Causes of Global Elastic Vertical Land Movement from 1900 to 2022

Elastic vertical land movement (eVLM) is the lithosphere's immediate elastic response to the loading and unloading of the Earth's surface mass. Understanding eVLM is crucial for interpreting relative sea level changes, particularly in coastal regions where subsidence or uplift can significantly alter the impacts of sea level changes recorded by tide gauges. Here we present a comprehensive global eVLM model, offering valuable insights for geodesy and related fields, especially in assessing observations from tide gauges and GNSS.

Our eVLM model spans from 1900 to 2022, featuring a 0.5-degree spatial resolution. It provides annual data from 1900 to 1990 and monthly data from 1991 to 2022, enabling both long-term and seasonal assessment. The dataset is available in three different reference frames: Centre of Mass (CM), Centre of Figure (CF), and ITRF2020, and thus suitable for many geodetic applications.

This study incorporates mass change estimations from Greenland, Antarctica, global glaciers, and land water storage (LWS), divided into natural LWS variations and anthropogenic water management like groundwater depletion and dam retention. Thus, we can explain regional VLM patterns that cannot be solely attributed to Glacial Isostatic Adjustment (GIA) models, for example, subsidence across Australia or uplift in Scandinavia that is larger than modeled GIA.

Methodology: We employed a composite loading model, integrating ice models from Greenland (Mankoff et al., 2021) and Antarctica (Otosaka et al, 2022; Nilsson et al, 2022) and glacier models (Hugonnet et al., 2022), GRACE observations, and a land water storage model (Müller-Schmied et al, 2023). Each of the aforementioned five causes of eVLM was perturbed with its uncertainty a thousand times, and the sea level equation was resolved for each variant using the ISSM-SEESAW framework (Adhikari et al., 2016). To align the results with observations in the ITRF2020 reference frame, which mirrors CM on secular timescales and CF on non-secular timescales (Dong et al, 2003). To accommodate this, we applied CM and CF Love loading numbers (Blewitt, 2003) in our calculations, enabling analysis in all three reference frames.