Sun et al. (2026) BiGRU-dXAJ: A hybrid hydrologic model enabling differentiable parameter learning for runoff simulation

Identification

• Journal: Environmental Modelling & Software
• Year: 2026
• Date: 2026-08-06
• Authors: Baowei Yan, Jianming Chen, Shuairu Xu, Jialang Chen, Jianbo Chang, Xuerui Zhou
• DOI: 10.1016/j.envsoft.2026.107136

Research Groups

Not explicitly stated in the provided text.

Short Summary

This paper introduces BiGRU-dXAJ, a novel hybrid hydrologic model that integrates deep learning with a conceptual model, enabling differentiable parameter learning for enhanced runoff simulation.

Objective

• To develop and evaluate BiGRU-dXAJ, a hybrid hydrologic model that combines Bidirectional Gated Recurrent Units (BiGRU) with a conceptual hydrologic model (dXAJ) to facilitate differentiable parameter learning for improved runoff simulation.

Study Configuration

• Spatial Scale: Not specified in the provided text.
• Temporal Scale: Not specified in the provided text.

Methodology and Data

• Models used: BiGRU-dXAJ, a hybrid hydrologic model integrating BiGRU (Bidirectional Gated Recurrent Unit) for deep learning and a conceptual hydrologic model (implied by "dXAJ", likely a variant of the Xinanjiang model). The methodology focuses on enabling differentiable parameter learning.
• Data sources: Not specified in the provided text.

Main Results

Specific quantitative results are not available in the provided pre-proof text. However, the paper aims to demonstrate that the BiGRU-dXAJ model successfully enables differentiable parameter learning and achieves improved performance in runoff simulation compared to traditional or standalone models.

Contributions

This article contributes a novel hybrid hydrologic modeling framework, BiGRU-dXAJ, which innovatively integrates deep learning capabilities (BiGRU) with a conceptual hydrologic model to enable differentiable parameter learning. This approach offers a significant advancement in hydrologic modeling by potentially improving the accuracy and efficiency of runoff simulation through a more robust parameter optimization process.

Funding

Not specified in the provided text.

Citation

@article{Sun2026BiGRUdXAJ,
  author = {Sun, Mingbo and Yan, Baowei and Chen, Jianming and Xu, Shuairu and Chen, Jialang and Chang, Jianbo and Zhou, Xuerui},
  title = {BiGRU-dXAJ: A hybrid hydrologic model enabling differentiable parameter learning for runoff simulation},
  journal = {Environmental Modelling & Software},
  year = {2026},
  doi = {10.1016/j.envsoft.2026.107136},
  url = {https://doi.org/10.1016/j.envsoft.2026.107136}
}