The mass loss of the Greenland Ice Sheet is substantially affected by surface melt , but the amount of liquid water that stored form in the firn layer is not well constrained. While the interior of the ice sheet is only occasionally experiencing surface melt , the ablationzone is prone to intensive melting over the summer months . In highly crevassed shear zones and other regions of high stress changes firn layers are even more accessible to melt water.
In soil, ponding of water over cracks in summer leads to distinct infiltration characteristics. We assume that this also applies for firn, and therefore investigate whether areas of ponding over crevasses are locations where radar images indicate the retention of water. To this end we will use different SAR missions in X, C, and L-band, with a focus on L-band.

Our study is focused on the area of 79°N Glacier, an outlet glacier of the North East Greenland Ice Stream. We conduct a Pauli decomposition of ALOS-1 and ALOS-2 data to distinguish between different scattering mechanism. We find features of very low backscatter in all polarizations, that are moving at the speed of the glacier. Furthermore, we use data acquired with AWI’s ultrawideband (UWB) radar from 2018 and 2021 to investigate the internal structure in the upper hundred meters of the glacier across such features. We compare our findings from observations with simulated firn/ice transition depth using an Lagrangian along-track approach. In addition to that, we classify crevasse zones and aim to constrain their depth with the ice penetrating radar. This way, we can assess whether water infiltration in summer is the origin of the observed features.

It is essential to advance our knowledge of the hydrology of the Greenland Ice Sheet, to assess its current contribution to global sea-level rise, and to improve its projection into a future warming climate. The high latitudes of the Northern Hemisphere have experienced a large regional warming over the last decades, and as a consequence the amount of liquid water at the surface of the Greenland Ice Sheet has significantly increased.
The 4DGreenland project (https://4dgreenland.eo4cryo.dk/) is funded by the ESA as part of the POLAR+ programme. It runs for two years (2020-2022) and its main objective is to maximize the use of Earth Observation (EO) to assess and quantify the hydrology of the Greenland ice sheet. The project is focused on dynamic variations in the hydrological components of the ice sheet, and on quantifying the water fluxes between the hydrological reservoirs.
Here, we present the initial results and data products, and analysis from the project. These include:
• improved mapping of drainage and refilling of active subglacial lakes using different satellite sensors
• basin-wide subglacial melt from an EO-informed model on a monthly resolution
• meltwater volume contained in supraglacial lakes based on optical imagery and ICESat-2 derived lake bathymetry
• identification of surface melt extent from SAR and PMW, with value added products such as melt onset, duration and intensity.
We have used a wide range of satellite data sets, and have developed, improved, implemented and tested methods to map the different hydrological components from these. We have combined the EO-derived data products with output from regional climate models to generate a basin-scale monthly time series of runoff from the Greenland ice sheet for selected test sites. In the next phases of the project, the analysis will be carried out ice sheet wide and for the time period 2010-present (the CryoSat-2 period). All products will be available at (as a minimum) monthly resolution. Some products will be available for a longer time period and at a higher temporal resolution.

As climate warming intensifies the melting of the Greenland Ice Sheet, increased fluxes of meltwater drain from its surface to the ocean. The passage of water on top, through, and ultimately underneath the ice is important, because of the influence it exerts upon future ice sheet evolution. This critical, but elusive, system is being studied using a suite of satellite data by ESA’s Polar+ 4D Greenland study. Here, we report analysis of ~35,000 high-resolution (2 metre) time-stamped Digital Surface Models (DSMs), to search for evidence of previously unknown hydrological phenomena around the entire margin of the Greenland Ice Sheet. This new dataset provides unprecedented detail of localised surface elevation changes, and allows us to identify previously unknown hydrological signals.

In this presentation, we focus on a detailed case study of a large – and highly unusual – subglacial lake drainage event identified beneath the Brikkerne Glacier, in northern Greenland. In summer 2014, approximately 9x107 m3 of water drained from this previously unknown subglacial lake, causing an 80 metre drop in the ice surface. Further downstream, the resulting subglacial flood breached back through the surface of the ice sheet, causing fracturing of the ice sheet surface, and depositing ice debris measuring 25-metres in height. To our knowledge this is the first time that such behaviour has been observed in an ice sheet setting, and reveals a complex, bidirectional coupling between the surface and basal hydrological systems. Analysis of historical satellite imagery suggests that the subglacial lake drained previously, in 1990; however, on that occasion the subglacial floodwater failed to breach the ice surface. This suggests that climate warming and ice sheet thinning over the past three decades may have helped to facilitate the passage of subglacial floodwater to the surface. As such, our observations demonstrate a new, and poorly understood, aspect of Greenland hydrology and show, more broadly, how high-resolution streams of satellite data can be used to obtain insight into the hydrological system. This work forms a contribution to ESA’s Polar+ 4D Greenland study.

The ice velocity of the Greenland Ice Sheet can be monitored from space using data from the ESA Sentinel-1 mission. The PROMICE ice velocity product exploits the high temporal and spatial resolution of the Sentinel-1 SAR data to produce Greenland Ice Sheet wide ice velocity mosaics that resolves variations on the seasonal scale. The product is a continuously updated time series (September 2016 - present) of mosaics produced using intensity offset tracking of Sentinel-1 SAR data and is posted on a 500 m grid. Each mosaic spans two Sentinel-1 A cycles (24 days) and a new one is made available every 12 days, and is provided within approximately 10 days of the last acquisition.

In this contribution, we use machine learning to investigate the seasonal dynamics of the Greenland Ice Sheet as well as the interannual variations. The length, coverage and the high temporal and spatial resolution of the time series make it possible to study not only a few locations on the ice sheet, but to investigate the flow of ice over large parts of the ice sheet and over several years.

We perform a fully automated, unsupervised clustering of annual time series in all grid points in the fast flowing part of the Greenland Ice Sheet. The analysis reveals how a number of well defined clusters describe the general seasonal flow patterns. It also shows how the clusters dominate different regions of the ice sheet and different areas of the outlet glaciers. The patterns of where different clusters dominate are not stationary but show significant interannual variations. By correlating to other data sets e.g. modelled surface run-off and topography, we can infer information on the processes linked to the different seasonal behaviours.

We identify spatial patterns in the seasonal velocity along all the marine outlet glaciers of the Greenland ice sheet and show that the patterns depend on the local availability of meltwater and vary upstream from the glacier front. Our findings represent a significant step forward in understanding the processes controlling the marine outlet glaciers compared to previous studies that were restricted to include a limited number of datapoints and only considered a fraction of the outlet glaciers. Our results underline the importance of obtaining long time series of ice velocity covering the entire Greenland ice sheet and in high spatial and temporal resolution, and provide insights needed to improve estimates of the future dynamic mass loss from the Greenland ice sheet.

Near-surface density is a key parameter for assessing the overall mass balance of ice sheets. Specifically, it is used during the conversion of surface elevation changes inferred from laser and/or radar altimetry into mass changes. As such, constraining the full mass balance of an ice sheet requires insight of spatiotemporal patterns in near-surface density that themselves are typically derived from global and/or regional climate models. While these models are calibrated with a handful of in situ measurements, there is currently no observational approach for constraining large-scale spatial and temporal patterns in near-surface density. In face of a changing climate, a measurement-driven approach to understanding of the spatial and temporal variability in surface density would represent a major step forward in reducing the uncertainty in estimates of ice sheet dynamics as well as future sea-level rise.

Beginning with ERS-1 in the early 1990’s, conventional space-borne radar altimetry studies of ice sheets have been overwhelmingly focused on using the time delay between signal transmission and the reception the reflected echo to determine how far the spacecraft is from the surface. Repeating these measurements through time and space then allows for ice sheet volume to be monitored across spatiotemporal scales unattainable using either airborne or in situ methods. At the same time however, little attention has been paid to the strength with which radar waves are reflected from the Earth’s surface. This is not unsurprising as measured radar wave reflection strengths are affected by several physical factors including, but not limited to, the contrast in electromagnetic properties across the Earth’s surface interface as well as the roughness of that interface.

The desire to quantitatively assess radar surface echo strengths represents a unique opportunity to introduce new methodologies initially developed for other planetary surfaces into the analysis of Earth Observation datasets. Specifically in this study we investigate using the Radar Statistical Reconnaissance (RSR) method originally developed for and applied on Mars. RSR leverages the statistical distribution of radar echo strengths over an assumed to be homogeneous region in order to separate the coherent and incoherent components of the reflected echo. The coherent component is related to the contrast in electromagnetic properties across the surface interface while the incoherent component is a function of surface roughness. Therefore, by applying the RSR methodology to surface echo powers recorded within a region, we can begin to quantitatively describe the geophysical parameters of that area.

In this work, we present early results from applying the RSR methodology to Ku- and Ka-band satellite radar altimetry measurements of the Greenland Ice Sheet (GrIS). We make use of minimally processed waveforms recorded by the SIRAL (Ku-band) instrument on-board ESA’s CryoSat-2 satellite as well as the ALtiKa (Ka-band) instrument on-board the joint ISRO/CNES SARAL satellite. To maximize surface echo density, and therefore RSR spatial resolution, we have focused our CryoSat-2 efforts thus far on the analysis of SARIn waveforms covering the marginal portions of the GrIS as opposed to less-dense LRM waveforms which are restricted to the interior. We generate RSR results for two instruments with different inherent range resolutions (SIRAL and ALtiKa signal bandwidths of 320 and 500 MHz respectively) in order to assess any impact of near-surface structure on our interpretations.

In addition to clear patterns in both coherent and incoherent powers, the RSR results reveal unique features dependent on data source. As SIRAL SARIn waveforms are acquired at a much higher rate (bursts of 64 pulses at 17.8 kHz, with a burst repetition frequency of 85 Hz) than those from ALtiKa (pulse repetition frequency of 40 Hz), the maximum radial distance around each RSR grid point required to gather a relevant statistical population of surface echo powers is much greater for ALtiKa as opposed to SIRAL. It then follows that for complex surfaces, such as those near the ice sheet margin, because surface echoes from a larger area must be included in the SARAL/ALtiKa RSR analysis, the likelihood that the echo powers reflect a homogenous surface is reduced. This stationarity assumption is implicit in RSR, which when violated, degrades the reliability of the ALtiKa results in these areas as opposed to SIRAL. In contrast, as SARIn measurements are performed only along the margins of the GrIS, the reliability of the current CryoSat-2/SIRAL RSR results diminishes towards the interior.

Moving forward, in order to incorporate the RSR results into future GrIS mass balance estimates, the coherent component must be calibrated to reflect near-surface density. Calibration will be performed using the SUMup database, which contains a comprehensive record of near-surface density measurements from firn/ice cores and snow pits from across the GrIS. To achieve maximum compatibility, calibration will be performed using only RSR results from the same months as the in situ density measurements were collected before being applied across the entirety of the RSR dataset.

In closing, by expanding the scope of how radar satellite altimetry datasets are analyzed, this research endeavors to provide new quantifiable and observational insights into the mass balance of the GrIS that serve to help reduce uncertainty with regards to estimating its contribution to future sea-level rise.

Runoff from the Greenland Ice Sheet has increased over recent decades affecting global sea level, regional ocean circulation, and coastal marine ecosystems. Runoff now accounts for most of Greenland’s contemporary mass imbalance, driving a decline in its net surface mass balance as the regional climate has warmed. Although automatic weather stations provide point measurements of surface mass balance components, and satellite observations have been used to monitor trends in the extent of surface melting, regional climate models have been the principal source of ice sheet wide estimates of runoff. To date however, the potential of satellite altimetry to directly monitor ice sheet surface mass balance has yet to be exploited. Here, as part of the Polar+ Earth Observation for Surface Mass Balance (EO4SMB) project, we explore the feasibility of measuring ice sheet surface mass balance from space by using CryoSat-2 satellite altimetry to produce direct measurements of Greenland’s runoff variability, based on seasonal changes in the ice sheet’s surface elevation. Between 2011 and 2020, Greenland’s ablation zone thinned on average by 1.4 ± 0.4 m each summer and thickened by 0.9 ± 0.4 m each winter. By adjusting for the steady-state divergence of ice, we estimate that runoff was 357 ± 58 Gt/yr on average – in close agreement with regional climate model simulations (root mean square difference of 47 to 60 Gt/yr). As well as being 21 % higher between 2011 and 2020 than over the preceding three decades, runoff is now also 60 % more variable from year-to-year as a consequence of large-scale fluctuations in atmospheric circulation. In total, the ice sheet lost 3571 ± 182 Gt of ice through runoff over the 10 year survey period, with record-breaking losses of 527 ± 56 Gt/yr first in 2012 and then 496 ± 53 Gt/yr in 2019. Because this variability is not captured in global climate model simulations, our satellite record of runoff should help to refine them and improve confidence in their projections.