Challenges in applying machine learning models for hydrological inference
- Machine learning (ML) algorithms are being increasingly used in Earth and Environmental modeling studies owing to the ever-increasing availability of diverse data sets and computational resources as well as advancement in ML algorithms. Despite advances in their predictive accuracy, the usefulness of ML algorithms for inference remains elusive. In this study, we employ two popular ML algorithms, artificial neural networks and random forest, to analyze a large data set of flood events across Germany with the goals to analyze their predictive accuracy and their usability to provide insights to hydrologic system functioning. The results of the ML algorithms are contrasted against a parametric approach based on multiple linear regression. For analysis, we employ a model-agnostic framework named Permuted Feature Importance to derive the influence of models' predictors. This allows us to compare the results of different algorithms for the first time in the context of hydrology. Our main findings are that (1) the ML models achieve higherMachine learning (ML) algorithms are being increasingly used in Earth and Environmental modeling studies owing to the ever-increasing availability of diverse data sets and computational resources as well as advancement in ML algorithms. Despite advances in their predictive accuracy, the usefulness of ML algorithms for inference remains elusive. In this study, we employ two popular ML algorithms, artificial neural networks and random forest, to analyze a large data set of flood events across Germany with the goals to analyze their predictive accuracy and their usability to provide insights to hydrologic system functioning. The results of the ML algorithms are contrasted against a parametric approach based on multiple linear regression. For analysis, we employ a model-agnostic framework named Permuted Feature Importance to derive the influence of models' predictors. This allows us to compare the results of different algorithms for the first time in the context of hydrology. Our main findings are that (1) the ML models achieve higher prediction accuracy than linear regression, (2) the results reflect basic hydrological principles, but (3) further inference is hindered by the heterogeneity of results across algorithms. Thus, we conclude that the problem of equifinality as known from classical hydrological modeling also exists for ML and severely hampers its potential for inference. To account for the observed problems, we propose that when employing ML for inference, this should be made by using multiple algorithms and multiple methods, of which the latter should be embedded in a cross-validation routine.…
Author details: | Lennart SchmidtORCiD, Falk HesseORCiDGND, Sabine AttingerORCiDGND, Rohini KumarORCiDGND |
---|---|
DOI: | https://doi.org/10.1029/2019WR025924 |
ISSN: | 0043-1397 |
ISSN: | 1944-7973 |
Title of parent work (English): | Water resources research |
Subtitle (German): | a case study for flooding events across Germany |
Publisher: | American Geophysical Union |
Place of publishing: | Washington |
Publication type: | Article |
Language: | English |
Date of first publication: | 2020/04/23 |
Publication year: | 2020 |
Release date: | 2023/05/04 |
Tag: | floods; inference; machine learning |
Volume: | 56 |
Issue: | 5 |
Article number: | e2019WR025924 |
Number of pages: | 10 |
Funding institution: | Helmholtz AssociationHelmholtz Association; Deutsche; ForschungsgemeinschaftGerman Research Foundation (DFG) [HE-7028-1/2] |
Organizational units: | Mathematisch-Naturwissenschaftliche Fakultät / Institut für Geowissenschaften |
DDC classification: | 5 Naturwissenschaften und Mathematik / 55 Geowissenschaften, Geologie / 550 Geowissenschaften |
Peer review: | Referiert |
Publishing method: | Open Access / Hybrid Open-Access |
License (German): | CC-BY - Namensnennung 4.0 International |