Xu et al. (2026) Graph‐Based Automated Discovery of Concise Soil Hydraulic Functions From Data: Beyond Empirical Formulations
⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.
Identification
- Journal: Geophysical Research Letters
- Year: 2026
- Date: 2026-11-14
- Authors: Hao Xu, Sun Jinshen, Yuntian Chen, Dongxiao Zhang
- DOI: 10.1029/2026gl126105
Research Groups
Not explicitly mentioned in the abstract.
Short Summary
This study introduces a graph-based equation discovery framework to derive soil hydraulic functions from experimental data, yielding a novel model that outperforms the Mualem–van Genuchten model in predicting unsaturated hydraulic conductivity across diverse soil textures.
Objective
- To introduce and evaluate a graph-based equation discovery framework for distilling soil hydraulic functions from experimental data, aiming to improve predictive accuracy and parametric interpretability compared to existing expert-designed formulations like Mualem–van Genuchten.
Study Configuration
- Spatial Scale: Pore-to-plot scale (vadose zone), evaluated across diverse soil textures.
- Temporal Scale: Not explicitly defined for the discovery process, but the derived functions are for dynamic water flow and transport.
Methodology and Data
- Models used: Graph-based equation discovery framework; Mualem–van Genuchten (MvG) model and its common variants (for comparison).
- Data sources: Experimental data, specifically the original datasets that motivated the MvG parameterization.
Main Results
- A concise water retention function and an analytical unsaturated hydraulic conductivity expression were recovered, whose mathematical structure departs from classical empirical forms.
- The discovered model predicts unsaturated hydraulic conductivity more accurately than MvG and its common variants across soils of contrasting texture.
- The fitted parameters of the discovered model retain interpretable correlations with measurable soil physical properties.
Contributions
- Introduction of a novel graph-based equation discovery framework for deriving soil hydraulic functions from experimental data.
- Demonstration that data-driven discovery can yield compact, analytically tractable constitutive relations that move beyond fixed empirical formulations.
- Development of a new soil hydraulic model that offers improved predictive accuracy for unsaturated hydraulic conductivity compared to the Mualem–van Genuchten model and its variants.
- The discovered model's parameters exhibit interpretable correlations with soil physical properties, enhancing understanding and applicability.
Funding
Not mentioned in the abstract.
Citation
@article{Xu2026GraphBased,
author = {Xu, Hao and Jinshen, Sun and Chen, Yuntian and Zhang, Dongxiao},
title = {Graph‐Based Automated Discovery of Concise Soil Hydraulic Functions From Data: Beyond Empirical Formulations},
journal = {Geophysical Research Letters},
year = {2026},
doi = {10.1029/2026gl126105},
url = {https://doi.org/10.1029/2026gl126105}
}
Original Source: https://doi.org/10.1029/2026gl126105