A novel approach for assessing regionally differentiated ocean contributions to Earth Energy Imbalance from GRACE(-FO) and multi-mission altimetry

Earth Energy Imbalance (EEI) results from a net positive radiative flux at the top of atmosphere. Over 90% of this excess energy is stored as heat in the ocean causing present day ocean heat content (OHC) change. This in turn leads to volumetric or steric expansion of the water column and sea level rise. Utilizing observed bias-corrected short- and long-wave energy fluxes from the CERES project, it is possible to estimate EEI at the top of atmosphere. However, bias corrections rely on reanalysis OHC, potentially resulting in overestimation of ocean heat uptake (OHU).

Combining GRACE(-FO) and altimetry observations and constructing global sea level budgets allows to derive (thermo-)steric sea level change and convert this to OHU; the latter is generally achieved considering a literatute-based ocean-mean expansion efficiency of 0.52 [W/m^2 / mm/yr]. Nonetheless, this approach is valid for global mean steric sea level change only and it is unclear to what extent one can use it for investigating regional OHU.

Here, we develop a novel approach for deriving global and regional observation based OHC and OHU, which consists of three steps. (1) Fitting mass and steric spatial patterns, so called fingerprints, to GRACE(-FO) and altimetry data in a joint least-squares inversion. (2) Projecting reanalysis OHC onto the same spatial patterns that we use to explain steric variability. (3) Rescaling reanalysis OHC based on the observed steric sea level changes and reconstruction of spatial maps of OHC. These can then be further analyzed in order to derive global and regional OHU.

Based on preliminary results for years 2005-01 till 2015-12, we find ~1.2 mm/yr (thermo-)steric sea level change. Global-mean OHU of 0.62 [W/m^2] can be derived from the literature expansion efficiency above, while we find 0.63 [W/m^2] from the novel rescaling approach and 0.87 [W/m^2] based on ORAS5 ocean model data only. Regionally analyzing these results regarding individual ocean contributions reveals that the ocean model seems to significantly overestimate the uptake of the Atlantic and Pacific oceans, while slightly underestimating the Indian ocean contribution.