AIFL: A Global Daily Streamflow Forecasting Model Using Deterministic LSTM Pre-trained on ERA5-Land and Fine-tuned on IFS

2026-02-18

AI Summary

This study addresses the performance degradation of data-driven models when transferring from reanalysis data to operational numerical weather prediction by proposing a two-stage transfer learning strategy. First, a deterministic LSTM model is pretrained on ERA5-Land (1980–2019), then fine-tuned on IFS control forecasts (2016–2019) to align with their error structure. This approach enables the first end-to-end global daily streamflow forecasting within the CARAVAN ecosystem, effectively bridging the domain gap between reanalysis and operational forecasts. Evaluated on an independent test set (2021–2024), the model achieves a median KGE' of 0.66 and NSE of 0.53, demonstrates reliable skill in capturing extreme events, and matches the performance of current state-of-the-art global systems—all within a transparent and reproducible framework.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Data Mining & Knowledge Management: Data Stream Mining
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Systems and Infrastructure for Web, Mobile and WoT: Data management and stream processing for Web, mobile and wireless applications
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

Reliable global streamflow forecasting is essential for flood preparedness and water resource management, yet data-driven models often suffer from a performance gap when transitioning from historical reanalysis to operational forecast products. This paper introduces AIFL (Artificial Intelligence for Floods), a deterministic LSTM-based model designed for global daily streamflow forecasting. Trained on 18,588 basins curated from the CARAVAN dataset, AIFL utilises a novel two-stage training strategy to bridge the reanalysis-to-forecast domain shift. The model is first pre-trained on 40 years of ERA5-Land reanalysis (1980-2019) to capture robust hydrological processes, then fine-tuned on operational Integrated Forecasting System (IFS) control forecasts (2016-2019) to adapt to the specific error structures and biases of operational numerical weather prediction. To our knowledge, this is the first global model trained end-to-end within the CARAVAN ecosystem. On an independent temporal test set (2021-2024), AIFL achieves high predictive skill with a median modified Kling-Gupta Efficiency (KGE') of 0.66 and a median Nash-Sutcliffe Efficiency (NSE) of 0.53. Benchmarking results show that AIFL is highly competitive with current state-of-the-art global systems, achieving comparable accuracy while maintaining a transparent and reproducible forcing pipeline. The model demonstrates exceptional reliability in extreme-event detection, providing a streamlined and operationally robust baseline for the global hydrological community.

Problem

• streamflow forecasting
• domain shift
• global hydrological modeling
• flood preparedness
• data-driven models

Innovation

• two-stage training
• deterministic LSTM
• ERA5-Land pre-training
• IFS fine-tuning
• global streamflow forecasting