Sea ice plays a critical role in Earth's climate system since it provides a barrier between the ocean and atmosphere, restricting the transfer of heat between the two. Due to its high albedo, the presence of sea ice reduces the amount of solar energy absorbed by the ocean. Arctic sea ice rejects brine during formation and fresh water during melting, and it is therefore a driving force of the global thermohaline circulation as well as the stratification of the upper layer of the ocean. The sea ice provides a critical habitat for marine mammals and for biological activity (e.g. Tynan et al., 2009), and it is a platform that enables subsistence hunting and travel for indigenous coastal communities.

The sea ice cover of the Arctic Ocean is waning rapidly. By 2019, the decline in September Arctic sea ice extent was about 13 % per decade relative to the 1981–2010 average, and the older, thicker, multi-year ice cover comprised ∼ 20 % of the winter ice pack compared to ∼ 45 % in the 1980s (Perovich et al., 2017; IPCC/SROCC, 2019). In the Southern Ocean sea ice is undergoing regional changes, with a decline observed in the Amundsen and Bellingshausen seas (Shepherd et al., 2018). These losses are having a profound impact on the climate, environment, and ecosystems of both polar regions. Monitoring the polar oceans is therefore of regional and global importance, and the long-term continuity of sea ice measurements is essential to extending both climate and operational data services.

Global warming and its Arctic amplification continue to contribute to the decrease in multi-year ice in the central Arctic Ocean (north of 81.5∘ N). It is therefore critical to obtain continuous, pan-Arctic observations of sea ice thickness, extending as close as possible to the North Pole. Continuous monitoring of Arctic Ocean sea ice conditions is necessary for safe navigation through ice-covered waters. When linked to previous measurements from Envisat, ICESat, CryoSat-2, and ICESat-2, the CRISTAL mission will deliver observations that provide a long-term record of sea ice thickness variability and trends that are critical to supporting climate services. Since sea ice thickness is an essential climate variable (see GCOS, 2011), its continuous measurement is required to understand the Arctic system and how ice loss is impacting global climate.

Shipping in ice-covered Arctic waters has increased significantly in recent years and is expected to continue to do so over the coming decades (IPCC/SROCC, 2019). In addition to traditional maritime operations and fishing in the high Arctic, several polar-class cruise liners are under construction. This means an increase in the need and scope of operational ice information services. A primary data source for national ice services is currently synthetic-aperture radar (SAR) imagery, specifically data acquired by Sentinel-1A and Sentinel-1B, RADARSAT-2, and the RADARSAT Constellation Mission. Thus, independent measurements of sea ice thickness distribution at reasonable latencies provided by CRISTAL will complement existing SAR measurements and benefit operational ice charting. Furthermore, observed sea ice thickness or freeboard distributions can be assimilated into sea ice models to generate ice forecasts needed for ice navigation and offshore operations.

Historically, satellite observations had primarily been used to monitor ice extent until Laxon et al. (2003) produced the first Arctic-wide sea ice thickness estimates from European Remote Sensing (ERS) satellite radar altimetry. Since then, various methods for converting the received signal to physical variables have been established (Giles et al., 2008a; Laxon et al., 2013; Kurtz et al., 2014; Ricker et al., 2014; Price et al., 2015; Tilling et al., 2018; Hendricks et al., 2018). The capability to obtain an estimate of sea ice freeboard and thickness and convert it to estimates of ice volume has enabled scientists to better understand the changing Arctic ice cover. Most recently, sea ice freeboard has been estimated from both Ka- and Ku-band measurements (Armitage and Ridout, 2015; Guerreiro et al., 2016; Lawrence et al., 2018).

Most sea ice thickness products are currently provided on a 25 km grid (see e.g. Sallila et al., 2019, for an overview of different products currently available), which corresponds to the GCOS user requirements (GCOS, 2011) but does not meet the specified accuracy requirements of 0.1 m. The residual, systematic uncertainty in sea ice thickness is estimated to be 0.56–0.61 m for ICESat (Connor et al., 2013), and it is 0.6 m for CryoSat-2 observations over first-year ice and 1.2 m for those over multi-year ice (Ricker et al., 2014). The uncertainty in ice thickness derived from CryoSat-2 observations is driven mainly by the unknown penetration of the radar pulse into the snow layer as a result of variable snow properties (Nandan et al., 2017, 2020) as well as the choice of retracker (Ricker et al., 2014). Reference is also made to Mallett et al. (2020), who find that assumptions concerning the time evolution of overlying snow density can lead to underestimates of sea ice thickness from radar altimetry.

While the focus of the Copernicus programme is on the Arctic, comprising all areas north of the southernmost tip of Greenland (∼ 60∘ N), the parameters specified for polar regions should equally be provided for its southern counterpart, the Antarctic, as well as all non-polar snow- and ice-covered surfaces.

The requirements for CRISTAL are currently stated to provide sea ice freeboard with an accuracy of 0.03 m along orbit segments of less than or equal to 25 km during winter months and to provide meaningful freeboard measurements during summer months. Winter months are months from October to April in the Northern Hemisphere and from May to October in the Southern Hemisphere. The system shall be capable of delivering sea ice thickness measurements with a vertical uncertainty of less than 0.15 m along orbit segments ≤ 25 km in winter months and of providing meaningful sea ice thickness estimates during summer months. The along-track resolution of sea ice thickness measurements shall be at least 80 m. The uncertainty requirement for sea ice thickness comes with a caveat as the thickness uncertainty depends on the uncertainty of auxiliary products. In the case of CRISTAL, snow thickness will be measured by the system, but snow and ice densities will still have to be estimated by other means. In light of the current 0.2 m sea ice thickness uncertainty from CryoSat-2 data assessed by Tilling et al. (2018) for a gridded, monthly product and the anticipated improvement from the dual-altimetry technology, especially in the snow depth and propagation estimates, a higher vertical uncertainty would seem reachable but requires further study. Currently, the retrieval accuracy of sea ice freeboard is limited by the range resolution of a radar altimeter. The large bandwidth of 500 MHz is an important driver for the CRISTAL instrument concept generation. A bandwidth of 500 MHz will improve the range resolution from 50 cm (as for CryoSat-2, with 320 MHz bandwidth) to ∼ 30 cm for CRISTAL. A radiometer will help in active–passive synergy to classify sea ice type (see e.g. Tran et al., 2009, for further justification).

An accurate estimate of snow depth over Arctic sea ice is needed for signal propagation speed correction to convert radar freeboard to sea ice freeboard and freeboard to sea ice thickness (Laxon et al., 2003, 2013). The penetration aspects of a dual-frequency snow depth retrieval algorithm over Antarctica are complex (Giles et al., 2008b; Shepherd et al., 2018) and are not further elaborated here. In addition to uncertainty reduction for ice thickness and freeboard computation, the variation in snow depth is a parameter that is highly relevant for climate modelling, ice navigation, and polar ocean research. The snow climatology of Warren et al. (1999) is still the single most used estimate of snow depth in sea ice thickness processing (Sallila et al., 2019). The uncertainty in the original Warren et al. (1999) snow depth estimates is halved over first-year ice (Kurtz and Farrell, 2011; Zhou et al., 2020), but snow still represents the single most important contribution to uncertainty in the estimation of sea ice thickness and volume (Tilling et al., 2018). The studies of Lawrence et al. (2018) and Guerreiro et al. (2016) show the possibility of using Ku- and Ka-bands in mitigating the snow depth uncertainty. Dual-frequency methods improve the ability to reduce and estimate the uncertainties related to snow depth and sea ice thickness retrieval. The modelling community is particularly interested in the uncertainty information according to the user requirement study in the PEG 1 report. Having better abilities to estimate the related uncertainties improves prediction quality assessment of annual snowmelt over Arctic sea ice (Blockley and Peterson, 2018). The stratigraphy and electromagnetic properties of the snow layer contrast with those of the underlying ice and can be exploited to retrieve information on the snow layer properties if contemporaneous measurements are acquired from multiple scattering horizons (for details see Giles et al., 2007, who demonstrated the propagating uncertainties associated with snow depth and other geophysical parameters). A dual-frequency satellite altimeter, as proposed for the CRISTAL mission, will address this need. CRISTAL aims to provide an uncertainty in snow depth retrieval over sea ice of less than or equal to 0.05 m. The additional Ka-band measurements, with a 500 MHz bandwidth, support the discrimination between the ice and snow interfaces.

Earth's land ice responds rapidly to global climate change. For example, melting of glaciers, ice caps, and ice sheets over recent decades has altered regional and local hydrological systems and has impacted sea levels and patterns of global ocean circulation. The Antarctic and Greenland ice sheets are Earth's primary freshwater reservoirs and, due to their progressive imbalance, have made an accelerating contribution to global sea level rise during the satellite era (Shepherd et al., 2018, 2019). Glaciers outside of the ice sheets constituted nearly one-third of all sea level rise over the past 2 decades (Gardner et al., 2013; Wouters et al., 2019) Although ice dynamical models have improved, future losses from the polar ice sheets remain the largest uncertainty in sea level projections. Due to their continental scale, remote location, and hostile climatic environment, satellite measurements are the only practical solution for spatially and temporally complete monitoring of the polar ice sheets.

Estimates of ice sheet surface elevation change provide a wealth of geophysical information. They are used as the basis for computing the mass balance and sea level contribution of ice sheets of both Greenland and Antarctica (McMillan et al., 2014, 2016; Shepherd et al., 2012), for identifying emerging signals of mass imbalance (Flament and Rémy, 2012; Wingham et al., 2009), and for determining the loci of rapid ice loss (Hurkmans et al., 2014; Sørensen et al., 2015). Through combination with regional climate and firn models of surface processes, surface elevation change can be used to isolate ice dynamical changes at the scale of individual glacier catchments (McMillan et al., 2016).

A unique and continuous record of elevation measurements is provided by radar altimeters dating back to 1992. The maps are typically delivered in (1) high-resolution (5–10 km) rates of surface elevation change (for single or multiple missions, typically computed as a linear rate of change over a period of several years to decades) and (2) frequently (monthly, quarterly) sampled time series of the cumulative change, averaged across individual glacier basins. In addition to being used to quantify rates of mass balance and sea level rise, they also have a range of other applications, such as detection of subglacial lake drainage (Siegert et al., 2016) and investigations of the initiation and speed of inland propagation of dynamic imbalance (Konrad et al., 2017) that provide valuable information relating to the underlying physical processes that drive dynamical ice loss.

CRISTAL will extend the decades-long record of the generation of elevation measurements provided by radar altimeters. It will produce maps of ice surface elevation with an uncertainty of 2 m (the vertical accuracy threshold is 2 m, an absolute accuracy of 0.5 m can be assumed, and there is a relative accuracy goal of 0.2 m). The system shall be capable of delivering surface elevation with an along-track resolution of at least 100 m and a monthly temporal sampling. CRISTAL will be capable of tracking steep terrain with slopes less than 1.5∘ using its SARIn (interferometric synthetic-aperture radar) mode. High-resolution swath processing over ice sheets (about 5 km wide) can reveal complex surface elevation changes related to climate variability and ice dynamics as well as subglacial geothermal and magmatic processes (see e.g. Foresta et al., 2016). Elevation measurements of regions with smaller glaciers are often missing in CryoSat-2 data. Indeed, tracking algorithms are not efficient when rough terrain is encountered. Improvement in the tracking over glaciers is thus a key element in the instrument concept generation.

Over the years and through constant improvement of the data quality, satellite altimetry has been used in a growing number of applications in Earth sciences. The altimeter measurements are helping us to understand and monitor the ocean: its topography, dynamics, and variability at different scales. Satellite observations for studying, understanding, and monitoring the ocean are more than essential over polar areas, where in situ data networks are very sparse and where profound and dramatic changes occur. This has also been expressed and emphasized by the Copernicus Marine Environmental Monitoring Service (CMEMS) as “ensuring continuity (with improvements) of the CryoSat-2 mission for sea level monitoring in polar regions” (CMEMS, 2017). “Reliable retrieval of sea level in the sea ice leads to reach the retrieval accuracy required to monitor climate change” is another CMEMS recommendation for polar and sea ice monitoring (see CMEMS, 2017).

Current data from the CMEMS catalogue do not allow a satisfactory sampling north of 81.5∘ N. It is of prime importance that the CRISTAL orbit configuration allows measurement coverage of the central Arctic Ocean with an omission not exceeding 2∘ of latitude around the poles. Sea level anomalies (SLAs) over frozen seas can only be provided by measurements in the leads. CRISTAL will contribute to the observation system for global observation of mean sea level, (sub-)mesoscale currents, wind speed, and significant wave height as a critical input to operational oceanography and marine forecasting services, and it will support sea ice thickness retrieval in the Arctic.

The high-inclination orbit of CRISTAL associated with high-resolution SAR and SARIn bi-band altimetry measurements would considerably extend our monitoring capability over the polar oceans. The development of tailored processing algorithms should not only have to track the low-frequency sea level trend in the presence of sea ice and to characterize large-scale and mesoscale ocean variations over regions not covered by conventional ocean altimeters. Beyond the observations of ice elevation variations, CRISTAL would offer the unique opportunity to improve our knowledge of the mutual ocean–cryosphere interactions over short- and long-term timescales for both poles. Southern Ocean circulation plays a key role in shaping the Antarctic cryosphere environment. First, it regulates sea ice production: as sea ice forms and ejects brine into the ocean, the ocean destabilizes and warms submerged waters that reach the ocean surface, limiting further ice production. Second, it impacts Antarctic ice sheet melt when warm and salty ocean currents access the base of floating glaciers through bathymetric troughs of the Antarctic continental shelf. These ocean currents melt the ice shelves from below and are the main causes of the current decline in floating ice shelves (Shepherd et al., 2019; Smith et al., 2020). Thus melting of ice shelves represents one of the largest uncertainties in the current prediction of global sea level change (Edwards et al., 2019), creating a major gap in our ability to respond and adapt to future climate change. Tightly linked with glacier melt, polar shelf circulation and its interaction with large-scale circulation also control the rate of bottom water production and deep-ocean ventilation, which impact the world's oceans on a timescale ranging from decades to millennia. Therefore, with a designed operational lifetime of at least 7.5 years (including in-orbit commissioning), the observation from the same sensor of each component of these multi-scale ice–ocean interactions would make CRISTAL unique in its capability to address climate issues of regional and global relevance. Over oceans, a secondary objective for the mission, the satellite will be able to measure sea surface height with an uncertainty of less than 3 cm. The main advantages and drawbacks of the Ka-band over the oceanic surface have been reviewed in Bonnefond et al. (2018). Given its high along-track resolution of less than 10 km and high temporal resolution of sea level anomalies, the mission can further contribute a suite of sea level products including sea surface height and mean sea surface (vertical accuracy in sea level anomaly retrieval of less than 2 cm is requested). The radiometer on board CRISTAL corrects the satellite altimeter data for the excess path delay resulting from tropospheric humidity. The microwave radiometer measurements will complement wet tropospheric corrections derived from numerical weather prediction and non-collocated atmospheric data from other satellite instruments to help meet the range accuracy requirement (Picard et al., 2015; Legeais et al., 2014; Vieira et al., 2019).

Observation of water level at the (Arctic) coast as well as of rivers and lakes is a key quantity in hydrological research (e.g. Jiang et al., 2017). Rivers and lakes not only supply fresh water for human use, including agriculture, but also maintain natural processes and ecosystems. The monitoring of global river discharge and its long-term trend contributes to the monitoring of global freshwater flux, which is critical for understanding the mechanism of global climate change. Satellite radar altimetry is a promising technology to do this on a regional to global scale. Satellite radar altimetry data have been used successfully to observe water levels in lakes and (large) rivers and have also been combined with hydrologic and hydrodynamic models. Combined with gravity-based missions like the NASA and Deutsches Zentrum für Luft- und Raumfahrt (DLR) GRACE and GRACE-FO missions, the joint use of the data will give information for ground water monitoring in the future.

Iceberg detection, volume change, and drift have been listed as a priority user requirement (Duchossois et al., 2018a, b).

Icebergs present a significant hazard to marine operations. Detection of icebergs in open water and in sea ice generally places a priority on wider satellite swaths to obtain greater geographic coverage. There is a need for automatic detection of icebergs for the safety of navigation and chart production. Iceberg concentration is given in CMEMS' catalogue at 10 km resolution covering Greenland waters. SAR imagery is the core input for iceberg detection. However, iceberg detection (in particular small icebergs) is also possible using high-resolution altimeter waveforms. Tournadre et al. (2018) demonstrated detection of icebergs from CryoSat-2 altimeter data using several modes and mention promising results with the Sentinel-3 data, which would be fed into a comprehensive dataset already built as part of the ALTIBERG project (Tournadre et al., 2016). The volume of an iceberg is valuable information for operational services and climate monitoring. For climate studies, the freshwater flux from the volume of ice transported by icebergs is a key parameter, with large uncertainties related to the volume of the icebergs. Measuring volume is currently only possible with altimetry by providing the iceberg freeboard elevation from the ocean surface. Iceberg volume has been calculated with altimetry with Envisat, Jason-1, and Jason-2 (e.g. Tournadre et al., 2015).

CryoSat-2 tracking over icebergs is operational, but icebergs with high freeboards may be missed in the current range window. The range window definition for CRISTAL is defined in order to ensure that echoes from icebergs are correctly acquired. In-flight performances for the measurement of the angle of arrival from CryoSat-2 are around 25 arcsec. An equivalent performance is necessary to retrieve across-track slopes and elevations. The CRISTAL design of the instrument and the calibration strategy will be designed to comply with the specification of 20 arcsec. CRISTAL will provide the unprecedented capability to detect icebergs at a horizontal resolution (gridded product) of at least 25 m. The products will be produced every 24 h in synergy with other high-resolution data such as SAR imagery. Iceberg distribution and volume products will be produced at 50 km resolution (gridded) on a monthly basis.

CRISTAL may support and contribute to studies and services in relation to seasonal snow cover and permafrost applications over land. These are considered a secondary objective for the mission. The ability to retrieve snow depth with Ku- and Ka-band altimetry is limited over land (Rott et al., 2018). Snow studies over land area are so far largely limited to scatterometer when the Ku-band is used; examples of such retrievals are reviewed in Bartsch (2010). Measurements as provided by CRISTAL may, however, be useful in retrieving internal properties of the snowpack such as the existence of ice layers (e.g. due to rain on snow; Bartsch et al., 2010). The relevant properties of an upper snow layer contrast with those of an underlying ice layer (see also Sect. 4.2). Further, snow structure is reflected in differences observed in radar observations using different frequencies (Lemmetyinen et al., 2016). Snow structure anomalies as well as land surface state (freeze and thaw) are expected to be identified by time series analyses as such processes alter penetration depth. Altimeter data are also rarely used for permafrost studies. Such data can also be applied for monitoring lake level as a proxy for permafrost change (Zakharova et al., 2017). Surface status is closely interlinked with ground temperature (e.g. Kroisleitner et al., 2018), but usage of satellite altimetry in this context remains unexplored. Signal interaction with vegetation limits the applicability of Ku- and Ka-bands for soil observations regarding freeze and thaw status (Park et al., 2011) as well as surface height. Wider use of altimetry for snow and permafrost applications requires higher spatial resolution and temporal coverage than what is available to date. An improvement regarding the latter issues is expected with CRISTAL, which will expand the utility of altimeter observations for permafrost and snow monitoring over land.