IceBench-S2S: A Benchmark of Deep Learning for Challenging Subseasonal-to-Seasonal Daily Arctic Sea Ice Forecasting in Deep Latent Space

2026-01-31

AI Summary

Current deep learning models struggle to extend Arctic sea ice forecasting beyond subseasonal-to-seasonal (S2S) timescales at daily resolution, limiting their utility in polar shipping and scientific research. This work proposes IceBench-S2S, the first deep learning benchmark specifically designed for 180-day continuous sea ice concentration prediction. By compressing spatial sea ice features into a deep latent space and incorporating a temporal concatenation strategy, the framework enables end-to-end S2S modeling. It integrates multiple backbone architectures within a unified pipeline for data processing, training, and evaluation, addressing the lack of a systematic benchmark for daily-resolution S2S sea ice forecasting. Experiments demonstrate that IceBench-S2S substantially improves forecast accuracy over six months, offering a standardized platform for model selection and optimization, and advancing the practical deployment of AI in polar climate prediction.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Computer Vision: Remote Sensing / Geospatial AI
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Large language models for search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Arctic sea ice plays a critical role in regulating Earth's climate system, significantly influencing polar ecological stability and human activities in coastal regions. Recent advances in artificial intelligence have facilitated the development of skillful pan-Arctic sea ice forecasting systems, where data-driven approaches showcase tremendous potential to outperform conventional physics-based numerical models in terms of accuracy, computational efficiency and forecasting lead times. Despite the latest progress made by deep learning (DL) forecasting models, most of their skillful forecasting lead times are confined to daily subseasonal scale and monthly averaged values for up to six months, which drastically hinders their deployment for real-world applications, e.g., maritime routine planning for Arctic transportation and scientific investigation. Extending daily forecasts from subseasonal to seasonal (S2S) scale is scientifically crucial for operational applications. To bridge the gap between the forecasting lead time of current DL models and the significant daily S2S scale, we introduce IceBench-S2S, the first comprehensive benchmark for evaluating DL approaches in mitigating the challenge of forecasting Arctic sea ice concentration in successive 180-day periods. It proposes a generalized framework that first compresses spatial features of daily sea ice data into a deep latent space. The temporally concatenated deep features are subsequently modeled by DL-based forecasting backbones to predict the sea ice variation at S2S scale. IceBench-S2S provides a unified training and evaluation pipeline for different backbones, along with practical guidance for model selection in polar environmental monitoring tasks.

Problem

• Arctic sea ice forecasting
• subseasonal-to-seasonal (S2S)
• deep learning
• forecasting lead time
• daily prediction

Innovation

• subseasonal-to-seasonal forecasting
• deep latent space
• Arctic sea ice concentration
• deep learning benchmark
• spatiotemporal modeling