G3.5

Holocene relative sea-level (RSL) records from regions distal from ice sheets (far-field) are commonly characterized by a mid-Holocene highstand, when RSL reached higher than present levels. The magnitude and timing of the mid-Holocene highstand varies spatially due to hydro-isostatic processes including ocean syphoning and continental levering. While there are open questions regarding the timing, magnitude and source of ice-equivalent sea level in the middle to late Holocene.

Here, we compare Glacial Isostatic Adjustment (GIA) model predictions to a standardized database of sea-level index points (SLIPs) from Southeast Asia where we have near-complete Holocene records. The database has more than 130 SLIPs that span the time period from ~9.5 ka BP to present. We investigate the sensitivity of mid-Holocene RSL predictions to GIA parameters, including the lateral lithospheric thickness variation, mantle viscosity (both 1D and 3D), and deglaciation history from different ice sheets (e.g., Laurentide, Fennoscandia, Antarctica).

We compute gravitationally self-consistent RSL histories for the GIA model with time dependent coastlines and rotational feedback using the Coupled Laplace-Finite Element Method. The preliminary results show that the timing of the highstand is mainly controlled by the deglaciation history (ice-equivalent sea level), while the magnitude is dominated by Earth parameters (e.g., lithospheric thickness, mantle viscosity). We further investigate whether there is meltwater input during middle to late Holocene and whether the RSL records from Southeast Asia can reveal the meltwater source, like Antarctica.

How to cite: Li, T., Chua, S., Khan, N., Wu, P., and Horton, B.: Parameters controlling mid-Holocene highstand in Glacial Isostatic Adjustment modelling, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-4224, https://doi.org/10.5194/egusphere-egu21-4224, 2021.

This study focusses on improved constraint of the millennial time-scale glacial isostatic adjustment (GIA) signal at present-day, and its role as a contributor to present-day sea-level budgets. The study area extends from the coastal regions of northern Europe to Scandinavia. Both Holocene relative sea level (RSL) data as well as vertical land motion (VLM) data are incorporated as constraints in a semi-empirical GIA model. Specifically, 71 geological rates of GIA-driven RSL change are inferred from Holocene proxy data. Rates of vertical land motion from GNSS at 108 sites provide an additional measure of regional GIA deformation; within the study area, the geological RSL data complement the spatial gaps of the VLM data and vice versa. Both datasets are inverted in a semi-empirical GIA model to yield updated estimates of regional present-day GIA deformations. A regional validation is presented for the North Sea, where the GIA signal may be complicated by lateral variations in Earth structure and existing predictions of regional and global GIA models show discrepancies. The model validation in the North Sea region suggests that geological data are needed to fit independent estimates of GIA-related RSL change inferred from tide gauge rates, indicating that geological rates from Holocene data can provide an important additional constraint for data-driven approaches to GIA estimation. The geological proxy rates therefore provide a unique dataset with which to complement or validate existing data-driven approaches that use satellite era rates of change.

How to cite: Simon, K., Riva, R., and Vermeersen, B.: Constraint of GIA in Northern Europe and the North Sea with Geological RSL and GPS Data, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-13144, https://doi.org/10.5194/egusphere-egu21-13144, 2021.

How to cite: Pallisgaard-Olesen, G., Pedersen, V. K., and Gomez, N.: Sea level response to Quartenary erosion and deposition in Scandinavia, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9537, https://doi.org/10.5194/egusphere-egu21-9537, 2021.

The Last Interglacial (LIG) period (130 - 115 ka) was the last time in Earth’s history that the Greenland and Antarctic ice sheets were smaller than those of today due, in part, to polar temperatures reaching 3 - 5 °C above pre-industrial values. Similar polar temperature increases are predicted in the coming decades and the LIG period could therefore help to shed light on ice sheet and sea level mechanisms in a warming world.

The North Sea region is a promising study site for the reconstruction of both the magnitude and rate of LIG sea-level change as well as the identification of relative, individual ice sheet contributions to sea level. The impact of glacial isostatic adjustment (GIA) is particularly significant for the North Sea region due to its proximity to the former Eurasian ice sheet, which deglaciated during the penultimate deglaciation leading into the LIG. The evolution of the local Eurasian and global ice sheets during the penultimate glacial cycle has left a complex spatio-temporal pattern of GIA during the LIG, both regionally and globally. In addition, interpretation of the LIG record is further complicated by uncertainties in ongoing earth deformation and sea level evolution since the LIG. However, there are large uncertainties in the geometry and evolution of global ice sheets before the Last Glacial Maximum and, in particular, a major source of uncertainty for North Sea LIG records is the geometry and evolution of the Eurasian ice sheet during the Penultimate Glacial Maximum (PGM).

We produce a range of plausible global ice sheet histories spanning the last 400 thousand years that vary in penultimate deglaciation characteristics including glacial maximum ice sheet volume, deglaciation timing, and the ice volume distribution of the Eurasian ice sheet. This novel PGM Eurasian component is constructed with the use of a simple ice sheet model (Gowan et al. 2016) enabling systematic variation in the thickness of each ice sheet region within known uncertainty ranges. We then employ a gravitationally consistent sea level model (Kendall et al. 2005) with a range of viscoelastic Earth structure models to calculate the global GIA response to each ice history and to infer which input parameters the North Sea LIG signal is most sensitive to. This work will improve our understanding of the GIA effects on near field relative sea level during previous interglacials and will enable a systematic quantification of uncertainties in LIG sea level in the North Sea.

How to cite: Pollard, O., Barlow, N., Gregoire, L., Gomez, N., and Cartelle, V.: The Impact of Global Ice Sheet Evolution on North Sea Glacial Isostatic Adjustment during the Last Interglacial, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9841, https://doi.org/10.5194/egusphere-egu21-9841, 2021.

The evolution of the Greenland Ice Sheet (GIS) is an important indicator of climate change and driver of sea level rise. However, providing accurate GIS ice mass balance remains a challenge today. Here, we propose to combine a unique set of geodetic measurements to improve our knowledge of the GIS spatial and temporal evolution. We attempt at reconciling satellite observations of ice volume with regional GNSS velocities estimates and time variable space gravity measurements over the 2003-2009 and 2011-2015 periods. The GIS mass variations are inferred from satellite altimetry for large ice sheets (IceSat and CryoSat-2; Sorensen et al.,2018, Simonsen et al.,2017) and digital elevation models (DEMs) generated from multiple satellite archives for peripheral glaciers (Hugonnet et al.,2020), associated with IMAU-FDM firn model (Ligtenberg et al., 2011). The spatial and temporal variations of the gravity field are given by the GRACE mission for which we use a solution where smaller wavelength signals are preserved (Prevost et al., 2019).

To resolve short wavelengths load variations affecting the displacement of nearby GNSS stations, we use Green’s functions for vertical crustal displacements assuming purely elastic Earth properties (Martens et al., 2019). We first assume that the deformation is entirely due to recent ice melting and show that vertical elastic displacements predicted by our refined ice loading model, while in good agreement with observations in some regions, cannot explain observations overall. In particular, observations and model disagree in the Southeastern and the Northern parts of Greenland.

We then explore potential viscoelastic deformation associated with short-term rheology of the asthenosphere induced by recent ice melting that could explain the observed GNSS displacements. We define a history of ice loading from 1900 to 2009 using both in situ and satellite altimetric measurements, compute today’s associated viscoelastic deformation for various mantle rheologies and discuss the potential contribution of ice melting since the little ice age to current observations. Remaining differences between observations and viscoelastic models may reflect a viscoelastic deformation induced by glacial isostatic adjustment. We discuss implications in terms of regional rheological constraints, and impact on estimates of present-day GIS ice mass budget.

Hugonnet, R. (2020). A globally complete, spatially, and temporally resolved estimate of glacier mass change: 2000 to 2019. In EGU 2020. 

Ligtenberg, S. R. M., et al (2011). An improved semi-empirical model for the densification of Antarctic firn. The Cryosphere, 5, 809-819.

Martens, H. R.,et al (2019). LoadDef: A Python‐based toolkit to model elastic deformation caused by surface mass loading on spherically symmetric bodies. Earth and Space Science, 6(2), 311-323.

Prevost, P., et al (2019). Data-adaptive spatio-temporal filtering of GRACE data. Geophysical Journal International, 219(3), 2034-2055.

Simonsen, S. B., & Sørensen, L. S. (2017). Implications of changing scattering properties on Greenland ice sheet volume change from Cryosat-2 altimetry. Remote Sensing of Environment, 190, 207-216.

Sørensen, L. S., et al (2018). 25 years of elevation changes of the Greenland Ice Sheet from ERS, Envisat, and CryoSat-2 radar altimetry. Earth and Planetary Science Letters, 495, 234-241.

How to cite: Sanchez, A., Métivier, L., Fleitout, L., Chanard, K., and Marianne, G.: Spatio-temporal evolution of the Greenland ice sheet and associated deformation of the Earth: a multi-technic geodetic approach, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-7341, https://doi.org/10.5194/egusphere-egu21-7341, 2021.

Models of glacial-isostatic adjustment (GIA) exhibit large differences in north-east Greenland, reflecting uncertainties about glacial history and solid Earth rheology. The GIA uncertainties feed back to uncertainties in present-day mass-balance estimates from satellite gravimetry. We present results from repeated and continuous GNSS measurements which provide direct observables of the bedrock displacement. The repeated measurements were conducted within five measurement campaigns between 2008 and 2017. They reveal uplift rates in north-east Greenland in the range of 2.8 to 8.9 mm yr^-1. We used the observed uplift rates to validate different GIA models in conjunction with estimates of the elastic load deformation induced by present-day ice-mass changes and ocean mass redistribution. To determine present-day ice-mass changes for both the Greenland Ice Sheet and the peripheral glaciers, we combined CryoSat-2 satellite altimetry data with GRACE satellite gravimetry data. The different GIA models were consistently used in all processing steps. Our comparison between observed and predicted uplift rates clearly favours GIA models that show low rates (0.7 to 4.4 mm yr^-1 at the GNSS sites) over GIA models with higher rates of up to 8.3 mm yr^-1. Applying the correction predicted by the GIA model favoured in north-east Greenland we estimate an ice-mass loss of 233 ± 43 Gt yr^-1 for entire Greenland (including peripheral glaciers) over the period July 2010 to June 2017.

How to cite: Kappelsberger, M. T., Strößenreuther, U., Scheinert, M., Horwath, M., Groh, A., Knöfel, C., Lunz, S., and Khan, S. A.: Validation of GIA models in north-east Greenland using densified GNSS measurements and refined estimates of present-day ice-mass changes, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8019, https://doi.org/10.5194/egusphere-egu21-8019, 2021.

The lithospheric structure of Greenland is still poorly known due to its thick ice sheet, the sparseness of seismological stations, and the limitation of geological outcrops near coastal areas. As only a few geothermal measurements are available for Greenland, one must rely on geophysical models. Such models of Moho and LAB depths and sub-ice geothermal heat-flow vary largely.

Our approach is to model the lithospheric architecture by geophysical-petrological modelling with LitMod3D. The model is built to reproduce gravity observations, the observed elevation with isostasy assumptions and the velocities from a tomography model. Furthermore, we adjust the thermal parameters and the temperature structure of the model to agree with different geothermal heat flow models. We use three different heat flow models, one from machine learning, one from a spectral analysis of magnetic data and another one which is compiled from a similarity study with tomography data.

For the latter, a new shear wave tomography model of Greenland is used. Vs-depth profiles from Greenland are compared with velocity profiles from the US Array, where a statistical link between Vs profiles and surface heat flow has been established. A similarity function determines the most similar areas in the U.S. and assigns the mean heat-flow from these areas to the corresponding area in Greenland.

The geothermal heat flow models will be further used to discuss the influence on ice sheet dynamics by comparison to friction heat and viscous heat dissipation from surface meltwater.

How to cite: Wansing, A., Ebbing, J., Lösing, M., Lebedev, S., Celli, N., Karlsson, N. B., and Solgaard, A.: Implications for the lithospheric structure of Greenland by applying different heat flow models, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10689, https://doi.org/10.5194/egusphere-egu21-10689, 2021.

Although currently microtidal, the Arctic Ocean is known to have been megatidal at Last Glacial Maximum (LGM) due to the Arctic Ocean basin being nearly entirely enclosed, with only Fram Strait connecting it to the global ocean. This allowed for the propagation of a gravest mode coastal Kelvin wave traveling anti-clockwise around the Arctic ocean. The transition from the megatidal regime at LGM to the mircotidal regime observed today is not well understood, and the factors which control the amplitude of the semidiurnal tidal constituents in the Arctic Ocean have not been fully determined in the literature. We investigate the Arctic tidal regime across the Bolling-Allerod (B-A) onset, 14.6-14.1 ka, finding that the Arctic Ocean is megatidal prior to B-A onset and weakens considerably thereafter. The period of time during which the Arctic tidal regime is enhanced is precisely the time at which high Arctic ice streams begin to deglaciate, indicating that the tides may play a causal role in forcing the rapid deglaciation of the sector of the Laurentide abutting the Arctic Ocean. We further show that the deglaciation of the Laurentide ice sheet, through the mechanisms of Glacial Isostatic Adjustment (GIA) and gravitationally self-consistent local reduction in sea level, causes an increase in the amplitude of the principal lunar and solar semidiurnal tidal constituents in the Arctic Ocean. Additionally, it is the collapse of the Barents sea ice sheet which significantly weakens the Arctic Ocean tidal regime. We report the contribution of each major terrestrial ice sheet to the relative sea-level rise at each of Barbados, Tahiti, and Sunda Shelf, finding that the gravitationally self-consistent GIA model employed accurately predicts the RSL change at each of these sites and determines that the contribution at Barbados from the Laurentide is smaller than the contribution at Tahiti or Sunda Shelf due to the flow of ocean water away from the deglaciating Laurentide and into the "far field."  We further show that the contribution to RSL at Barbados due to the collapse of the Barents Sea ice sheet is significant. 

How to cite: Velay-Vitow, J. and Peltier, W. R.: Arctic Ocean tidal regime change across the Bolling-Allerod onset, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-29, https://doi.org/10.5194/egusphere-egu21-29, 2021.

A suite of forward GIA model predictions, spanning a wide range of layered mantle viscosity and lithospheric thickness values, is compared to observed horizontal crustal motions in North America to discern optimal model parameters in order to minimize a root-mean-square (RMS) measure of the velocity residuals. To obtain the Earth model response, a combination of the full normal mode analysis and the collocation method is implemented. It provides a means to determine the surface loading response automatically and robustly to 1-dimensional (radially varying) Earth models, while retaining as much of the physics of the normal mode method as numerically feasible, given documented issues with singularities along the negative inverse-time axis in the Laplace transform domain. This method enables the exploration across a wide parameter range (for the lower mantle, transition zone, asthenosphere, and thickness of the elastic lithosphere) to find optimal combinations to explain horizontal crustal motion in North America. The analysis utilizes crustal motion rates from approximately 300 GNSS sites in central North America (Canada and United States) provided by the Nevada Geodetic Laboratory.  Preliminary results indicate that as the lithospheric thickness increases, from 60 km to 240 km, the horizontal motion residuals gradually decrease with no minimum apparent for the thicknesses thus far considered. The residual velocities for the best-fitting models appear to carry a remaining signal, confirming previous inferences of limitations to spherically symmetric Earth models in modeling horizontal crustal motions in North America.

How to cite: Brierley-Green, C., James, T., Robin, C., Simon, K., and Craymer, M.: North American Crustal Motion and Glacial Isostatic Adjustment Model Predictions, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9007, https://doi.org/10.5194/egusphere-egu21-9007, 2021.

In Southeast Alaska, extreme uplift rates are primarily caused by glacial isostatic adjustment (GIA), as a result of ice load changes from the Little Ice Age to the present combined with a low viscosity asthenosphere. Current GIA models adopt a one-dimensional (1-D) stratified Earth structure. However, the actual (3-D) structure is more complex due to the presence of a subduction zone and the transition from a continental to an oceanic plate. A simplified 1-D Earth structure may not be an accurate representation in this region and therefore affect the GIA predictions. In this study we will investigate the effect of 3-D variations in the shallow upper mantle viscosity on GIA in Southeast Alaska. In addition, investigation of 3-D variations also gives new insight into the most suitable 1-D viscosity profile.

We test a number of models using the finite element software ABAQUS. We use shear wave tomography and mineral physics to constrain the shallow upper mantle viscosity structure. We investigate the contribution of thermal effects on seismic velocity anomalies in the upper mantle using an adjustable scaling factor, which determines what fraction of the seismic velocity variations are due to temperature changes, as opposed to non-thermal causes. We search for the combination of the scaling factor and background viscosity that best fits the GPS data. Results show that relatively small lateral variations improve the fit with a best fit background viscosity of 5.0×10¹⁹ Pa s, resulting in viscosities at ~80 km depth that range from 1.8×10¹⁹ to 4.5×10¹⁹ Pa s.

How to cite: Marsman, C., van der Wal, W., Riva, R., and Freymueller, J.: The impact of a 3-D Earth structure on glacial isostatic adjustment following the Little Ice Age in Southeast Alaska, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-914, https://doi.org/10.5194/egusphere-egu21-914, 2021.

The Coast Mountains in British Columbia and southeastern Alaska contain around 9040 km² of glaciers and ice fields at present. While these glaciers have followed an overall trend of mass loss since the Little Ice Age (or LIA around 300 years before present), the past decade has seen a significant increase in melting rate that is likely to continue due to the effects of climate change. The region is home to a complex tectonic setting, having proximity to the Queen Charlotte-Fairweather transform plate boundary in the northern region and the Cascadia subduction zone (CSZ) in the southern region, which has an associated active volcanic arc underlying the glaciated area. Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) glacier melt data collected between 2000 and 2019 represent a melt rate that is averaged between periods of relatively low mass loss (2000-2009) and high mass loss (2010-2019). As a preliminary test, this average melt rate was assumed to be constant back to the LIA. A history of gridded ice thicknesses was calculated to create an ice loading model for input to a series of forward modelling calculations to determine the crustal response. Predictions of vertical crustal motion are compared to available Global Navigation Satellite System (GNSS) measurements of uplift rate to constrain Earth rheology. The results using this simplified loading model favour a thin lithosphere (around 20-40 km thick) and asthenospheric viscosities on the order of 10¹⁹ Pa s. These values are significantly lower than those of rheological profiles used in extant global GIA models, but are in general agreement with previous GIA modelling of the forearc region of the CSZ. To improve the glacial history model, the Open Global Glacier Model (OGGM), driven by historic climate data and statistically downscaled climate projections, is being employed to create a more accurate loading model and refine our estimates of Earth rheology and regional crustal motion. The best-fitting models will be employed to separate GIA and tectonic components of crustal motion and to generate improved regional sea-level projections.

How to cite: Lauch, M., James, T., Leonard, L., Jiang, Y., Henton, J., and Brierley-Green, C.: Glacial Isostatic Adjustment Modelling of the Coast Mountains of British Columbia and Southeastern Alaska, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9011, https://doi.org/10.5194/egusphere-egu21-9011, 2021.

Studies of glacial isostatic adjustment (GIA) often use paleoshorelines and present-day deformation to constrain the viscosity of the mantle and the thickness of the lithosphere. However, different studies focused on similar locations have resulted in different estimates of these physical properties even when considering the same model of viscoelastic deformation. We argue that these different estimates infer apparent viscosities and apparent lithospheric thicknesses, dependent on the timescale of deformation. We use recently derived relationships between these frequency dependent apparent quantities and the underlying thermodynamic conditions to produce predictions of mantle viscosity and lithospheric thickness across a broad spectrum of geophysical timescales for three locations (Western North America, Amundsen Sea, and the Antarctic Peninsula). Our predictions require the self-consistent consideration of elastic, viscous, and transient deformation and also include non-linear steady state deformation, which have been determined by several laboratories. We demonstrate that these frequency dependent predictions of apparent lithospheric thickness and viscosity display a significant range and that they align to first order with estimates from GIA studies on different timescales. Looking forward, we suggest that observationally based studies could move towards a framework of determining the frequency trend in apparent quantities – rather than single, frequency independent values of viscosity – to gain deeper insight into the rheological behavior of Earth materials.

How to cite: Lau, H., Austermann, J., Holtzman, B., Book, C., Havlin, C., Hopper, E., and Lloyd, A.: Frequency Dependent Mantle Viscoelasticity via the Complex Viscosity: cases from Antarctica and Western North America, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1869, https://doi.org/10.5194/egusphere-egu21-1869, 2021.

The Wilkes Subglacial Basin in East Antarctica hosts one of the largest marine-based and hence potentially more unstable sectors of the East Antarctic Ice Sheet (EAIS). Predicting the past, present and future behaviour of this key sector of the EAIS requires that we also improve our understanding of the lithospheric cradle on which it flows. This is particularly important in order to quantify geothermal heat flux heterogeneity in the region.   

The WSB stretches for almost 1600 km from the Southern Ocean towards South Pole. Like many intracratonic basins, it is a long-lived geological feature, which originated and evolved in different tectonic settings. A wide basin formed in the WSB in a distal back arc basin setting, likely in response to a retreating West Antarctic Paleo-Pacific active margin from Permo-Triassic times. Jurassic extension then led to the emplacement of part of a huge flood basalt province that extends from South Africa to Australia.  The region was then affected by relatively minor upper crustal Mesozoic to Cenozoic(?) extension and transtension, producing narrow graben-like features that were glacially overdeepened, and presently steer enhanced glacial flow of the Matusevich, Cook and Ninnis glaciers.

Here we present the results of our enhanced geophysical imaging and modelling in the WSB region performed within the 4D Antarctica project of ESA, which aims to help quantify the spatial variability in subglacial Antarctic geothermal heat flux (GHF), one of the least well constrained parameters of the entire continent.

We exploit a combination of airborne radar and aeromagnetic data compilations and crustal and lithosphere thickness estimates from both satellite and airborne gravity and independent passive seismic constraints to develop new geophysical models for the region. To help constrain the starting models, including depth to basement beneath the Permian to Jurassic cover rocks, we applied a variety of depth to magnetic and gravity source estimation approaches from both line and gridded datasets. Given the huge differences between recent satellite gravity estimates of crustal thickness (Pappa et al., 2019, JGR) and sparse seismological constraints, we examine different scenarios for isostatic compensation of Rock Equivalent Topography and intracrustal loads, as a function of variable effective elastic thickness (Te) across the WSB and its flanks.  

Our models reveal a major lithospheric-scale boundary along the northeastern margin of the WSB, separating the Ross Orogen from a cryptic and composite Precambrian Wilkes Terrane. At the onset of enhanced flow for the central Cook ice stream, we image a Precambrian basement high with a felsic bulk composition. We suggest based on the similarity in potential field signatures that it represents late Paleoproterozoic to Mesoproterozoic igneous basement as exposed in South Australia, where it also associated with high GHF (80-120 mW/m²), primarily caused by anomalously radiogenic granitoids.

We hypothesise that the differences in basement depth and metasediment/sediment thickness, coupled with differences in intracrustal heat production give rise to significantly greater heterogeneity in GHF beneath different sectors of the WSB than previously recognised. To help quantify such heterogeneity we develop a suite of new probabilistic thermal models for the study region.

How to cite: Lowe, M., Ferraccioli, F., Young, D., Blankenship, D., Armadillo, E., Siegert, M., and Ebbing, J.: Unveiling lithosphere heterogeneity beneath the East Antarctic Ice Sheet in the Wilkes Subglacial Basin, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12953, https://doi.org/10.5194/egusphere-egu21-12953, 2021.

We established a new Geothermal Heat Flow (GHF) model for Antarctica by using a machine learning approach. GHF is substantially related to the geodynamic setting of the plates, and global geophysical and geological data sets can provide information for remote regions like Antarctica, where only sparse direct measurements exist. We applied a Gradient Boosted Regression Tree algorithm in order to build an optimal prediction model relating GHF to the observables.
Employed data sets are reviewed for their reliability and quality in polar regions and we emphasize the need for adding reasonable data to the algorithm. The validity of our approach is indicated by predictions for Australia, where an extensive database of GHF measurements exists. Our new estimated GHF map exhibits rather moderate values compared to previous models, ranging from 35 to 156 mWm^-2, and shows visible connections to the conjugate margins in Australia, Africa, and India.
Such estimates on the geothermal structure of Antarctica are for example needed for studies on ice sheet modeling. The internal thermal structure and the mass balance of the modeled Antarctic ice sheet (AIS) are significantly affected by the prescribed GHF distribution.  Applying a wide range of possible GHF maps within estimated uncertainties to ice-sheet-shelf set-ups, the influence of GHF on the modeled AIS response to a variety of climate scenarios is quantified. 

How to cite: Lösing, M., Bernales, J., and Ebbing, J.: A Machine Learning Heat Flow Model of Antarctica, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-5764, https://doi.org/10.5194/egusphere-egu21-5764, 2021.