Preprints
https://doi.org/10.5194/egusphere-2025-4055
https://doi.org/10.5194/egusphere-2025-4055
06 Oct 2025
 | 06 Oct 2025
Status: this preprint is open for discussion and under review for Hydrology and Earth System Sciences (HESS).

Never Train a Deep Learning Model on a Single Well? Revisiting Training Strategies for Groundwater Level Prediction

Marc Ohmer and Tanja Liesch

Abstract. Deep learning (DL) models are increasingly used for hydrological forecasting, with a growing shift from site-specific to globally trained architectures. This study tests whether the widely held assumption that global models consistently outperform local ones also applies to groundwater systems, which differ substantially from surface water due to slow response dynamics, data scarcity, and strong site heterogeneity. Using a benchmark dataset of nearly 3000 monitoring wells across Germany, we systematically compare global Long Short-Term Memory (LSTM) models with locally trained single-well models in terms of overall performance, training data characteristics, prediction of extremes, and spatial generalization.

For groundwater level prediction, we find that global models provide no systematic accuracy advantage over local models. Local models more often capture site-specific behavior, while global models yield more robust but less specialized predictions across diverse wells. Performance gains arise primarily from dynamically coherent training data, whereas random data reduction has little effect, indicating that similarity matters more than quantity in this setting. Both model types struggle with extreme groundwater conditions, and global models generalize reliably only to wells with comparable dynamics.

These findings qualify the assumption of global model superiority and highlight the need to align modeling strategies with groundwater-specific constraints and application goals.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.
Share
Marc Ohmer and Tanja Liesch

Status: open (until 17 Nov 2025)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
Marc Ohmer and Tanja Liesch
Marc Ohmer and Tanja Liesch
Metrics will be available soon.
Latest update: 06 Oct 2025
Download
Short summary
We compared global vs. local deep learning models for groundwater level prediction using ~3,000 wells. Unlike surface water, groundwater is complex and data-scarce. Results: global models show no systematic accuracy advantage over local ones. Data similarity matters more than quantity for better predictions. Successful groundwater modeling requires strategies tailored to these unique complexities, not just larger datasets.
Share