Efforts to estimate past global mean temperature and latitudinal gradients must contend with spatial heterogeneity in sea surface temperatures (SSTs). Here, we use modern SSTs to show that the environments from which most paleoclimatic data are drawn, shallow epeiric seas and continental margins, are systematically offset from zonal mean temperatures. Epeiric seas are warmer and more seasonal than open‐ocean values from the same latitudes, while continental margins exhibit consistent and predictable deviations related to gyre circulation. Warm temperatures inferred from Paleozoic proxy data may largely reflect that these data derive almost entirely from epeiric seas. Moreover, pseudoproxy analysis using Paleogene sampling localities demonstrates how undersampling of the full range of dynamical environments associated with gyre circulation can generate spurious estimates of latitudinal temperature gradients. Recognition of these global patterns permits a predictive framework within which to more robustly interpret proxy data, improve Earth system models, and reconstruct ancient dynamic regimes.

中文翻译：

---

解释古代海表温度的动力学框架

估算过去全球平均温度和纬度梯度的努力必须与海表温度（SSTs）的空间异质性相抗衡。在这里，我们使用现代SST来表明，从中提取大多数古气候数据的环境，浅海和大陆边缘系统地偏离了纬向平均温度。与相同纬度的远洋海域相比，远洋海域更暖和更季节性，而大陆边缘显示出与回旋环流相关的一致且可预测的偏差。从古生代代用资料推断出的温暖温度可能在很大程度上反映了这些数据几乎完全来自上海。此外，使用古近纪采样地点进行的伪代理分析表明，与回旋环流相关的整个动态环境的欠采样如何导致对纬度温度梯度的虚假估计。对这些全球模式的认识允许建立一个预测框架，在该框架内可以更可靠地解释代理数据，改善地球系统模型并重建古老的动态制度。