Reliable information on snow cover across the Northern Hemisphere and Arctic and sub-Arctic regions is needed for climate monitoring, for understanding the Arctic climate system, and for the evaluation of the role of snow cover and its feedback in climate models. In addition to being of significant interest for climatological investigations, reliable information on snow cover is of high value for the purpose of hydrological forecasting and numerical weather prediction. Terrestrial snow covers up to 50 million km² of the Northern Hemisphere in winter and is characterized by high spatial and temporal variability making satellite observations the only means for providing timely and complete observations of the global snow cover. The ESA Snow CCI project will continue the efforts from ESA GlobSnow project to improve methodologies for snow cover extent (SE) and snow water equivalent (SWE) retrieval [1] using satellite data and construct long term data records of terrestrial snow cover for climate research purposes.

The efforts to improve the satellite-based SWE data record within ESA Snow CCI will deal with:

• improving the retrieval methodology (the snow emission model and the snow density consideration within the SWE retrieval scheme)
• improving the consistency and stability of the synoptic weather station based snow depth data record, applied as auxiliary data source for SWE retrieval
• Improving the spatial resolution of the processing system
• Utilization of enhanced resolution input satellite radiometer data
• Utilization of optical-satellite derived Snow Extent information in SWE retrieval
• Advanced uncertainty characterization, and improved uncertainty estimates with the SWE product

The new retrieval methodology will combine medium resolution optical data with passive microwave-based SWE retrieval to produce coarse resolution (approx. 12.5km or 6.25km spatial resolution) estimates on hemispherical scale. The new approach will be assessed and validated with the goal of producing a daily SWE data record spanning from 1980 to present day. The resulting data record will be characterized and analyzed in regard to retrieval accuracy and stability, changes in long term snow conditions and climatological trends in Northern Hemisphere, Eurasia and Asia.


References
[1] Takala, M., Luojus, K., Pulliainen, J., Derksen, C., Lemmetyinen, J., Kärnä, J.-P, Koskinen, J., Bojkov, B., “Estimating northern hemisphere snow water equivalent for climate research through assimilation of space-borne radiometer data and ground-based measurements”, Remote Sensing of Environment, Vol 115 (2011).

Earth’s cold regions are showing rapid changes, necessitating improvements in our understanding of how future variations will impact human activity at all latitudes. Unlike the recent past where the primary research focus was to determine whether ice sheet changes were occurring, the issue now is how fast changes in sea level and land ice will progress and what should be the human response to these changes. To fully answer these questions and to provide improved predictability, new and complementary information can be obtained from radiometry in the low-microwave bands that are able to penetrate deep in the ice. Indeed, up to now spaceborne sensors measure ice sheet surface parameters such as skin temperature, deformation, the presence of sea ice, and its concentration. The availability of new L-band spaceborne radiometers has opened the possibility to probe the inner characteristics of the ice, e.g. to determine the thickness of thin sea ice (<1m). The concurrent development of new radiofrequency interference mitigation techniques extends the possibility of using the radiometry in bands that are not protected but which are also able to probe deeper within the ice sheet. In this framework, The Ohio State University has developed a new instrument that acquires data from 0.5 to 2 GHz. Electromagnetic radiation in this band can penetrate the ice from several hundred meters up to the bedrock, thus it can be used to retrieve the internal geophysical parameters of the internal ice sheet, the ice shelf, and the sea ice. The science objectives behind developing the sensor are to determine the physical temperature of ice sheets and ice shelves at 10 m depth, to measure the average physical temperature from surface to base, and to measure the vertical temperature profile. The instrument is also designed to estimate more accurately sea ice thickness up to several meters to support scientific investigations of the polar oceans and to provide additional operational information for ships operating in these regions. This new sensor has been developed in the framework of a NASA project, and the first airborne measurements were carried out in Greenland in September 2016 and 2017. The first results from these campaigns clearly demonstrated the potential of this technique.
With this heritage, a new airborne campaign in Antarctica is being conducted during November 2018. The campaign is part of the project ISSIUMAX (Ice Sheet and Sea Ice Ultrawideband Microwave Airborne eXperiment) funded by the Italian Antarctic Programme, and has the objective to investigate multiple geophysical areas in the region. In particular, airborne measurements will be conducted first in a coastal region close to Mario Zucchelli Station ( 74.695°, 164.114°) over sea ice, ice shelf, and glaciers, and then in the inner part of the Antarctic plateau from Mario Zucchelli Station to the Dome C region (-75° , 123°) . In the vicinity of Dome-C, areas having different ice sheet physical characteristics will then be observed. For instance, places having lakes at the bedrock level, areas with megadunes, and areas having surface windcrust are included on the planned flight path. Figure 1 shows the planned flight route which covers a total of approximately 6500 km. An analysis of ESA’s SMOS 1.4 GHz data over the same locations (Figure 2) provide preliminary information on the expected Tb variability. Data from SMOS and other satellites have been used to design the flight paths and will also be used for intercalibration purposes.
The 0.5-2 GHz brightness temperature data obtained will be used along with ancillary data (including conventional ground measurements as well as L-, C-, X-band spaceborne SAR images, satellite brightness temperatures, GPR acquisitions, glaciological models estimations, etc.), to derive (i) sea ice thickness; (ii) ice sheet internal temperature profiles within the plateau; and (iii) classification of glacial facies of the ice sheet. These products will be obtained by means of retrieval algorithms which are currently under development.
Besides the geophysical goals of the experiment, the project will also aim at analyzing the synergy between active and passive sensors in polarregions and at the intercalibratio of different instruments (e.g. including the Domecair L-band 1.4 GHz passive microwave measurements collected by DTU and emerging Icesat-2 data). The first results will be shown at the time of the conference.

The Copernicus Imaging Microwave Radiometer (CIMR) mission is a high priority candidate mission within the European Copernicus Expansion program, specifically designed to observe the Arctic environment, its sea ice and ocean. CIMR is a conically scanning microwave radiometer imager, with a large swath (>1400km) and full coverage of the poles, in a Sun synchronous orbit similar to MetOp-SG. It includes channels at 1.4, 6.9, 10.65, 18.7 and 36.5 GHz (at least dual polarizations) with low noise receivers and mitigation of the Radio Frequency Interferences. The requirement for spatial resolution is < 55 km at 1.4 GHz, <15 km at 6.9 and 10.65 GHz, and < at 5 km at 36 GHz. It will provide all weather Sea Ice Concentration (SIC) and Sea Surface Temperature (SST), as primary products, along with a large range of other ocean and ice parameters as secondary products (e.g., Sea Ice Thickness for thin ice, Ocean Wind Speed including high wind speed, Sea Surface Salinity)

An end-to-end simulator framework is developed to analyze the performances of the CIMR mission. It is based on optimal estimation principles. The methodology will be described during this presentation, along with the derived results obtained under a large of environmental conditions. Preliminary results have already been documented (Kilic et al., 2018). There will be consolidated and extended to better account from the instrument specifications. The retrieval methodologies will be detailed and improved.

Over ice-free ocean, radiative transfer models are inter-compared with existing satellite observations to derive a unified and consistent approach across frequencies and environmental conditions. Over sea ice, the analysis is based on satellite observations under known conditions (the Round Robin Data Package of the ESA sea ice Climate Change Initiative project) that provide the 0 and 100% sea ice signatures from which a simple linear model can be developed. The expected theoretical performances of the retrievals from CIMR are estimated using a careful information content analysis, under realistic assumptions, and including a large range of environments. They are systematically compared to the current satellite possibilities using ASMR2 and SMAP. The retrieval methodology is based on optimal estimation, for both SIC and SST. For SIC, the demonstrated performance covers the influence of the uncertainty of the tie points on the retrieval at SIC=0 and 1, and propagates at intermediate SIC.

We show that CIMR can provide SIC and SST with a spatial resolution of better than 5 km and 15 km, and a precision of 5% and 0.2 K, respectively, twice daily, using algorithms developments that fully benefit from the CIMR capacity in terms of frequency coverage and spatial resolution. The SST can be retrieved at better than 30 km from the coast. In addition, algorithms are developed to estimate coincident Ocean Wind Speed, even under extreme conditions, Sea Surface Salinity, and Sea Ice Thickness from CIMR, with improved precision and / or accuracy as compared to the current missions.

Potential synergies of the CIMR mission with other satellite observations using complementary techniques will also be discussed.

Reference:

Kilic, L., Prigent, C., Aires, F., Boutin, J., Heygster, G., Tonboe, R. T., Roquet, H., Jimenez, C., and Donlon, C. (2018). Expected performances of the Copernicus Imaging Microwave Radiometer (CIMR) for an all-weather and high spatial resolution estimation of ocean and sea ice parameters. Journal of Geophysical Research: Oceans, 123. https://doi.org/10.1029/2018JC014408

For almost 9 years, Soil Moisture and Ocean Salinity (SMOS) L-Band interferometer has been providing an unprecedented time series of Sea Surface Salinity (SSS) estimates from space. SMOS provides SSS maps with resolutions of around 50 km and 3 days revisit time. In addition, since 2015, Soil Moisture Active and Passive (SMAP) completes the SSS satellites constellation with a similar spatial resolution and 2-3 days revisit time.
The purpose of this study is to present SMOS ability to monitor SSS in cold water and in polar oceans. SMOS SSS accuracy and precision are assessed by using a set of various in-situ data available in international data bases, as well as russian data bases, including ship Thermosalinographs (TSG), Conductivity, Temperature, Depth (CTD) stations, Argo measurements, in situ data from field campaigns and In Situ Analysis System (ISAS) gridded salinity products. Variability of SMOS SSS are further compared with SMAP SSS (from Remote Sensing System and Jet Propulsion Laboratory).
The L-Band measurement sensitivity to SSS decreases with Sea Surface Temperature (SST). In addition, ice variability, roughness conditions (high wind speed) and land vicinity makes SSS monitoring from space in polar oceans very challenging.
Considering these difficulties, a methodology allowing for the mitigation of SMOS systematic errors have been developed by the LOCEAN Expertise Center of the Centre Aval de Traitement des Données SMOS (CATDS) (Kolodziejczyk et al. 2016, Boutin et al. 2018a). In the last version (debias_v3, Boutin et al. 2018b), the SSS correction takes into account sea surface temperature (SST) dependency of the L-band brightness temperature with the sea surface salinity (SSS) and flaws in the Klein and Swift (1977) dielectric constant model. With this new correction, SMOS SSS are much better consistent with in situ measurements, in particular with the SSS from the Arctic Tara expedition in 2013. Nevertheless, in some locations, uncertainty related to the absolute calibration of the SSS over the whole satellite period remain, which origins are being investigated. These SMOS SSS are combined with other SSS products through an Optimal Interpolation method originally developed in the subtropical Atlantic Ocean (Kolodziejczyk et al. 2015) and extended to the global ocean at LOPS laboratory. Taking advantage of both L-Band satellites, SMOS and SMAP SSS estimates are merged.
The results of inter-comparison among SMOS, SMAP and in situ data will be discussed in the context of monitoring SSS variability in the Polar Oceans. Seasonal to interannual variability of SSS, as well as physical features, such as fronts and fresh river plumes will be examined in detail in order to assess the capability of the various satellite SSS products to monitor SSS variability at various spatio-temporal scales in cold water oceans and explore the various components of the error budget.

Boutin J., Vergely J. L., Marchand S., D'Amico F, Hasson A., Kolodziejczyk Nicolas, Reul Nicolas, Reverdin G., Vialard J. (2018a). New SMOS Sea Surface Salinity with reduced systematic errors and improved variability. Remote Sensing Of Environment, 214, 115-134. Publisher's official version : https://doi.org/10.1016/j.rse.2018.05.022 , Open Access version : https://archimer.ifremer.fr/doc/00441/55254/.

Boutin Jacqueline, Vergely Jean-Luc, Khvorostyanov Dmitry (2018b). SMOS SSS L3 maps generated by CATDS CEC LOCEAN. debias V3.0. SEANOE. https://doi.org/10.17882/52804.

Gaillard Fabienne, Reynaud Thierry, Thierry Virginie, Kolodziejczyk Nicolas, Von Schuckmann Karina (2016). In-situ based reanalysis of the global ocean temperature and salinity with ISAS: variability of the heat content and steric height. Journal Of Climate, 29(4), 1305-1323. Publisher's official version : https://doi.org/10.1175/JCLI-D-15-0028.1 , Open Access version : https://archimer.ifremer.fr/doc/00309/42030/.

Kolodziejczyk, N., J. Boutin, O. Hernandez, A. Sommer, G. Reverdin, S. Marchand, N. Martin,J.-L. Vergely, and X. Yin, Argo SSS and SMOS SSS combination helps monitoring SSS variability from basin scale to mesoscale, 16-21, Mercator Ocean—Coriolis Quarterly Newsletter #52, Mai 2015.

Snow on sea ice plays an important role for the Arctic energy budget. A layer of snow in early summer increases the surface albedo, and delays the onset of the sea-ice melt period. During winter, the snow acts as an insulator, thereby reducing heat loss from the sea ice to the cold atmosphere. Additionally, the snow layer adds significant uncertainty to sea-ice thickness estimates from satellite altimetry. The snow significantly affects the microwave signatures of sea ice, suggesting that the observed brightness temperatures can be used for inferring properties of the sea-ice snow cover.

We have developed two methods to retrieve pan-Arctic spring-time snow depths over sea ice from 2013 to 2018 based on AMSR-2 microwave radiometer data. All seven frequency channels (6.9 to 89.0 GHz) at vertical and horizontal polarization were employed. First, snow depths were retrieved using Bayesian inversion of an empirical surface emission model. The optimization procedure selected a consistent set of physical parameters, which produced the best fit to the observed microwave emissions, while accounting for the associated uncertainties. The skill in snow thickness retrieval was evaluated using a Round Robin Data Package containing collocated sets of AMSR-2 brightness temperatures and Operation Ice Bridge (OIB) observations of snow thicknesses along flight tracks. Secondly, a random forest regression model was trained to predict snow thickness on Arctic sea ice from the collocated set of OIB snow thickness observations and measured brightness temperatures. The resulting snow thicknesses derived from applying the two methods are in reasonable agreement with each other. A comparison to snow thicknesses from two modelled reanalysis data sets (ECMWF-ORAS5, DMI-HYCOM-CICE) shows some similarity in the regional pattern, although the absolute values differ.

We observe a large inter-annual variability in snow thicknesses across the Arctic sea ice areas. Regions of multi-year sea ice tend to have the thickest snow cover, as expected since multi-year ice allows snow to collect during the entire winter period. Averaging over the total Arctic, the spring snow depths over the last five years are significantly less than the Warren climatologic mean. This can, at least partly, be explained as result of the decreasing areas of multi-year ice.

The high inter-annual variability in spring snow-cover thicknesses impacts the year-to-year onset of the sea-ice melt period. In addition, it emphasizes the need for using daily snow depth observations, instead of snow depth climatologies, when correcting for the snow layer in satellite altimetry retrievals of sea ice thicknesses.

Sea-ice concentration (SIC) is a key variable to observe for understanding and monitoring the polar oceans. Satellite remote sensing, and especially passive microwave radiometry (PMR), have allowed the sub-daily, all-weather, pan-polar mapping of SIC for the last 40 years, enabling both climate monitoring and operational met-ocean applications.

Considering the lack of PMR capability in its Earth Observation programme Copernicus, and the absence of adequate, firmly committed, operational missions by other European or international agencies, the European Union (EU) tasked the European Space Agency (ESA) to study the design of an advanced PMR mission -the Copernicus Imaging Microwave Radiometer (CIMR)-, in support for its Integrated Arctic Policy. A candidate for the Expansion phase of the Copernicus Space Component (CSC), CIMR would fly on a Sentinel satellite at the horizon 2025. CIMR is a conically scanning PMR mission, with frequency bands covering L- to Ka-band (1.4 to 37 GHz), at much higher spatial resolution than past or future missions (see Figure). More information on CIMR can be found on https://cimr.eu.

In this contribution, we design and test a (set of) new SIC retrieval algorithm(s) for CIMR. Our retrieval algorithms stem from those successfully used by the EUMETSAT Ocean and Sea Ice Satellite Application Facility (OSI SAF) service, and further developed in the ESA Climate Change Initiative (CCI) projects with today’s PMR sensors, and are here extended to cover the specificities of CIMR, and especially explore the unique new combination of frequency channels.

A test-bench for PMR-based SIC algorithms is built. In the test-bench, algorithms are assessed both in terms of accuracy (standard deviation of retrieval noise), and achievable spatial resolution. To control these two characteristics simultaneously is essential, since high accuracy is often achieved at the expense of poor spatial resolution. Accuracy of CIMR SICs is assessed against signatures of known 100% and 0% SIC conditions, as gathered in the Round Robin Data Package (RRDP) of the ESA CCI Sea Ice project. Spatial resolution is assessed against simulated images of both idealized and realistic sea-ice scenes.

Using the test-bench, we conclude on which channel combinations have the highest potential for SIC retrieval from CIMR data, and how this capability compares to accuracy and resolution of other PMR missions.