Papers

研究检验平流扩散先验在概率化海表温度预测中的实际作用

Where Should Physics Enter? Information, Identifiability and Calibration in Probabilistic Sea Surface Temperature Forecasting

Why we picked it

这篇论文用合成数据加真实数据双重对照,把"物理先验放哪"这个常见做法拆开验证了。结论挺反直觉:学出来的速度场根本没用,洋流当输入通道就够,校准修正才是提升 CRPS 的大头。做物理信息神经网络的人应该看看,避免踩同样的坑。

一项研究在孟加拉湾 NOAA OISST v2.1 数据(1982-2024)和同网格合成基准上,测试平流扩散先验应注入异方差 ConvLSTM 的哪个位置。结果显示:带学习速度场的软残差在 lead-15 上 CRPS 变化仅 -0.007°C(95% 区间 [-0.022, +0.008]),与简单时间平滑惩罚无可靠差异。学习到的速度场即使在合成数据中也无法恢复已知洋流,相关性与打乱日期的零假设相当,而正确利用洋流本可提升 day-15 预测 16-23%。

Body

将洋流作为输入通道提供(不做传输步)即可恢复传输设计 92% 的 lead-5 增益;但用 GLORYS12 再分析洋流在真实数据上反而使 lead-1 和 lead-5 差 6% 和 4%,因为再分析平流与 OISST 日变化相关性仅 0.02。真正的关键缺陷是各模型共有的随预测时长变化的校准偏差:按 lead 拟合一个方差乘子即可将 90% 覆盖率恢复到 0.85-0.91,并将 day-1 CRPS 降低 13-21%。作者撤回了此前基于不受控比较的 81% CRPS 降低结论。

Original · arXiv cs.LG

Where Should Physics Enter? Information, Identifiability and Calibration in Probabilistic Sea Surface Temperature Forecasting

Advection-diffusion priors are a common way to make neural sea surface temperature (SST) forecasts physics-informed. We test where such a prior must enter a heteroscedastic ConvLSTM to help, using daily NOAA OISST v2.1 for the Bay of Bengal (1982-2024) and a synthetic benchmark on the same grid with known currents of controlled strength. (i) On real data, a soft advection-diffusion residual with a learned velocity gives no reliable benefit (lead-15 CRPS -0.007 degrees C, 95% interval [-0.022, +0.008]); a plain temporal-smoothness penalty matches it, and equivalence within 3% is not established either. (ii) Learned velocities do not recover the currents, whether placed in the loss, in an explicit transport step, or in a transport step without neural correction: they correlate with the reference no better than a shuffled-date null, even on synthetic data where knowing the currents would improve day-15 forecasts by 16-23%. (iii) Supplied currents help on synthetic data, but mostly as information: giving them to the network as input channels, without transport, recovers 92% of the transport design's lead-5 gain, and a frozen-current numerical forecast does as well as either. (iv) With GLORYS12 reanalysis currents on real data, the transport design is 6% and 4% worse at leads 1 and 5; reanalysis advection explains essentially none of the daily OISST change (correlation 0.02), and frozen-current transport is worse than persistence. (v) The defect that matters is lead-dependent miscalibration shared by every model. One variance multiplier per lead, fitted on validation data, brings 90% coverage to 0.85-0.91 at all leads and lowers day-1 CRPS by 13-21%; fine-tuning on 15-day rollouts fixes long-lead coverage but costs short-lead skill. This version withdraws an earlier claim of an 81% CRPS reduction that came from an uncontrolled comparison.

Industry Coordinates & Source Verification

This report is categorized under AITOP's "Papers" desk. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include ConvLSTM, physics-informed neural networks, CRPS, NOAA.

Source attribution: curated from arXiv cs.LG (@Parmitha Vangapandu, Jonnakuti Pavan Kumar, Sathwik Narkedimilli, MSVPJ Sathvik, T. M. Balakrishnan Nair, Santanu Bhattacharya).