Computationally Efficient Ranking of Groundwater Monitoring Locations

Radvanyi, P., Miller, C. , Alexander, C. , Low, M. , Jones, W. R. and Rock, L. (2023) Computationally Efficient Ranking of Groundwater Monitoring Locations. In: 37th International Workshop on Statistical Modelling (IWSM), Dortmund, Germany, 16-21 Jul 2023, pp. 332-338. ISBN 9783947323425

Abstract

Sampling groundwater quality monitoring wells is a costly and time intensive process that incurs health and safety risks. Reducing the number of wells whilst minimising information loss can greatly increase the sustainability of long-term monitoring. Wells that provide redundant information can be identified by assessing their observations’ influence on statistical model estimates. Well-based cross-validation (WBCV) could be used to obtain such a measure of influence for each well, however, the associated computational cost renders this option unfavourable. In this paper, we propose a method based on influence statistics of regression-based, groundwater solute concentration models, as a computationally efficient, approximate alternative. The method, named well influence analysis (WIA), approximated WBCV results in a simulation study and real groundwater contaminant observations with an average 77% and 73% accuracy respectively. WIA will be implemented in the "well redundancy analysis" feature of GWSDAT, an open-source software for the spatiotemporal modelling of groundwater monitoring observations.

• Item Type: Conference Proceedings
• Status: Published
• Refereed: Yes
• Glasgow Author(s) Enlighten ID: Low, Dr Marnie and Jones, Dr Wayne and Alexander, Dr Craig and Miller, Professor Claire and Radvanyi, Mr Peter
• Authors: Radvanyi, P., Miller, C., Alexander, C., Low, M., Jones, W. R., and Rock, L.
• College/School: College of Science and Engineering > School of Mathematics and Statistics > Statistics; College of Science and Engineering > School of Mathematics and Statistics
• ISBN: 9783947323425
• Copyright Holders: Copyright © 2023 The Authors
• First Published: First published in Proceedings of the 37th International Workshop on Statistical Modelling (IWSM): 332-338
• Publisher Policy: Reproduced with the permission of the Author

Related URLs

• Organisation