Comment on tc-2021-353

The Los Alamos sea ice model (CICE) is being tested in standalone mode to identify biases that limit its suitability for seasonal prediction. The prescribed atmospheric forcings to drive CICE are from the NCEP Climate Forecast System Reanalysis (CFSR). A built-in mixed layer ocean model in CICE is used. Initial conditions for the sea ice and the mixed layer ocean are also from CFSR in the control experiments. The simulated sea ice extent is generally in good agreement with observations in the warm season at all lead times up to 12 months both in the Arctic and Antarctic, suggesting that CICE is able to provide useful sea ice edge information for seasonal prediction. However, the Arctic sea ice thickness forecast has a positive bias stemming from the initial conditions, and this bias often persists for more than a season, limiting the model’s seasonal forecast skill. When this bias is reduced by initializing ice thickness using the CryoSat-2 satellite observations while keeping all other initial fields unchanged in the CS2_IC experiments, both simulated ice edge and thickness improve. This confirms the important roles of sea ice thickness initialization in sea ice seasonal prediction seen in many studies.

In the past decade groundbreaking new satellite observations of the Arctic sea ice cover have been made, allowing researchers to understand the state of the Arctic sea ice system in greater detail than before. The derived estimates of sea ice thickness are useful but limited in time and space. In this study the first results of a new sea ice data assimilation system are presented. Observations assimilated (in various combinations) are monthly mean sea ice thickness and monthly mean sea ice thickness distribution from Cryosat-2, and NASA daily Bootstrap sea ice concentration. This system couples the Centre for Polar Observation and Modelling's (CPOM) version of the Los Alamos Sea Ice Model (CICE) to the Localised Ensemble Transform Kalman Filter (LETKF) from the Parallel Data Assimilation Framework (PDAF) library. The impact of assimilating a sub-grid scale sea ice thickness distribution is of particular novelty. The sub-grid scale sea ice thickness distribution is a fundamental component of sea ice models, playing a vital role in the dynamical and thermodynamical processes, yet very little is known of its true state in the Arctic. This study finds that assimilating Cryosat-2 products for the mean thickness and the sub-grid scale thickness distribution can have significant consequences on the modelled distribution of the ice thickness across the Arctic and particularly in regions of thick multi-year ice. The assimilation of sea ice concentration, mean sea ice thickness and sub-grid scale sea ice thickness distribution together performed best when compared to a subset of Cryosat-2 observations held back for validation. Regional model biases are reduced: the thickness of the thickest ice in the Canadian Archipelago is decreased, but the thickness of the ice in the Central Arctic is increased. When comparing the assimilation of mean thickness with the assimilation of sub-grid scale thickness distribution, it is found that the latter leads to a significant change in the volume of ice in each category. Estimates of the thickest ice improve significantly with the assimilation of sub-grid scale thickness distribution alongside mean thickness.

In the past decade groundbreaking new satellite observations of the Arctic sea ice cover have been made, allowing researchers to understand the state of the Arctic sea ice system in greater detail than before. The derived estimates of sea ice thickness are useful but limited in time and space. In this study the first results of a new sea ice data assimilation system are presented. Observations assimilated (in various combinations) are monthly mean sea ice thickness and monthly mean sea ice thickness distribution from Cryosat-2, and NASA daily Bootstrap sea ice concentration. This system couples the Centre for Polar Observation and Modelling's (CPOM) version of the Los Alamos Sea Ice Model (CICE) to the Localised Ensemble Transform Kalman Filter (LETKF) from the Parallel Data Assimilation Framework (PDAF) library. The impact of assimilating a sub-grid scale sea ice thickness distribution is of particular novelty. The sub-grid scale sea ice thickness distribution is a fundamental component of sea ice models, playing a vital role in the dynamical and thermodynamical processes, yet very little is known of its true state in the Arctic. This study finds that assimilating Cryosat-2 products for the mean thickness and the sub-grid scale thickness distribution can have significant consequences on the modelled distribution of the ice thickness across the Arctic and particularly in regions of thick multi-year ice. The assimilation of sea ice concentration, mean sea ice thickness and sub-grid scale sea ice thickness distribution together performed best when compared to a subset of Cryosat-2 observations held back for validation. Regional model biases are reduced: the thickness of the thickest ice in the Canadian Archipelago is decreased, but the thickness of the ice in the Central Arctic is increased. When comparing the assimilation of mean thickness with the assimilation of sub-grid scale thickness distribution, it is found that the latter leads to a significant change in the volume of ice in each category. Estimates of the thickest ice improve significantly with the assimilation of sub-grid scale thickness distribution alongside mean thickness.

To increase our capability to predict Arctic sea ice and climate, we have developed a coupled atmosphere‐sea ice‐ocean model configured for the pan‐Arctic with sufficient flexibility. The Los Alamos Sea Ice Model is coupled with the Weather Research and Forecasting Model and the Regional Ocean Modeling System in the Coupled Ocean‐Atmosphere‐Wave‐Sediment Transport modeling system. It is well known that dynamic models used to predict Arctic sea ice at short‐term periods strongly depend on model initial conditions. Parallel Data Assimilation Framework is implemented into the new modeling system to assimilate sea ice observations and generate skillful model initialization, which aid in the prediction procedures. The Special Sensor Microwave Imager/Sounder sea ice concentration, the CyroSat‐2, and Soil Moisture and Ocean Salinity sea ice thickness are assimilated with the localized error subspace transform ensemble Kalman filter. We conduct Arctic sea ice prediction for the melting seasons of 2017 and 2018. Predictions with improved initial sea ice conditions show reasonable sea ice evolution and small biases in the minimum sea ice extent, although the ice refreezing is delayed. Our prediction experiments suggest that the use of appropriate uncertainty for the observed sea ice thickness can lead to improved spatial distribution of the initial ice thickness and thus the predicted sea ice distribution. Our new modeling system initialized by the output of the National Centers for Environmental Prediction Climate Forecast System seasonal forecasts with data assimilation can significantly increase the sea ice prediction skills in sea ice extent for the entire Arctic as well as in the Northern Sea Route compared with the predictions by the National Centers for Environmental Prediction Climate Forecast System.

At the National Centers for Environmental Prediction (NCEP), a reanalysis of the atmosphere, ocean, sea ice and land over the period 1979–2009, referred to as the climate forecast system reanalysis (CFSR), was recently completed. The oceanic component of CFSR includes many advances: (a) the MOM4 ocean model with an interactive sea-ice, (b) the 6 h coupled model forecast as the first guess, (c) inclusion of the mean climatological river runoff, and (d) high spatial (0.5° × 0.5°) and temporal (hourly) model outputs. Since the CFSR will be used by many in initializing/validating ocean models and climate research, the primary motivation of the paper is to inform the user community about the saline features in the CFSR ocean component, and how the ocean reanalysis compares with in situ observations and previous reanalysis. The net ocean surface heat flux of the CFSR has smaller biases compared to the sum of the latent and sensible heat fluxes from the objectively analyzed air-sea fluxes (OAFlux) and the shortwave and longwave radiation fluxes from the International Satellite Cloud Climatology Project (ISCCP-FD) than the NCEP/NCAR reanalysis (R1) and NCEP/DOE reanalysis (R2) in both the tropics and extratropics. The ocean surface wind stress of the CFSR has smaller biases and higher correlation with the ERA40 produced by the European Centre for Medium-Range Weather Forecasts than the R1 and R2, particularly in the tropical Indian and Pacific Ocean. The CFSR also has smaller errors compared to the QuickSCAT climatology for September 1999 to October 2009 than the R1 and R2. However, the trade winds of the CFSR in the central equatorial Pacific are too strong prior to 1999, and become close to observations once the ATOVS radiance data are assimilated in late 1998. A sudden reduction of easterly wind bias is related to the sudden onset of a warm bias in the eastern equatorial Pacific temperature around 1998/1999. The sea surface height and top 300 m heat content (HC300) of the CFSR compare with observations better than the GODAS in the tropical Indian Ocean and extratropics, but much worse in the tropical Atlantic, probably due to discontinuity in the deep ocean temperature and salinity caused by the six data streams of the CFSR. In terms of climate variability, the CFSR provides a good simulation of tropical instability waves and oceanic Kelvin waves in the tropical Pacific, and the dominant modes of HC300 that are associated with El Nino and Southern Oscillation, Indian Ocean Dipole, Pacific Decadal Oscillation and Atlantic Meridional Overturning Circulation.

This study was motivated by the potential development of offshore oil exploration leases in the Canadian Southern Beaufort Sea, an area within the Inuvialuit Settlement Region. Sea ice concentration, extent, motion, and thickness data are vital to the success of potential oil operations in this region, and relevant data cannot be gleaned from larger-scale hemispheric studies. We therefore undertook regionally specific sea ice analyses in the southern Beaufort Sea during the summer drilling season (July, August, and September) in 2009 and over the long-term (1996 – 2010). On average, the Canadian oil lease areas contain mostly old sea ice during the drilling season and have not experienced significant decreasing trends in total or old sea ice. The average sea ice motion in the region for the period was anti-cyclonic at 20 – 25 cm·s-1, acting to transport sea ice southward toward the lease areas. Summer 2009 was used as a case study of regional ice concentration, motion, and thickness and to compare September sea ice thickness measurements to data collected in April 2009. In the summer of 2009, old sea ice was the predominant ice type in the lease areas. Sea ice motion was anti-cyclonic and faster than the long-term average, reaching 60 cm·s-1 west of Banks Island and across the north end of the lease areas. September 2009 sea ice thickness (mean = 1.03 m, σ = 0.97 m) was modal about the 0.20 – 0.29 m thickness bin. The sea ice thickness distri­bution was spatially variable, with the thickest ice occurring at the north end of the study area, in an area dominated by high old ice concentrations. Ice thicknesses greater than 10 m (the upper limit our instruments could measure) were encountered. Thinner sea ice predominated at the periphery of the core Beaufort Sea multi-year pack. Near the oil lease areas, the sea ice thickness distributions were shifted left on the histogram in comparison to those farther north, resulting in a greater proportion of relatively thick sea ice due to the thermodynamic loss of thinner (< 1.5 m) first-year ice during its southward movement. After enduring a summer’s melt, however, this thicker ice at the south end of the study region had thinned in comparison to the ice at the north end.

We present a first assessment of airborne laser and radar altimeter data over snow-covered sea ice, gathered during the National Aeronautics and Space Administration Operation IceBridge Mission. We describe a new technique designed to process radar echograms from the University of Kansas snow radar to estimate snow depth. We combine IceBridge laser altimetry with radar-derived snow depths to determine sea ice thickness. Results are validated through comparison with direct measurements of snow and ice thickness collected in situ at the Danish GreenArc 2009 sea ice camp located on fast ice north of Greenland. The IceBridge instrument suite provides accurate measurements of snow and ice thickness, particularly over level ice. Mean IceBridge snow and ice thickness agree with in situ measurements to within ~ 0.01 and ~ 0.05 m, respectively, while modal snow and ice thickness estimates agree to within 0.02 and 0.10 m, respectively. IceBridge snow depths were correlated with in situ measurements (R = 0.7, for an averaging length of 55 m). The uncertainty associated with the derived IceBridge sea ice thickness estimates is 0.40 m. The results demonstrate the retrieval of both first-year and multiyear ice thickness from IceBridge data. The airborne data were however compromised in heavily ridged ice where snow depth, and hence ice thickness, could not be measured. Techniques developed as part of this study will be used for routine processing of IceBridge retrievals over Arctic sea ice. The limitations of the GreenArc study are discussed, and recommendations for future validation of airborne measurements via field activities are provided.

The sensitivity of seasonal prediction (June to September) of Indian monsoon to initial state from two variants of coupled data assimilation (CDA) products, viz. the Climate Forecast System (CFS) Reanalysis (CFSR) and Indian Institute of Tropical Meteorology, University of Maryland‐ Weakly Coupled Analysis (IWCA) is explored in this study. The IWCA implements the local ensemble transform Kalman filter, and incorporates theoretically advanced features of flow‐dependency and ensemble‐based analysis compared to CFSR. The CFS version‐2 predictions using IWCA simulate the large‐scale monsoon features, and convection centers well, and improve prediction skills compared to CFSR predictions. The enhanced analysis quality and Ocean‐Atmospheric cross‐domain equilibrium in IWCA reduce initial shocks in springtime predictions. Further, the sustained ensemble consistency aided to simulate the variability better and improved the seasonal predictions. The study strongly advocates the adaptation of advanced CDA methods for seasonal monsoon and probable seamless predictions.

Inter-comparison of extra-tropical cyclone activity in nine reanalysis datasets

Ship‐based observations are used to describe regional and seasonal changes in the thickness distribution and characteristics of sea ice and snow cover thickness around Antarctica. The data set comprises 23,373 observations collected over more than 2 decades of activity and has been compiled as part of the Scientific Committee on Antarctic Research (SCAR) Antarctic Sea Ice Processes and Climate (ASPeCt) program. The results show the seasonal progression of the ice thickness distribution for six regions around the continent together with statistics on the mean thickness, surface ridging, snow cover, and local variability for each region and season. A simple ridge model is used to calculate the total ice thickness from the observations of level ice and surface topography, to provide a best estimate of the total ice mass, including the ridged component. The long‐term mean and standard deviation of total sea ice thickness (including ridges) is reported as 0.87 ± 0.91 m, which is 40% greater than the mean level ice thickness of 0.62 m. Analysis of the structure function along north/south and east/west transects revealed lag distances over which sea ice thickness decorrelates to be of the order of 100–300 km, which we use as a basis for presenting near‐continuous maps of sea ice and snow cover thickness plotted on a 2.5° × 5.0° grid.

&amp;lt;p&amp;gt;Arctic sea ice has experienced dramatic changes for the past few decades. Recent changes in the properties of Arctic sea ice have posed significant challenges to the research community to provide sea ice predictions. To improve our capability to predict Arctic sea ice and climate, we have developed a coupled atmosphere-sea ice-ocean model configured for the Arctic with sufficient flexibility. The Los Alamos sea ice model is coupled with the Weather Research and Forecasting (WRF) Model and the Regional Ocean Modeling System (ROMS) within the Coupled-Ocean-Atmosphere-Wave-Sediment Transport (COAWST) modeling system. A series of sensitivity experiments with different physics options have been performed to determine the &amp;amp;#8216;optimal&amp;amp;#8217; physics configuration that provides reasonable simulation of Arctic sea ice.&amp;lt;/p&amp;gt;&amp;lt;p&amp;gt;It is well known that dynamic models used to predict Arctic sea ice at short-term periods strongly depend on model initial conditions. Thus, a data assimilation that integrates sea ice observations to generate realistic and skillful model initialization is needed to improve predictive skill of Arctic sea ice. Parallel Data Assimilation Framework has been implemented into the new modeling system to assimilate SSMIS sea ice concentration, and CyroSat-2 and SMOS sea ice thickness using a localized error subspace transform ensemble Kalman filter (LESTKF). We have conducted Arctic sea ice predictions for the melting seasons of 2017 and 2018. Predictions with improved initial sea ice states show reasonably accurate sea ice evolution and small biases in the minimum sea ice extent.&amp;lt;/p&amp;gt;&amp;lt;p&amp;gt;Storms-induced ocean surface waves are capable of breaking pack ice into smaller floes and changing the sea ice melting rate. We have also coupled the Simulating Wave Nearshore (SWAN) with above atmosphere-sea ice-ocean coupled system and examined the impacts of wave-ice interactions on sea ice simulation. Preliminary results suggest ocean waves have direct and indirect impacts on sea ice. Direct impacts are the fracturing of ice pack and indirect impacts the change of ocean thermo-structure through the wave breaking.&amp;lt;/p&amp;gt;

Light transmission through a sea ice cover has strong implications for the heat content of the upper ocean, the magnitude of bottom and lateral ice melt, and primary productivity in the ocean. Light transmittance in the vicinity of the Multidisciplinary Drifting Observatory for the Study of Arctic Climate (MOSAiC) Central Observatory was estimated by driving a two-stream radiative transfer model with physical property observations. Data include point and transect observations of snow depth, surface scattering layer thickness, ice thickness, and pond depth. The temporal evolution of light transmittance at specific sites and the spatial variability along transect lines were computed. Ponds transmitted 4–6 times as much solar energy per unit area as bare ice. On July 25, ponds covered about 18% of the area and contributed roughly 50% of the sunlight transmitted through the ice cover. Approximating the transmittance along a transect line using average values for the physical properties will always result in lower light transmittance than finding the average light transmittance using the full distribution of points. Transmitted solar energy calculated using the standard five ice thickness categories and three surface types used in the Los Alamos sea ice model CICE, the sea ice component of many weather and climate models, was only about 1 W m−2 less than using all the points along the transect. This minor difference suggests that the important processes and resulting feedbacks relating to solar transmittance can be represented in models that use five or more categories of ice thickness distributions.

&amp;lt;p&amp;gt;The MOSAiC expedition took place in the Arctic from September 2019 to October 2020 while having measurements under, in, and above sea ice for a complete annual cycle. Airborne thermal infrared imaging was conducted during 41 helicopter survey flights along the MOSAiC drift track. We analyze the infrared brightness temperature of snow, sea ice, and ocean water surfaces from October 2019 until May 2020 from the airborne measurements. While the snow-covered sea ice appears very cold, thin ice and open water are significantly warmer. These surface types will be considered with particular attention because they dominate the heat exchange between the ocean, ice, and atmosphere during wintertime. This influences the Arctic Climate and becomes even more important in the currently changing Arctic, where the sea ice gets thinner, moves faster, and breaks up easier. After georeferencing and merging the recorded images to a mosaic, we can provide maps of infrared brightness temperatures in a high spatial resolution of 1 m for each flight. The spatial range of the maps varies from local (~5 km) up to regional (~30 km). This data set provides a basis to study the spatial and temporal variability of sea ice characteristics in the Arctic winter. We derive the physical surface temperature from the brightness temperature, surface emissivity, and downwelling radiation from the sky or clouds. Using the surface temperature, we calculate the heat flux from a local up to a regional scale based on thermodynamic assumptions and atmospheric measurements on the ice floe. From more complex thermodynamic simulations, we estimate ice thickness and ice age based on the airborne measured surface temperatures. The model calculates for each surface temperature a specific ice thickness and heat flux based on the knowledge about the surface&amp;amp;#8217;s thermodynamic history. The simulated ice thickness allows a sea ice classification which is compared to our first classification approach which deals with the flight's temperature distribution only. In the future, we will investigate the sub-footprint scale variability of ice surface temperature and thin ice thickness for satellite data, e.g. MODIS and Sentinel-3.&amp;lt;/p&amp;gt;

Abstract. Local analytical optimal nudging (LAON) is introduced and thoroughly evaluated for assimilating the Advanced Microwave Scanning Radiometer 2 (AMSR2) sea ice concentration (SIC) in the Norwegian High-resolution pan-Arctic ocean and sea ice Prediction System (NorHAPS). NorHAPS is a developing high-resolution (3–5 km) pan-Arctic coupled ocean and sea ice modeling and prediction system based on the HYbrid Coordinate Ocean Model (HYCOM version 2.2.98) and the Los Alamos multi-category sea ice model (CICE version 5.1.2), with the LAON for data assimilation. In this study, our focus is on the LAON assimilation of AMSR2 SIC, which is designed to update the model SIC in every time step such that the analysis will eventually reach the optimal estimate. The SIC innovation (observation minus model) is designed to be proportionally distributed to the multiple sea ice categories. A hindcast experiment is performed with and without the LAON assimilation for the period 1 January 2021 to 30 April 2022, in which the extra computational cost for the LAON assimilation is about 5 % of the free run without assimilation. The results show that the LAON assimilation greatly improves the simulated sea ice concentration, extent, area, thickness, and volume, as well as the sea surface temperature (SST). It also produces significantly more accurate sea ice edge and marginal zone (MIZ) than the observed AMSR2 SIC that is assimilated when evaluated against the Norwegian Ice Service (NIS) ice chart. The results are also compared with the Copernicus Marine Environment Monitoring Service (CMEMS) operational SIC analyses from NEMO, TOPAZ4, and neXtSIM, which use ensemble Kalman filters and direct insertion for data assimilation. It is shown that the LAON assimilation produces significantly lower integrated ice edge error (IIEE) and integrated MIZ error (IME) than the CMEMS SIC analyses when evaluated against the NIS ice chart. LAON also produces a continuous and smooth evolution of sub-daily SIC, which avoids abrupt jumps often seen in other assimilated products. This efficient and accurate method is promising for data assimilation in global and high-resolution models.

A hierarchy of sea ice rheologies is evaluated on the basis of a comprehensive set of observational data. The investigations are part of the Sea Ice Model Intercomparison Project (SIMIP). Four different sea ice rheology schemes are compared: a viscous‐plastic rheology, a cavitating‐fluid model, a compressible Newtonian fluid, and a simple free drift approach with velocity correction. The same grid, land boundaries, and forcing fields are applied to all models. As verification data, there are (1) ice thickness data from upward looking sonars (ULS), (2) ice concentration data from the passive microwave radiometers SMMR and SSM/I, (3) daily buoy drift data obtained by the International Arctic Buoy Program (IABP), and (4) satellite‐derived ice drift fields based on the 85 GHz channel of SSM/I. All models are optimized individually with respect to mean drift speed and daily drift speed statistics. The impact of ice strength on the ice cover is best revealed by the spatial pattern of ice thickness, ice drift on different timescales, daily drift speed statistics, and the drift velocities in Fram Strait. Overall, the viscous‐plastic rheology yields the most realistic simulation. In contrast, the results of the very simple free‐drift model with velocity correction clearly show large errors in simulated ice drift as well as in ice thicknesses and ice export through Fram Strait compared to observation. The compressible Newtonian fluid cannot prevent excessive ice thickness buildup in the central Arctic and overestimates the internal forces in Fram Strait. Because of the lack of shear strength, the cavitating‐fluid model shows marked differences to the statistics of observed ice drift and the observed spatial pattern of ice thickness. Comparison of required computer resources demonstrates that the additional cost for the viscous‐plastic sea ice rheology is minor compared with the atmospheric and oceanic model components in global climate simulations.