A Deep State Space Model for Rainfall-Runoff Simulations

2025-01-24

AI Summary

To address insufficient modeling accuracy and cross-basin generalizability in rainfall–runoff simulation, this study proposes S4D-FT—the first frequency-tuned diagonalized state-space model tailored for hydrological modeling. S4D-FT integrates structural hydrological priors with explicit frequency-domain dynamics, enabling end-to-end rainfall–runoff mapping without reliance on traditional conceptual models (e.g., Sacramento) or standard LSTMs. Trained jointly across multiple basins and rigorously benchmarked against established methods, S4D-FT demonstrates significant improvements over LSTM on 531 U.S. basins: it achieves higher Nash–Sutcliffe Efficiency (NSE), reduced peak-flow error, and superior cross-basin generalization. This work pioneers the application of state-space models to hydrological time-series modeling and establishes a new paradigm for physics-informed deep learning in water resources science.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Intelligent Robots: State Estimation
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Large language models for search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

The classical way of studying the rainfall-runoff processes in the water cycle relies on conceptual or physically-based hydrologic models. Deep learning (DL) has recently emerged as an alternative and blossomed in hydrology community for rainfall-runoff simulations. However, the decades-old Long Short-Term Memory (LSTM) network remains the benchmark for this task, outperforming newer architectures like Transformers. In this work, we propose a State Space Model (SSM), specifically the Frequency Tuned Diagonal State Space Sequence (S4D-FT) model, for rainfall-runoff simulations. The proposed S4D-FT is benchmarked against the established LSTM and a physically-based Sacramento Soil Moisture Accounting model across 531 watersheds in the contiguous United States (CONUS). Results show that S4D-FT is able to outperform the LSTM model across diverse regions. Our pioneering introduction of the S4D-FT for rainfall-runoff simulations challenges the dominance of LSTM in the hydrology community and expands the arsenal of DL tools available for hydrological modeling.

Problem

• Rainfall-Runoff Modeling
• LSTM Improvement
• Hydrological Simulation

Innovation

• S4D-FT model
• hydrology simulation
• frequency modulation