SIFusion: A Unified Fusion Framework for Multi-granularity Arctic Sea Ice Forecasting

Arctic sea ice performs a vital role in global climate and has paramount impacts on both polar ecosystems and coastal communities. In the last few years, multiple deep learning based pan-Arctic sea ice concentration (SIC) forecasting methods have emerged and showcased superior performance over physics-based dynamical models. However, previous methods forecast SIC at a fixed temporal granularity, e.g. sub-seasonal or seasonal, thus only leveraging intra-granularity information and overlooking the plentiful inter-granularity correlations. Specifically, inter-granularity correlations mean that SIC at various temporal granularities exhibits cumulative effects and are naturally consistent, with short-term fluctuations potentially impacting long-term trends and long-term trends provide effective hints for facilitating short-term forecasts in Arctic sea ice. Therefore, in this study, we pro-pose to cultivate temporal multi-granularity that naturally derived from Arctic sea ice reanalysis data and provide a unified perspective for modeling SIC via our S ea I ce Fusion framework. SIFusion is delicately designed to leverage intra-granularity and inter-granularity information to capture granularity-consistent representations that promote forecasting skills. Our extensive experiments indicate that SIFusion outperforms off-the-shelf fixed temporal granularity SIC forecasting deep learning models for their specific temporal granularity.