Global ocean latent and sensible heat fluxes from the maximum entropy production framework from 1988-2017

The global long-term ocean surface heat fluxes were revisited using an enhanced Maximum Entropy Production (MEP) framework. The ocean heat storage and four empirical Bowen ratio formulas were employed to refine the MEP method's estimation. Utilizing the improved MEP method alongside multi-source remote sensing observation data as the driving input, we generated a high-precision global estimate of ocean latent and sensible heat flux from 1988 to 2017. This new dataset achieved a coefficient of determination (R2) of 0.79 and a root mean square error (RMSE) of 16 W·m-2 when compared against 129 global distributed buoy stations. The newly derived latent heat flux dataset outperformed the existing mainstream heat fluxes products, including J-OFURO3, ERA5, MERRA2, and OAFlux.
The dataset files are all in NetCDF (.nc) format, with a monthly temporal resolution and a spatial resolution of 0.25-degree, covering the period from 1988 to 2017.

本研究采用增强型最大熵产生（Maximum Entropy Production, MEP）框架，对全球长期海洋表面热通量进行了重新反演。研究结合海洋热储量与四种经验鲍恩比公式，对MEP方法的估算过程进行了优化。本研究将改进后的MEP方法与多源遥感观测数据作为驱动输入，生成了1988年至2017年期间高精度的全球海洋潜热与感热通量估算结果。相较于全球129个分布式浮标站点的观测数据，该新数据集的决定系数（coefficient of determination, R²）为0.79，均方根误差（root mean square error, RMSE）为16 W·m⁻²。此次新生成的潜热通量数据集，性能优于当前主流的热通量产品，包括J-OFURO3、ERA5、MERRA2及OAFlux。
本数据集所有文件均采用NetCDF（.nc）格式，时间分辨率为月度，空间分辨率为0.25°，覆盖时段为1988年至2017年。