Shackleton Ice Shelf bathymetry - MCMC gravity inversion results

Overview

This dataset contains ensemble-based gravity inversion results for the Shackleton Ice Shelf region, East Antarctica. The products include bed topography realizations, the best-fitting bed model, and corresponding gravity realizations used to assess uncertainty in the recovered sub-ice shelf bathymetry.

The inversion combines airborne gravity observations with direct bed constraints from radar sounding and bathymetric measurements. An ensemble approach is used to characterize non-uniqueness and uncertainty in the recovered bed geometry.

Coordinate Reference System

All gridded products are provided in Antarctic Polar Stereographic coordinates.

Projection:

• Projection: Antarctic Polar Stereographic

• PROJ string: +proj=stere +lat_0=-90 +lat_ts=-71

File Descriptions

best_bed.nc

Best-fitting bed elevation model from the ensemble inversion.

Variables:

• beds (y, x)

  • Best model bed elevation.

  • Units: metres (m)

• bed_measured (y, x)

  • Directly observed bed elevations and bathymetry derived from radar sounding and sonar measurements.

  • Units: metres (m)

Global attributes:

• best_index

  • Index of the selected best-fitting realization within the ensemble (lowest RMS in gravity fit).

all_beds.nc

Complete ensemble of bed topography realizations generated during the inversion.

Dimensions:

• sim: simulation number

• y: northing coordinate

• x: easting coordinate

Variables:

• beds (sim, y, x)

  • Ensemble of bed elevation realizations.

  • Units: metres (m)

• bed_measured (y, x)

  • Directly observed bed elevations and bathymetry used as inversion constraints.

  • Units: metres (m)

The ensemble can be used to calculate summary statistics such as mean bed elevation, standard deviation, confidence intervals, and derived quantities including subglacial water routing and ice thickness uncertainty.

all_gravity_realizations.nc

Gravity realizations associated with the bed topography ensemble.

Variables:

• terrain_effect (simulation, point)

  • Predicted gravity response of the recovered bed topography for each realization.

  • Units: mGal

• non_terrain_effect (point)

  • Estimated non-terrain gravity component removed prior to inversion, representing long-wavelength geological contributions and other gravity signals not directly attributable to bed topography.

  • Units: mGal

• residual (simulation, point)

  • Difference between predicted and observed gravity values for each realization.

  • Units: mGal

Global attributes:

• best_index

  • Index of the best-fitting realization (lowest RMS in gravity fit).

Method Summary

The inversion estimates bed topography by matching observed gravity anomalies while honoring direct radar and bathymetric constraints. Multiple realizations are generated through a stochastic inversion framework to quantify uncertainty arising from sparse observations and gravity inversion non-uniqueness.

The resulting ensemble provides a probabilistic representation of bed geometry rather than a single deterministic solution. Regions close to radar and gravity observations are generally better constrained, whereas uncertainty increases with distance from direct measurements.

Gravity Data Configurations

Two alternative gravity datasets were used to investigate the sensitivity of the inversion results to the gravity observations included.

The first inversion uses only directly observed gravity measurements acquired during airborne and shipborne surveys in the study region. These observations provide the highest-resolution gravity constraints and represent the primary inversion dataset.

shackleton_final_icecap_only

The second inversion additionally incorporates gravity disturbance data from the Antarctic Gravity Grid 2021 (AntGG2021). The AntGG2021 product provides continent-wide gravity coverage derived from a combination of satellite, airborne, shipborne, and terrestrial gravity observations. AntGG2021: Scheinert, Mirko; Zingerle, Philipp; Schaller, Theresa; Pail, Roland (2024): Antarctic gravity anomaly and height anomaly grids (AntGG2021) [dataset]. PANGAEA, https://doi.org/10.1594/PANGAEA.971238

shackleton_final_with_antgg