30 Aug 2023
 | 30 Aug 2023
Status: this preprint is currently under review for the journal HESS.

On the challenges of global entity-aware deep learning models for groundwater level prediction

Benedikt Heudorfer, Tanja Liesch, and Stefan Broda

Abstract. The application of machine learning (ML) including deep learning models in hydrogeology to model and predict groundwater level in monitoring wells has gained some traction in recent years. By now, the dominant model class is so called single-well models, where one model is trained for each well separately. However, recent developments in neighbouring disciplines including hydrology (rainfall-runoff-modelling) have shown that global models, being able to incorporate data of several wells, may have advantages. These models are often called entity-aware models, as they usually rely on static data to differentiate the entities, i.e. groundwater wells in hydrogeology or catchments in surface hydrology. We test two kinds of static information to characterize the groundwater wells in a global, entity-aware deep learning model setup, first, environmental features that are continuously available and thus theoretically allow spatial generalization (regionalization), and second, timeseries features that are derived from the past time series at the respective well. Moreover, we test random integer features as entity information for comparison. We use a published dataset of 108 groundwater wells in Germany, and evaluate the models’ performances in terms of Nash-Sutcliffe efficiency (NSE) in an in-sample and an out-of-sample setting, representing temporal and spatial generalization. Our results show, that entity-aware models work well with a mean performance of NSE > 0.8 in an in-sample setting, thus being comparable to, or even outperforming single-well models. However, they do not generalize well spatially in an out-of-sample setting (mean NSE < 0.7, i.e. lower than a global model without entity information). The reason for this potentially lies in the small number of wells in the dataset, which might not be enough to take full advantage of global models. However, also more research is needed to find meaningful static features for ML in hydrogeology.

Benedikt Heudorfer et al.

Status: open (until 01 Nov 2023)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on hess-2023-192', Anonymous Referee #1, 15 Sep 2023 reply

Benedikt Heudorfer et al.

Benedikt Heudorfer et al.


Total article views: 366 (including HTML, PDF, and XML)
HTML PDF XML Total Supplement BibTeX EndNote
293 64 9 366 15 3 4
  • HTML: 293
  • PDF: 64
  • XML: 9
  • Total: 366
  • Supplement: 15
  • BibTeX: 3
  • EndNote: 4
Views and downloads (calculated since 30 Aug 2023)
Cumulative views and downloads (calculated since 30 Aug 2023)

Viewed (geographical distribution)

Total article views: 347 (including HTML, PDF, and XML) Thereof 347 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
Latest update: 26 Sep 2023
Short summary
We build a neural network to predict groundwater levels from monitoring wells. We want to predict all the wells at the same time, by learning the differences between wells with “static features”. Then it is an “entity aware global model”. We test different static features and find that the model doesn’t really use them to learn how exactly the wells are different, but only to uniquely identify them. So this model class isn’t actually entity aware, and we suggest next steps to make it so.