Understanding the long-term changes in the ice sheets of Greenland has global climate significance, especially on long term global sea level rise predictions. Having validated satellite data on current and recent past changes of the ice sheets are crucial for validating climate and earth system models, and give good opportunities for space geodesy to play an important role for society.

During the 2012-18 ESA Climate Change Initiative Greenland ice sheet project, a 25-year time series, validated and with error estimates, have been derived for five different ECV’s (surface elevation changes, ice velocities from radar and optical data, gravimetric mass balance, and outlet glacier calving fronts and grounding lines), and the data have been made for use by scientists, stakeholders and the general public. The data highlights the acceleration of the Greenland ice sheet melt, but also the highly varying dynamic behaviour of the ice sheet, with large variability in year-to-year changes.

With the different earth observation methods having different error characteristics, the combination of different ECVs, along with in-situ data from airborne, ground and model data, gives an overall improved high-resolution picture of the mass changes of the Greenland ice sheet; we show examples of such applications involving both GRACE, altimetry, ice velocities and GNET GPS data.

Numerous recent studies have revealed rapid change in ice discharge from Greenland, as many of the ice sheet’s outlet glaciers have accelerated dramatically over the last two decades. These observations are significant in that they show Greenland’s mass balance can fluctuate rapidly and unpredictably. To document and improve understanding of these changes, we are producing Greenland-wide ice-velocity maps using Landsat, Copernicus Sentinel 1A/B, TerraSAR-X/TanDEM-X, and COSMO-SkyMed data. Earlier products in this nearly two decade-long record were derived using RADARSAT-1, ALOS PALSAR, and ERS-1/2 data. With the advent of regular coastal SAR coverage for Greenland from the Copernicus Sentinel 1A/B satellites, supplemented with data from other sensors, we can now produce full coastal mosaics with relatively uniform temporal sampling, including monthly, quarterly, and annual sampling. For individual glaciers, we can resolve changes over intervals as small as 6 days with the Copernicus Sentinel 1A/B sensors. Fine temporal resolution, however, yields greater errors, leading to a trade-off between accuracy and temporal resolution. We examine changes on several glaciers in an effort to determine the optimal sampling interval. While Copernicus Sentinel 1A/B provides broad spatial coverage, X-band sensors provide far finer resolution, albeit over only limited areas. Thus, we examine the impact of spatial resolution on the measurement of fast-flowing glaciers. In general, we find that the consistency of the results is good through time and across sensors, but our analysis indicates substantial differences can arise in regions with high strain rates (e.g., shear margins) where sensor spatial resolution can become a factor. These results indicate that the best ice-sheet wide products are obtained when combining results from multiple sensors, taking advantage of best properties of each. We anticipate further improvements in data quality as new sensors come on line, particularly with the launch of the NASA-ISRO Synthetic Aperture Radar (NISAR) in early 2022. The Greenland Ice Mapping Project products are being used by numerous projects that assimilate the data into models with the goal understanding future sea level rise from Greenland.

Figure: Examples of the types of velocity products being produced by the Greenland Ice Mapping Project. These results were deriving using NASA/USGS Landsat 8, Copernicus Sentinel 1A/B, and TerraSAR-X/Tandem-X data.

Accurate ice velocity measurements are a requisite for monitoring and modelling processes that are determining the contributions of polar ice masses to sea level rise. After an era of intermittent coverage, the systematic acquisition planning of Sentinel-1 provides nearly uninterrupted observations of the Greenland Ice Sheet margin, of key regions in Antarctica and of Arctic ice caps and glaciers, including those in the Russian and Canadian Arctic, Svalbard and Patagonia. The dual-satellite configuration of Sentinel-1A and B reduces the repeat observation period of the SAR mission to 6 days. This changed the landscape for satellite remote sensing of the cryosphere, providing the basis for operational monitoring of key climate variables such as ice sheet velocity and ice discharge. Continuous coverage is now extended to include the main part of the Antarctic Ice Sheet margin, allowing retrieval of dense time series of ice flow for major Antarctic outlet glaciers previously rarely observed. The ongoing acquisition of ice sheet margins is augmented by dedicated annual ice sheet-wide campaigns over Greenland.

In the framework of the ESA CCI and Austrian ASAP/FFG programs we developed an automatic system for ice sheet velocity and discharge monitoring. Applying customized iterative offset tracking tools we generate surface velocity maps from repeat pass Sentinel-1 Interferometric Wide Swath (IWS) data. The Terrain Observation by Progressive Scans (TOPS) technology, with a spatial resolution of 4 m by 22 m in slant range and azimuth, respectively, and a swath width of 250 km, are well suited for comprehensive velocity retrievals over large ice bodies. These Sentinel velocity products compare very well with those derived from very high-resolution sensors, such as TerraSAR-X, and provide detailed coverage of ice motion even in regions with high snowfall rates, fast ice flow and high velocity gradients. Combined with ice thickness, derived from airborne radio echo sounding, the velocity maps form the basis for studying glacier dynamics, calculating ice discharge and estimating mass balance. We present ice sheet wide velocity maps of the Greenland and Antarctic mapping campaigns and show time series of velocity variations of major outlet glaciers of the ice sheets and other polar ice bodies and their evolution over time. Based on extended time series including velocity products from other missions (ERS, ALOS, TerraSAR-X), we show how velocity and ice discharge varies spatially and temporally over time scales ranging from days to years. The continuous repeat observation capability of Sentinel-1 with 6 day time intervals offers also excellent capabilities for mapping of grounding lines and monitoring their migration by means of SAR interferometry.

During winter 2018/19 the 5th ice sheet wide Greenland mapping campaign is planned and monitoring of the Antarctic margins is going on. Sentinel-1 continues delivering essential information for comprehensive monitoring of polar ice masses, a prerequisite for understanding and predicting there response to climatic change.

During the last years, the mapping of surface ice velocity at continental scale (Greenland, Antarctica) has been primarily realized through feature and speckle tracking of optical and SAR satellites, respectively. The technique assumes 2-D surface parallel flow and a single acquisition is in principle sufficient. The precision of such mappings is limited by the resolution of the raw images acquired by the satellite sensors, typically around 10 meter. Thus, the recent mappings of Antarctica or Greenland reach, at best, precision in velocity of tens of meters per year, which is not enough to observe the flow close to the divides where horizontal velocity is smaller than 1 meter per year.
Combining the ascending/descending interferogram phases from synthetic aperture radar is 10 to 100 times more accurate than tracking, however, it remains challenging to use at an ice sheet wide scale. Phase analysis requires more data (ascending and descending), phase unwrapping remains difficult in areas with complex motion patterns, L-band sensors are strongly impacted ionosphere phase shift and calibration requirements are more stringent due to the increased accuracy. Here, we present for the first time a new innovative approach to calibrate and combine the unwrapped phases acquired by 11 different sensors during the last 2 decades (ERS1&2, RADARSAT1&2, ALOS1&2, TSX, COSMO-SkyMed, Envisat/ASAR, Sentinel-1a&b) in Antarctica. The resulting phase-based ice velocity map covers more than 70% of the ice sheet surface, only limited by data availability and by phase aliasing in regions of fast flow. We will discuss how this new map helps to measure precisely the flow up to the divides, hence providing a new reference map for defining glacier drainage basins, calculating accurate strain rates or extending farther inland mass-conservation based methods to estimate ice thickness.

These results also help to evaluate and prepare the workflow for the coming NASA/ISRO mission NISAR to map the ice sheets at unprecedented accuracy using ascendin and descending coverages every 12 days.
The study is made possible thanks to 5 space agencies (NASA, ESA, JAXA, DLR, ASI) and the coordination from the Polar Space Task Group.

Advances in Earth Observation have transformed our capacity to study the polar regions, and the past three decades of satellite measurements have painted a new picture of how Earth’s ice sheets, sea ice, and polar oceans evolve. As global temperatures have risen, so too have rates of snowfall, ice melting, and sea level rise, and each of these changes impacts upon the neighbouring land, marine, and atmospheric environments. In this presentation, we report the satellite measurements that have proved central to our awareness and understanding of Arctic and Antarctic environmental change.

We have now 27 years of precise observations of ice dynamics in Greenland and Antarctica from SAR interferometry, coupled with historical data from Landsat and other optical sensors going back to the 1960s. We evolved from a basic mapping of ice velocity and grounding line position from the early satellites to the precise calculation of ice sheet mass balance using the component approach (mass flux vs inputs of snowfall minus surface melt), and more important to the detection of temporal and spatial changes in ice velocity and grounding line positions, especially in places that are deemed to be most vulnerable to climate warming. With Sentinel-1/2, Landsat-8, we evolved from annual time series to monthly time series which revealed an important seasonal dynamics in Greenland and temporal variability in Antarctica. In terms of grounding line dynamics, we observe complex and wide variations at tidal frequencies in regions where the bed topography is nearly flat, which has implications on how we model ocean-ice interactions in grounding zones and what intensity of melt should be used in models. In terms of ice dynamics, we can constrain the regime of calving of large glaciers, document ice shelf melt rates from above, and detect the impact of large calving events on the glacier flow. The results of these analyzes benefit enormously from the surge in data availability from Sentinel and Landsat, in anticipation of future programs such as the NASA ISRO SAR satellite, and sparse high temporal resolution data from the CSK constellation. It is of uttermost importance for constraining model projection of ice sheet evolution to maintain and improve that observation network, improve the quality of the observations and improve the temporal resolution of the data. It is also essential to maintain a free network of satellite data to encourage a young generation of researchers to participate in this research and encourage other agencies to follow the same pathways as ESA and NASA.