Revisiting Sea-level and global water budgets for the period 1993–2022 within ESA Climate change initiative

Abstract. Sea‑level rise, driven by anthropogenic greenhouse‑gas forcing, is one of the best indicators of climate change. Satellite altimetry, the global Argo profiling network, and space‑borne gravimetry (GRACE/GRACE‑FO) have enabled quantitative monitoring of the sea‑level budget, yet recent analyses reveal a persistent non‑closure after 2015. We present an updated assessment of the global and regional sea‑level budget for the satellite era (1993–2022), extending the previous ESA‑CCI evaluation with refined uncertainty characterisation and an objective inverse closure framework. Global-mean sea-level (GMSL) trend and associated uncertainties at 90 % confidence level are accelerating from 3.39 ± 0.20 mm yr^-1 for 1993–2022 to 3.86 ± 0.18 mm yr^-1 for 2004–2022. GMSL is driven primarily by land‑ice mass loss of 1.44 ± 0.09 mm yr^-1 and 1.74 ± 0.09 mm yr^-1, respectively, alongside the thermosteric contribution of 1.24 ± 0.14 mm yr^-1 and 1.32 ± 0.15 mm yr^-1, for the same periods. The budget closes robustly until 2015 with residuals under 0.3 mm yr^-1 (less than 10 % of the trend). Thereafter a statistically significant residual trend emerges, independent of the barystatic sea-level either from gravimetry or via the global water budget, indicating either a systematic inconsistency or multiple single‑dataset artifact.

Regional analysis identifies persistent non‑closure in the North Atlantic, around Australia, and to a lesser extent in the North Pacific—patterns that resemble the largest steric signals and suggest underestimation of deep‑ocean steric contributions or overestimation of altimetric sea-level. An objective weighted‑least‑squares inversion (applicable from 2004 onward) demonstrates that a closed solution exists within combined uncertainties for 2004–2016, but closure fails at the 1σ level from 2017 onward, with less than 32 % probability of consistency. The inversion requires modest adjustments to satellite altimetry but substantial corrections to the gravimetric mass term, suggesting the GRACE‑FO transition as a likely source of the budget breakdown.

These findings highlight the need for improved deep‑ocean observations, refined gravimetric processing, and spatial error‑correlation estimates to achieve reliable sea‑level budget closure at regional scales in the continued satellite era.