Preprints
https://doi.org/10.5194/egusphere-2026-6870
https://doi.org/10.5194/egusphere-2026-6870
20 Nov 2026
 | 20 Nov 2026
Status: this preprint is open for discussion and under review for Geoscientific Model Development (GMD).

dMoT v1.0: A Differentiable Library of 36 Conceptual Hydrological Models for Large-Sample Modelling

Xin Jing, Jungang Luo, Mingxiang Yang, Yaoxuan Xiao, and Xue Yang

Abstract. Conceptual hydrological models represent alternative hypotheses of catchment behaviour, yet their diverse numerical implementations complicate gradient-based, large-sample modelling. We present dMoT v1.0, a PyTorch-based library implementing 36 established lumped rainfall-runoff models within a common differentiable framework. Drawing on MARRMoT formulations, dMoT retains model-specific storages, fluxes, and routing processes while adopting explicit sequential state updates and GPU-parallel computation. We evaluate numerical reliability, automatic differentiation, computational efficiency, and scalability. Water-balance tests met their prescribed tolerances, and automatic-differentiation gradients agreed with finite-difference estimates under the evaluated conditions. Compiled GPU execution achieved a median 2.89-fold forward-execution speed-up over single-threaded CPU execution across the 36 models. We further demonstrate independent calibration and differentiable parameter learning across 531 CAMELS-US catchments. The two approaches achieved median model-level test scores of 0.624 and 0.612, respectively, using a numerically stabilised Kling-Gupta efficiency metric on temporally held-out data from the training catchments. These results establish dMoT as a computational framework for applying contrasting conceptual model structures under different parameter-estimation strategies. By combining a documented model library with automatic differentiation, numerical verification, and scalable execution, dMoT supports reproducible large-sample hydrological modelling and systematic comparisons of model structures and learning approaches.

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
Xin Jing, Jungang Luo, Mingxiang Yang, Yaoxuan Xiao, and Xue Yang

Status: open (until 15 Jan 2027)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
Xin Jing, Jungang Luo, Mingxiang Yang, Yaoxuan Xiao, and Xue Yang
Xin Jing, Jungang Luo, Mingxiang Yang, Yaoxuan Xiao, and Xue Yang
Metrics will be available soon.
Latest update: 20 Nov 2026
Download
Short summary
Rivers are studied using many computer models, but comparing them is often difficult. We brought 36 established models into one software library, checked their calculations, and measured how quickly they run. Tests across 531 river basins in the United States showed broadly similar overall results when model settings were chosen separately for each basin or learned from its landscape characteristics. The library makes it easier to compare models and study river flow across many places.
Share