OPEN-DDNN: Global 0.25°x 0.25° Monthly Ocean Heat Content Dataset (1993-2023) from Remote Sensing Reconstruction

This product used a Densely Deep Neural Network approach to reconstruct a high-resolution (0.25°x 0.25°) ocean heat content for upper 2000m over five different depths (0-100m, 0-300m, 0-700m, 0-1500m, 0-2000m) at a global scale from 1993-2023. It combined multisource remote sensing sea surface temperature (SST), altimetry absolute dynamic topography (ADT), sea surface wind (SSW) field data, and utilizing Argo-grid and EN4-profile data for training. The new 0.25°× 0.25° reconstruction effectively captures detailed thermal variations in regions with complex dynamics such as the Gulf Stream and Kuroshio Extension, surpassing traditional methods in resolution and accuracy. This dataset estimates the trend of ocean warming from 1993 to 2023 with higher resolution, revealing an intensifying trend of ocean warming over the past decade.