Evaluating twenty-year trends in earth's energy flows from observations and reanalyses

• Published in: Journal of Geophysical Research: Atmospheres
• Other Authors: Loeb, Norman G. (author), Mayer, Michael (author), Kato, Seiji (author), Fasullo, John T. (author), Zuo, Hao (author), Senan, Retish (author), Lyman, John M. (author), Johnson, Gregory C. (author), Balmaseda, Magdalena (author)
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2022
• Subjects: Pacific; Sea ice
• Online Access: https://doi.org/10.1029/2022JD036686

Satellite, reanalysis, and ocean in situ data are analyzed to evaluate regional, hemispheric and global mean trends in Earth's energy fluxes during the first 20 years of the twenty-first century. Regional trends in net top-of-atmosphere (TOA) radiation from the Clouds and the Earth's Radiant Energy System (CERES), ECMWF Reanalysis 5 (ERA5), and a model similar to ERA5 with prescribed sea surface temperature (SST) and sea ice differ markedly, particularly over the Eastern Pacific Ocean, where CERES observes large positive trends. Hemispheric and global mean net TOA flux trends for the two reanalyses are smaller than CERES, and their climatological means are half those of CERES in the southern hemisphere (SH) and more than nine times larger in the northern hemisphere (NH). The regional trend pattern of the divergence of total atmospheric energy transport (TEDIV) over ocean determined using ERA5 analyzed fields is similar to that inferred from the difference between TOA and surface fluxes from ERA5 short-term forecasts. There is also agreement in the trend pattern over ocean for surface fluxes inferred as a residual between CERES net TOA flux and ERA5 analysis TEDIV and surface fluxes obtained directly from ERA5 forecasts. Robust trends are observed over the Gulf Stream associated with enhanced surface-to-atmosphere transfer of heat. Within the ocean, larger trends in ocean heating rate are found in the NH than the SH after 2005, but the magnitude of the trend varies greatly among datasets. 1947282