Utilizing Earth Foundation Models to Enhance the Simulation Performance of Hydrological Models with AlphaEarth Embeddings

2026-01-04

arXiv.org

AI Summary

This study addresses the limitations of traditional hydrological models in ungauged basins, where reliance on manually designed static attributes fails to adequately capture complex environmental heterogeneity, thereby constraining predictive accuracy. To overcome this, the authors propose a novel data-driven approach that leverages embedding vectors derived from AlphaEarth—a foundation model trained on vast satellite imagery—to represent the integrated dynamic environmental characteristics of watersheds. By selecting donor basins for transfer learning based on embedding similarity, the method significantly improves streamflow prediction accuracy in basins excluded from model training. The results demonstrate the superiority of these learned embeddings over conventional static attributes in both representing watershed characteristics and guiding effective knowledge transfer, thereby transcending the constraints of traditional watershed feature representation.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Data Mining & Knowledge Management: Data Stream Mining
• Knowledge Representation and Reasoning: Knowledge Acquisition

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Predicting river flow in places without streamflow records is challenging because basins respond differently to climate, terrain, vegetation, and soils. Traditional basin attributes describe some of these differences, but they cannot fully represent the complexity of natural environments. This study examines whether AlphaEarth Foundation embeddings, which are learned from large collections of satellite images rather than designed by experts, offer a more informative way to describe basin characteristics. These embeddings summarize patterns in vegetation, land surface properties, and long-term environmental dynamics. We find that models using them achieve higher accuracy when predicting flows in basins not used for training, suggesting that they capture key physical differences more effectively than traditional attributes. We further investigate how selecting appropriate donor basins influences prediction in ungauged regions. Similarity based on the embeddings helps identify basins with comparable environmental and hydrological behavior, improving performance, whereas adding many dissimilar basins can reduce accuracy. The results show that satellite-informed environmental representations can strengthen hydrological forecasting and support the development of models that adapt more easily to different landscapes.

Problem

• ungauged basins
• hydrological modeling
• river flow prediction
• basin similarity
• environmental representation

Innovation

• Foundation Models
• AlphaEarth Embeddings
• Hydrological Modeling
• Ungauged Basins
• Satellite-based Environmental Representation