅。因此,地下水和地表水之间的相互作用可以相互联系或关联。开发了一个神经网络模型,根据附近的Chena河的流量数据预测用户指定的感兴趣地点的地下水位。使用由费尔班克斯附近的Fort Wainwright地区地下水流动的现场校准有限元模型提供的样本对神经网络模型进行了训练。自1948年以来,美国地质调查局一直在为Chena河保持泄洪量。在接受了可用的野外数据训练后,神经网络模型可以预测这段时间内整个研究区域内的地下水高程。这种策略和实施可以大大降低地下水监测成本和劳动力需求。该系统还提供了必要的基础,可以模拟污染物在较长时间内的命运和运输情况。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号