arXiv：cs.LG（机器学习，全量分类）· Gabriela Martinez Balbontin, Anastase Charantonis, Dominique Bereziat, Stefano Ciavatta·· 15 小时前AI 评分39
BG4Sea：基于渐进式信息缩放的海水生物地球化学季节性可预报性
BG4Sea: Biogeochemical Seasonal Forecastability via Progressive Information Scaling
AI 导读
BG4Sea 是首个全球性、数据驱动的海洋生物地球化学状态多变量季节性预报系统，采用列自编码器、潜在预报器、FiLM 表面强迫调节器和交叉注意力水平耦合模块的模块化架构。模型在 BIORYS4（NEMO/PISCES）全球海洋再分析数据上训练评估，可生成 1/4 度、月分辨率的六个月预报，覆盖溶解化学、生物和碳库变量，在多数变量和预报时效上优于持续性预报和气候态基准。
正文
Abstract:Marine biogeochemical forecasting is increasingly important for managing marine ecosystems and the carbon cycle, yet global, seasonal forecast products lag far behind physical oceanography, held back by the complexity of the processes involved and by data scarcity. We introduce BG4Sea, which to our knowledge is the first global, data-driven system to produce multivariate seasonal forecasts of the marine biogeochemical state. BG4Sea is a modular architecture with a column autoencoder that compresses the vertical column into a low-dimensional latent space, a latent forecaster propagates this representation forward in time, a surface-forcing conditioner that injects physical boundary information via Feature-wise Linear Modulation (FiLM), and a horizontal-coupling module that incorporates neighboring-column context through cross-attention. The model is trained and evaluated on the global ocean reanalysis BIORYS4 (NEMO/PISCES), and produces six-month forecasts at 1/4 degree, monthly resolution for dissolved chemistry, biology, and carbon-pool variables, outperforming persistence and climatology across most variables and lead times. We position BG4Sea as an interpretable baseline for future, more expressive approaches, and discuss predictability attribution to each component, alongside the model's structural limitations.
| Subjects: | Machine Learning (cs.LG) |
| Cite as: | arXiv:2607.16731 [cs.LG] |
| (or arXiv:2607.16731v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2607.16731 arXiv-issued DOI via DataCite |
Submission history
From: Gabriela Martinez Balbontin [view email]
[v1]
Sat, 18 Jul 2026 09:43:36 UTC (34,670 KB)
[v2]
Wed, 30 Sep 2026 19:34:35 UTC (48,012 KB)
来源：arXiv：cs.LG（机器学习，全量分类） · arxiv.org