Reconciling ocean mass change and GIA from satellite gravity and altimetry (OMCG-2)

OMCG-2 aims at resolving discrepancies in published estimates for ocean mass change (OMC), a major component of the sea level budget, and for providing new consolidated estimates at global and regional scale. Knowing OMC is important for deriving ocean heat content (OHC), and thus for improving our understanding of the Earth’s energy budget.

At the beginning of OMCG-1, published estimates of OMC, from essentially the same Gravity Recovery and Climate Experiment (GRACE) mission data, have disagreed significantly at the global scale. Estimates varied from 1.1 to more than 2.0 mm/a (with studies following the ‘Inverse’ approach at the lower end and those following the ‘Direct’ approach at the higher end), claiming errors and budget closure at the 0.1-0.2 mm/a level. This has posed an obvious problem for attempts for inferring global or even regional OHC change from altimetry. Many authors have claimed to achieve closure of the global sea level budget (GSLB), but the individual budget components have changed significantly between studies.

Satellite altimetry has proven that global mean sea level (GMSL) is rising at a rate of more than 3 mm/a since 1993, with an uncertainty of about 0.4 mm/a. Nerem et al. (2018) have recently suggested that, when removing interannual and decadal variability including sudden cooling phases due to volcanic eruptions, and considering potential instrument errors, GMSL is accelerating at 0.08 ± 0.025 mm/a2, broadly in agreement with climate predictions. It is thus of utmost importance to understand and quantify the factors that determine GMSL and thus enter the GSLB. The state of the art concerning the GSLB, at mid-2018, is described in a recent review paper by the WCRP (2018), which is meant to initiate a series of annual GSLB updates in between IPCC Assessment Reports. In this paper (Tab. 10), GRACE-derived OMC rates are reported between 1.1 and 2.1 mm/a (for differing time spans), based on publications from 2013 and newer, and mostly using the Paulson et al. (2007) or A et al. (2013) models of glacial isostatic adjustment (GIA). Both PIs of this proposal have contributed to this paper. However, the work in OMCG-1 (Uebbing et al., 2019) led to a reassessment and consolidation of these rates. It is also worth noting that the WCRP (2018) study contains inconsistencies between GIA corrections of the involved GRACE-based OMC estimates and GRACE-based ice sheet mass change estimates, in particular for Antarctica.

Figure 1: Global mean ocean mass time series from direct and indirect approaches. Left: Reconstruction of published direct and inverse OMC estimates. Right: OMC estimates from consistently processing the GRACE data within the direct and inverse methods. Dashed lines represent the total signal and solid lines are filtered with a 13 month boxcar filter. The trend estimates are derived from the time period Jun 2002 until May 2014, indicated by the gray background (based on Uebbing et al., 2019).

The central hypothesis of this proposal for a second phase (OMCG-2), unchanged with respect to OMCG-1, is that OMC discrepancies (Figure 1, left) are mostly related to

  1. method issues in the analysis of the GRACE data and
  2. the unsolved problem of correcting for the effect of GIA (Figure 2).

We believe that they can be resolved through (a) developing a unified methods framework, (b) devising and conducting realistic simulations, and (c) including GIA consistently in inverse GRACE-Altimetry schemes.

Figure 2: (A) Antarctic GIA fingerprint according to forward modeling based on the ICE-5G glacial history. (B) Result from the global inversion from adjusting the amplitude of the forward-modelled pattern. (C) Empirical GIA solution from the combination of GRACE, ice altimetry and firn modeling over the interval Apr 2002 until Aug 2016.

The OMCG project has two key objectives,

  1. Resolve conflicting mass/steric sea level budget partitioning. This is most important for the South-East Asian Island States region with current rates of up to 12 mm/a.
  2. Provide consolidated and consistent gridded OMC to the SPP and to the community. Address harmonizing the GRACE/GRACE-FO OMC estimates by available bridging solutions.

With these objectives, we believe the project to be instrumental for any integrated analysis of climate-related sea-level change at global to regional scale. It supports the central objective of the SPP and provides key data sets for many other SPP-projects.

To reach our objectives within the project, we will work on the following main challenges:

  • Express the 'Direct' and 'Inverse' approaches in a joint theoretical framework to understand their differences and their equivalences.
  • Investigate 'Direct' and 'Inverse' approaches in a joint comprehensive consistent simulation.
  • Resolve GIA together with OMC and the ice sheet contribution to OMC by integrating satellite altimetry over the ice sheets as an additional observable. This will reduce the GIA-related uncertainty of present-day OMC. The implied improvement of ice sheet change estimates will aid the understanding of ice sheet mass balance and ice-ocean-solid Earth interaction and will thereby aid OMC predictions.
  • Assess real-data results, disseminate them within and beyond the SPP.

The Bonn and Dresden groups will collaborate to develop a common theoretical framework for assessing various approaches and the resulting performance for OMC. Further both groups will collaborate to integrate, at different levels, the information from ice altimetry together with firn modeling into the global inversion framework, in order to improve the estimation of GIA.

OMCG-2 will seek to directly quantify physical processes that contribute to sea-level change and trends at the global but also regional scale, including ice-ocean-land processes (GIA) and land-ocean processes (contribution of terrestrial water storage change). OMCG-2 will thus provide important data sets for modelling projects inside and outside the SPP. In addition, OMCG-2 will provide regionalized data for Northern Europe and South east Asia regions and this will be helpful for improving predictions and assessments of sea-level rise effects at regional scale.