Barystatic sea level change observed by satellite gravimetry: 1993-2022

Abstract

Using time-variable gravity fields determined from satellite laser ranging (SLR) and the Gravity Recovery and Climate Experiment (GRACE) and GRACE Follow-On (GRACE-FO), we derive three decades of barystatic sea level (BSL) change estimates since 1993, providing a long record of satellite gravimetry-based BSL estimates comparable in duration to satellite altimetry. Over the 1993-2022 altimetry era, the BSL change rate is 1.75 ± 0.59 mm/y (2σ) as observed by SLR. The rate increases to 2.16 ± 0.59 mm/y (SLR) and 2.13 ± 0.38 mm/y (GRACE/-FO) over 2003-2022, mainly due to accelerated ice loss in Greenland since the 2000s. The remarkable agreement between SLR and GRACE/-FO estimates over the joint period cross-validates each other, suggesting that both work well in monitoring BSL changes. We also compare our satellite gravimetry estimates with the sum of individual mass change components (ice sheets, glaciers, and land water storage) obtained from multisource datasets (in situ, remote sensing and geophysical modeling). The differences are within 0.2 mm/y over the three decades, highlighting the global consistency of different techniques for observing Earth's surface mass changes. Moreover, we reconcile the global mean sea level (GMSL) rise budget using reprocessed altimetry data and updated thermosteric sea level ensembles. From 1993 to 2022, the sum of thermosteric and SLR-based barystatic contributions (3.16 ± 0.64 mm/y) agrees well with altimetry-observed GMSL rate (3.22 ± 0.28 mm/y), suggesting that the GMSL rise budget can be closed within uncertainties over the last three decades.

Keywords: barystatic sea level change; climate change; satellite gravimetry; sea level budget.

Fig. 1.

(A) Time series of GMSL change from satellite altimetry, thermosteric sea level change from in situ measurements, and BSL change from satellite gravimetry (SLR and GRACE/-FO). (B) Comparison between thermosteric-corrected altimetry GMSL and BSL change derived from satellite gravimetry. (C) GMSL budget residuals (i.e., GMSL minus total contributions) with BSL estimates from SLR and GRACE/-FO. Seasonal signals are removed from the time series. The curves are further smoothed with a 5-mo moving average filter and are offset for better visibility.

Fig. 2.

(A) Altimetry-derived GMSL change and uncertainty (2σ) (shaded area). (B) Thermosteric sea level change and uncertainty (2σ) (shaded area). The GMSL is from GSFC, with uncertainty calculated as the RMS of the differences between the AVISO+, CU, and NOAA GMSL estimates relative to GSFC. The thermosteric sea level is the ensemble mean of eight in situ ocean temperature datasets, with the uncertainty represented by their spread. Seasonal signals and the 2005–2015 baseline means are removed in both (A) and (B).

Fig. 3.

GMSL budget assessments for the periods 1993–2022 (A) and 2003–2022 (B). BSL change rates are derived from individual surface mass change estimates (Model), SLR gravity fields, and GRACE/-FO gravity fields (2003-2022 only). Uncertainties are twice the SD (2$\sigma$)

Fig. 4.

GMSL budget residuals (i.e., GMSL minus total contributions) using different altimetry solutions (GSFC, CU, AVISO+, and NOAA) with BSL estimates from GRACE/-FO (A) and SLR (B). The curves are smoothed with a 5-mo moving average filter.