Evaluating global models of deformation from ongoing ice mass changes and long-term GIA

Systematic subsidence of ~1 mm/yr is observed across the Pacific at GNSS sites on islands that lack significant local tectonic and volcanic processes (Altamimi et al., 2023; Ballu et al., 2019; Hammond et al., 2021).  However, the horizontal motion of these sites is well described by Pacific plate motion.  This suggests that the observed subsidence represents a deeply rooted geophysical signal, rather than just localized deformation. 

Both ongoing and past ice mass redistribution are known to produce global deformation. Previous models of recent global ice mass redistribution (Coulson et al., 2021; Riva et al., 2017) predict tenths of a mm/yr of subsidence in far field locations such as the Pacific.  In addition, post-LGM GIA models like ICE-6G also predict subsidence in the Pacific on the order of tenths of a mm/yr.

Here we evaluate three new models of the global deformation associated with present day ice mass redistribution. These models use the methods of the elastic loading model originally published by Coulson et al. (2021), utilizing new mass change estimates with increased spatial and temporal coverage as input. The three updated models all use Velicogna et al.’s (2020) mass change estimates for the Antarctic and Greenland ice sheets, paired with global glacier mass change estimates from either Ciraci et al. (2020), the Copernicus group (Dussaillant et al., 2024), or Hugonnet et al. (2021). These models are evaluated at selected GPS sites near field to glaciers in regions such as SE Alaska, Greenland, etc., as well as 27 Pacific GPS sites located far field from ice mass change. We also use these far field sites to evaluate 39 different long-term GIA models that predict the present-day viscoelastic response of the earth to past loading. 

We find that all three models of elastic deformation due to recent global ice mass change produce very similar results in the far field. The most significant differences in the models are seen in the near field in SE Alaska and Svalbard. Additionally, there is a set of 15 long-term GIA models that improve the fit of both the horizontal and vertical observations in the Pacific when used in combination with an ongoing cryospheric loading model to correct the GPS data. Overall, we find that the sum of the deformation due to ongoing ice mass changes and long-term GIA explains about half of the subsidence signal that we observe in the far field. 

Studies of the contribution of different components of barystatic sea level (BSL) suggest that though cryospheric melting is the largest contributor, non-cryospheric terrestrial water storage (TWS) could be responsible for ~17 – 25% of BSL over the past couple of decades (Nie et al., 2025; McGirr et al., 2024).  Since changes in TWS may represent a non-trivial global loading signal, we choose to also consider if deformation associated with TWS may explain part of the residual ~0.5 mm/yr subsidence signal that we see in the far field after our cryospheric loading corrections.