AI 中文总结
该研究提出跨变量代理学习方法，基于M-JEPA在无降水监督下预训练，实现了优于ECMWF集合的季风降水预测，为大气表示迁移提供诊断框架。
AI 中文摘要
深大气对流决定了南亚季风的变率，但尝试直接从零膨胀、重尾的降水数据中学习其潜在世界模型会产生次优的预测表示。连续大气代理变量（如向外长波辐射OLR）能更连贯地表达这种对流组织。我们通过跨变量代理学习解决这一不匹配问题：M-JEPA（多尺度季风联合嵌入预测架构）在追踪移动对流系统的拉格朗日斑块上的五个连续代理场进行预训练，全程未使用降水监督。将得到的冻结表示通过共享解码器主干迁移到日降水预测，该主干包含并行概率分支与确定性分支。由于预训练期间完全未观测到降水，下游性能直接衡量潜在滚动中捕获的预测信息。采用含两个对照组的冻结主干探测框架（仅在降水上训练的相同架构、随机初始化的主干），将迁移效果归因于代理预训练：直接降水训练的CRPS误差高出36%（7.52 vs. 5.54 mm/day）。对阵51成员的欧洲中期天气预报中心（ECMWF）业务集合，该迁移模型使用1540万参数在单个消费级GPU上实现了统计显著的CRPS优势（6.81 vs. 6.89 mm/day）和更高的Brier技巧评分（+0.05 vs. -0.04），优势集中在强降水阈值和精细空间尺度；而集合在邻域技巧评分和点指标的确定性参考上仍保留优势。该研究提供了基于季节内动力学的有竞争力的季风降水预测模型，以及评估迁移大气表示的诊断框架。
英文摘要
Deep atmospheric convection governs South Asian monsoon variability, yet attempting to learn its latent world model directly from zero-inflated, heavy-tailed precipitation yields suboptimal predictive representations. Continuous atmospheric proxies, such as outgoing longwave radiation (OLR), express this convective organization far more coherently. We address this mismatch with \emph{cross-variable proxy learning}: M-JEPA, a multiscale Monsoon Joint-Embedding Predictive Architecture, is pretrained on five continuous proxy fields over Lagrangian patches tracking moving convective systems---without rainfall supervision at any point. The resulting frozen representation is transferred to daily precipitation forecasts through a shared decoder trunk featuring parallel probabilistic and deterministic branches. Because rainfall is strictly unobserved during pretraining, downstream skill directly measures the predictive information captured in the latent rollout. A frozen-backbone probing framework with two controls (an identical architecture trained on rainfall alone, and a randomly initialized backbone) attributes the transfer specifically to proxy pretraining: direct rainfall training exhibits $36\%$ higher CRPS error ($7.52$ vs.\ $5.54$\,mm/day). Against the 51-member operational ECMWF ensemble, the transferred model attains a statistically resolved CRPS advantage ($6.81$ vs.\ $6.89$\,mm/day) and higher Brier skill ($+0.05$ vs.\ $-0.04$) using $15.4$M parameters on a single consumer GPU, concentrated at heavy-rain thresholds and fine spatial scales, while the ensemble retains an advantage in neighborhood skill and deterministic references on point metrics. The result provides a competitive monsoon precipitation forecast grounded in intraseasonal dynamics and a diagnostic framework for evaluating transferred atmospheric representations.
Comments8 pages, 3 figures, plus supplementary material
