Methodology:

This dataset includes three components: Input geophysical compilations, inversion results and interpreted geological boundaries. Details of the data compilation, inversion and basic geological interpretation are presented here. This dataset is the companion to the following paper Jordan et al., (in prep) Geological sketch map and implications for ice flow of Thwaites Glacier, West Antarctica, from integrated aerogeophysical observations, Science Advances., which contains full details of the geological interpretation and should be cited when this dataset is used.

Input geophysical compilations:

This dataset includes compilations of aerogeophysical data collected across the Thwaites Glacier region of West Antarctica over the past several decades. Surveys used include: AGASEA and BBAS which collected radar, magnetic and gravity data (Diehl et al., 2008; Holt et al., 2006; Jordan et al., 2010; Vaughan et al., 2006). NASA Operation Ice Bridge (OIB) which collected radar and gravity data (Paden et al., 2010, updated 2021; Tinto et al., 2010, updated 2019). The International Thwaites Glacier Collaboration (ITGC) which collected radar, magnetic and gravity data (Jordan et al., 2020a; Jordan and Robinson, 2021a; Jordan and Robinson, 2021b, c; Jordan et al., 2020c). Additional magnetic data was taken from the ADMAP2 compilation (Golynsky et al., 2018).

The presented bed-elevation dataset was created from the radar-derived bed elevation picks from the surveys noted above interpolated onto a 2 km mesh raster, padded with gravity derived bathymetry beneath adjacent ice shelves (Jordan et al., 2020b) and BEDMAP2 in more distal areas (Fretwell et al., 2013). All topographic data were corrected to an ellipsoidal reference frame prior to integration.

The presented subglacial topographic roughness grid was created from the along track standard deviation of the picked radar bed elevation within a 1 km wide moving window, interpolated onto a 2 km mesh raster. Areas where the bed was poorly resolved, defined as having along-line data gaps more than 50 percent of the analysis window (500 m), were excluded from the gridded product.

The presented aeromagnetic compilations were created from line data available through the ADMAP2 project (Golynsky et al., 2018) and incorporating new ITGC data (Jordan et al., 2020a; Jordan and Robinson, 2021c). Data were upward/downward continued to 500 m above the ice/sea surface to place all observations on a common surface prior to reapplication of statistical levelling. After levelling and continuation crossover errors were 18.79 nT. An alternative continuation to 2500 m above the underlying rock was also carried out to give a view of the data with a consistent distance to the closest possible source. The two versions of the final continued line data were interpolated onto a 1 km mesh, and subject to a 3 pass Hanning filter.

The presented free air gravity anomalies are from the surveys noted above after detailed data editing and levelling. Note that strictly the presented gravity compilation reflects ''gravity disturbances'', rather than anomalies as they are referenced to the ellipsoid. The line data from ITGC, AGASEA and BBAS were statistically levelled and edited using the OIB dataset as a reference after high altitude (more than 1500 m above ice surface) OIB survey lines had been removed. The line free air data was continued to a uniform level of 2500 m. After levelling and continuation the free air gravity data had an internal crossover error of 2.94 mGal. The line data were interpolated onto a 2 km raster mesh using a minimum curvature method.

The presented Bouguer gravity anomaly includes correction for a full 3D model of the gravity effect of the topography, bathymetry and ice, based on the topographic compilation noted above and ice surface elevations taken from the BEDMAP2 (Fretwell et al., 2013). An o...(54)

Data quality:

See Lineage for data quality of data compilations.