HydroStartML: A combined machine learning and physics-based approach to reduce hydrological model spin-up time

2025-04-24

AI Summary

Hydrological models require prolonged iterative spin-up to determine the initial depth to water table (DTWT), incurring high computational costs and convergence difficulties. To address this, we propose a physics-informed, data-driven machine learning surrogate model—the first framework capable of robust, generalizable spatial prediction of steady-state DTWT. Our approach employs an ensemble random forest trained on physically constrained features—including hydraulic conductivity and surface slope—to enable reliable extrapolation to unseen topographies. Compared to conventional multi-year spin-up methods, our framework reduces spin-up duration by over 70% and cuts convergence iterations by more than 50%, while yielding predictions that closely approximate the physical steady-state solution. This advancement significantly enhances computational efficiency and scalability of watershed-scale hydrological simulations.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Data Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal Data
• Humans and AI: Human-in-the-loop Machine Learning

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Web Mining and Content Analysis: Models for Web evolution
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Finding the initial depth-to-water table (DTWT) configuration of a catchment is a critical challenge when simulating the hydrological cycle with integrated models, significantly impacting simulation outcomes. Traditionally, this involves iterative spin-up computations, where the model runs under constant atmospheric settings until steady-state is achieved. These so-called model spin-ups are computationally expensive, often requiring many years of simulated time, particularly when the initial DTWT configuration is far from steady state. To accelerate the model spin-up process we developed HydroStartML, a machine learning emulator trained on steady-state DTWT configurations across the contiguous United States. HydroStartML predicts, based on available data like conductivity and surface slopes, a DTWT configuration of the respective watershed, which can be used as an initial DTWT. Our results show that initializing spin-up computations with HydroStartML predictions leads to faster convergence than with other initial configurations like spatially constant DTWTs. The emulator accurately predicts configurations close to steady state, even for terrain configurations not seen in training, and allows especially significant reductions in computational spin-up effort in regions with deep DTWTs. This work opens the door for hybrid approaches that blend machine learning and traditional simulation, enhancing predictive accuracy and efficiency in hydrology for improving water resource management and understanding complex environmental interactions.

Problem

• Reducing hydrological model spin-up time using ML
• Predicting initial water table depth for faster convergence
• Combining ML and physics to improve simulation efficiency

Innovation

• Machine learning emulator predicts steady-state DTWT
• Hybrid approach combines ML and physics-based modeling
• Reduces computational spin-up time significantly