The present study investigates several factors related to the assimilation of surface-sensitive radiance observations over land from the Advanced Microwave Sounding Unit-A within a non-cycling one-dimensional Ensemble-Variational (1D-EnVar) framework, with the aim of informing future improvements to numerical weather prediction (NWP) at Environment and Climate Change Canada (ECCC). These observations are currently not assimilated in ECCC's operational systems, since it requires accurate knowledge of both surface emissivity and skin temperature. Typically, surface emissivity is retrieved using a surface-sensitive channel observation combined with prior knowledge of the atmospheric state and skin temperature through an inverse radiative transfer calculation. The present study examines the impact of assimilating multiple surface-sensitive channels to simultaneously estimate surface emissivity, together with the standard set of state variables (including skin temperature), by including surface emissivity as an additional analysis variable in a 1D-EnVar system. In addition, other related factors are examined, including modifications to the topographic criteria in the observation quality control to allow more channel 6 observations over higher-elevation surfaces to be assimilated as surface-sensitive channels. Although these changes lead to analysis improvements in the mid-troposphere, they also result in an increased analysis error in the lower troposphere, caused by unexpectedly large lower-tropospheric air temperature and skin temperature increments. This increase in error is mitigated by modifying the ensemble-based background error covariances. Lastly, given the limited knowledge on the true surface emissivity uncertainty, the impact of varying the emissivity background error covariances is examined. The combined impact of integrating the modifications from all factors while assimilating channels 4-14 resulted in up to a 3.4% reduction in the normalized root-mean-squared error for the mid- and lower-tropospheric temperature compared with assimilating only channels 6-14 and without considering the discussed factors. These findings provide a basis for future advancements in assimilating surface-sensitive microwave radiances over land in 4D-EnVar in the full ECCC NWP system.

In cold regions, snow serves as the primary water source for downstream rivers and lakes. Accurate gridded snow water equivalent (SWE) estimation is hindered by the sparse ground observation network and the low resolution of satellite passive microwave products. To address this, Environment and Climate Change Canada (ECCC), the Canadian Space Agency (CSA), and Natural Resources Canada (NRCan) are developing the Terrestrial Snow Mass Mission (TSMM), a dual Ku-band satellite mission designed to measure backscatter at 13.5 and 17.25 GHz across the Northern Hemisphere at a 500 m spatial resolution with a weekly temporal resolution. This study assesses the feasibility of assimilating Ku-band backscatter to enhance SWE estimates in a synthetic experiment. We used the Soil-Vegetation-Snow version 2 (SVS2) land surface model, which incorporates the snowpack model Crocus, coupled with the Snow Microwave Radiative Transfer model (SMRT). Synthetic observations of SWE and backscatter extracted at weekly intervals from synthetic truths (model simulations) were assimilated with a particle filter at point-scale. This was done at three sites representing three different Canadian climates (Arctic, humid continental, Alpine) over three winter seasons. Meteorological forcing derived from the high-resolution Canadian meteorological model was perturbed to generate ensembles of snow simulations for assimilation. Results indicate that assimilating synthetic observations of backscatter improved SWE estimates at the Arctic and humid continental sites, reducing the mean continuous ranked probability score (CRPS) by up to 32% compared to the open-loop ensemble. This performance was comparable to assimilating the SWE synthetic observations with observation errors larger than 20 %. As similating synthetic observations of backscatter at the Alpine site only improved the SWE estimates by 5 % as backscatter signals seemed to lose sensitivity to SWE values greater than similar to 300 kgm-2 in our experimental setup. Assimilating backscatter and SWE synthetic observations also improved the estimations of vertical profiles of snow density and specific surface area. These findings demonstrate the potential of direct assimilation of Ku-band backscatter to enhance both estimates of SWE and snowpack properties.

In this study, we derived pan-Arctic multi-year (MYI) and first year sea ice (FYI) retrievals at 1.6 km resolution under no melt condition using the RADARSAT Constellation Mission (RCM) ScanSAR dual-polarization HH-HV synthetic aperture radar (SAR) data. The RCM products consistently provide fine-scale MYI/FYI details that are not resolved by low-resolution passive microwave observations. Verification of the RCM MYI/FYI products against the corresponding Canadian Ice Service image analyses revealed that 60.28% of MYI/FYI image analysis samples were classified with a very high accuracy of 99.58% over 2020-2025. We also generated pan-Arctic RCM MYI/FYI gridded products at 2 km resolution aggregated over 3- and 7-day rolling temporal windows from 2022 to the present. We found the RCM MYI/FYI products to be in good agreement with existing passive microwave datasets. The proportion correct total scores were 0.90 and 0.69 when compared against daily Ocean and Sea Ice Satellite Application Facility and weekly National Snow and Ice Data Center products, respectively. The RCM high-resolution MYI/FYI retrievals are being assimilated in the Environment and Climate Change Canada ice type data assimilation system that efficiently combines them with low-resolution passive microwave and scatterometer observations.

It has been documented that spread-error equality and a flat rank histogram are necessary but insufficient for demonstrating ensemble forecast reliability. Nevertheless, these metrics are heavily relied upon-both in the literature and at operational numerical weather prediction centers-as if they were valid indicators of perfect ensemble dispersion. In this study, we show that, in the presence of a climatological variance bias, several reliability diagnostics can indicate falsely that forecasts are perfectly reliable. This is demonstrated theoretically for the spread-error relationship and related reliability budget. Idealized experiments based on the multivariate normal distribution reveal that the same joint-distribution structure causing insufficiency also leads to false diagnoses of reliability with the rank histogram and the reliability component of the continuous rank probability score. Under this structure, and when the ensemble-mean state is meaningfully different from climatology, the truth lies systematically among the least extreme members when climatological variance is excessive in each member, and among the most extreme members when climatological variance is deficient. Importantly, this behavior is also shown to be plausible in operational ensemble weather forecasts. Combining these results with calibration principles from statistical postprocessing leads us to conclude that both "perfect dispersion" and "underdispersion" are ill-defined. When diagnostics are misinterpreted as indicating the latter, improper tuning of forecasts can lead to further deterioration of forecast quality, even while improving spread-error and rank histogram behavior. To address these issues, we propose a reliability diagnostic based on three easily computed statistics, directly related to the structure of the joint distribution of ensemble members and the reference truth up to second order. The diagnostic separates contributions to unreliability originating from climatology and predictability, enabling a more precise and robust characterization of ensemble behavior.

Operational weather forecasting has traditionally relied on physics-based numerical weather prediction (NWP) models, but the rise of AI-based weather emulators is reshaping this paradigm. However, most data-driven models for medium-range forecasting still face limitations, such as a narrow range of predicted variables and low effective spatiotemporal resolution. This presentation will compare the strengths and weaknesses of these two approaches, using Environment and Climate Change Canada’s Global Environmental Multiscale (GEM) model and Google DeepMind’s GraphCast model. It will demonstrate that GraphCast outperforms GEM in predicting large-scale features, particularly for longer lead times.Building on these findings, we propose a new hybrid NWP-AI system, in which GEM’s large-scale state variables are spectrally nudged towards GraphCast’s inferences, while GEM continues to generate fine-scale details critical for weather extremes. Results show that this hybrid system improves GEM’s forecast accuracy, reducing RMSE for the 500-hPa geopotential height by 5-10% and extending predictability by 6-12 hours in the extratropics, peaking at day 7 of the forecast. It also yields significant improvements in tropical cyclone trajectory prediction without degrading intensity forecasts. Unlike state-of-the-art AI-based models, the hybrid system ensures meteorologists retain access to all forecast variables, including those critical for high-impact weather. Preparations are currently well underway for the operationalization of this hybrid system at the Canadian Meteorological Centre.

Most ensemble verification diagnostics are sensitive to ensemble size, complicating the evaluation of a system’s underlying quality and the comparison of different ensemble systems. This study examines how the Mean Squared Error (MSE) of the ensemble mean, and the Spread-Error relationship used to evaluate ensemble consistency, vary as a function of ensemble size. As the MSE of the ensemble mean (“Error” in Spread-Error) is affected by ensemble size, but the average sample ensemble variance (“Spread” in Spread-Error) is not, these effects must be removed from the MSE for a robust assessment of ensemble consistency. Although the dependence of these diagnostics on ensemble size has been sparsely addressed over several decades, gaps remain concerning the assumptions necessary for quantification. Evidence also suggests these effects are not widely known or fully understood. The impact of ensemble size is examined by assuming exchangeability between ensemble members, allowing us to derive the MSE and Spread-Error relationship (expressed as a difference) that would be obtained with an infinite-sized ensemble. Ensemble-size effects are removed from both scores by subtracting the average ensemble variance divided by the ensemble size. The unbiased MSE can be used to estimate the error reduction achievable by increasing ensemble size and allows for an “apples-to-apples” comparison of forecast error across systems. The unbiased Spread-Error relationship eliminates the effects of ensemble size on the original diagnostic, which, when ignored, make ensembles appear too underdispersive.

The Modular and Integrated Data Assimilation System (MIDAS) software (version 3.9.1) is described in terms of its range of functionality, modular software design, parallelization strategy, and current uses within real-time operational and experimental systems. MIDAS is developed at Environment and Climate Change Canada for both operational and research applications, including all atmospheric data assimilation (DA) elements of the Canadian operational numerical weather prediction systems. The described version of MIDAS is part of the Canadian prediction systems that became operational in June 2024. The software is designed to be sufficiently general to enable other DA applications, including atmospheric constituents (e.g. ozone), sea ice, and sea surface temperature. In addition to describing the current MIDAS applications, a sample of the results from these systems is presented to demonstrate their performance in comparison with either systems from before the switch to using MIDAS software or similar systems at other numerical weather prediction (NWP) centres. The modular software design also allows the code that implements high-level components (e.g. observation operators, error covariance matrices, state vectors) to easily be used in many different ways depending on the application, such as for both variational and ensemble DA algorithms, for estimating the observation impact on short-term forecasts, and for performing various observation pre-processing procedures. The use of a single common DA software package for multiple components of the Earth system provides both practical and scientific benefits, including the facilitation of future research on DA approaches that explicitly include the coupled connections between multiple Earth system components. To this end, work is currently underway to allow the use of MIDAS DA algorithms for initializing both deterministic and ensemble three-dimensional ocean model forecasts.

Arctic sea ice type information is essential for various operational and scientific applications including the support of marine users and guiding ice thickness retrieval algorithms operating with SMOS and CryoSat-2 data for improved sea ice prediction. A sea ice type analysis system developed at Environment and Climate Change Canada’s (ECCC) generates pan-Arctic ice type analyses at 5 km resolution every 6 hours. The ice type analysis system assimilates ice type information provided by passive microwave (AMSR2, SSMIS) and scatterometer (ASCAT) data, but assimilation of these observations is not reliable in the areas near land and in the narrow channels of the Canadian Arctic Archipelago due to their low spatial resolution of ~20-50 km. Therefore, assimilation of high-resolution ice type observations from synthetic aperture radar (SAR) is highly desired. In this study, we extended our approach for automated detection of winter multi-year ice (MYI) and first-year ice (FYI) at 1.6 km scale from RADARSAT-2 to RCM data. To this end, we collected more than 2,000 RCM images and corresponding image analyses products that were manually generated by the Canadian Ice Service (CIS) ice experts for the period of time between July 1, 2020 and July 31, 2023. From these RCM images we extracted SAR information for more than 30,000 pure MYI and more than 619,000 pure FYI data samples under no melt conditions as identified by the CIS image analyses. We demonstrated that separability measures for MYI and FYI classes in the spaces of the two predictor parameters (HV/HH polarization ratio and standard deviation of HV signal) are consistent with those we previously observed for RADARSAT-2. Furthermore, we found that our RCM-based MYI/FYI detection approach allows us to classify 60% of the winter CIS ice data samples with the accuracy of 99.6%. Preliminary results of assimilating RCM-based MYI/FYI high-resolution retrievals in combination with passive microwave and scatterometer data in the ECCC ice type analysis system will be also presented.

Operational meteorological forecasting has long relied on physics-based numerical weather prediction (NWP) models. Recently, this landscape has faced disruption by the advent of data-driven artificial intelligence (AI)-based weather models, which offer tremendous computational performance and competitive forecasting accuracy. However, data-driven models for medium-range forecasting generally suffer from major limitations, including low effective resolution and a narrow range of predicted variables. This study illustrates the relative strengths and weaknesses of these competing paradigms using the physics-based GEM (Global Environmental Multiscale) and the AI-based GraphCast models. Analyses of their respective global predictions in physical and spectral space reveal that GraphCast-predicted large scales outperform GEM, particularly for longer lead times, even though fine scales predicted by GraphCast suffer from excessive smoothing. Building on this insight, a hybrid NWP-AI system is proposed, wherein temperature and horizontal wind components predicted by GEM are spectrally nudged toward GraphCast predictions at large scales, while GEM itself freely generates the fine-scale details critical for local predictability and weather extremes. This hybrid approach is capable of leveraging the strengths of GraphCast to enhance the prediction skill of the GEM model while generating a full suite of physically consistent forecast fields with a full power spectrum. Additionally, trajectories of tropical cyclones are predicted with enhanced accuracy without significant changes in intensity. Work is in progress for operationalization of this hybrid system at the Canadian Meteorological Centre.

The assimilation of surface-sensitive microwave radiance observations over land can improve numerical weather prediction accuracy at Environment and Climate Change Canada. However, the benefits of these observations are limited by large errors in surface emissivity and skin temperature, along with challenges accounting for these errors in the data assimilation system. Typically, the contribution of surface emissivity error is treated as a deficiency in the observation operator and represented by inflating a fixed observation error covariance matrix (R). However, this study demonstrates the error contributions of surface emissivity in observation space for each radiance observation profile can vary substantially due to variations in the surface emissivity Jacobians, even if the surface emissivity error covariances are fixed. An approach is introduced to account for this variability by representing the surface emissivity errors in the background error covariance matrix (B) and including surface emissivity as an additional analysis variable. Using an idealized one-dimensional variational assimilation framework, the approach is compared with a reference experiment that uses a fixed R to represent the average error contribution of surface emissivity. The proposed approach results in up to 20% and 35% greater error reduction in the mid-lower tropospheric air temperature and skin temperature analyses, respectively, when compared to the reference experiment. Both methods would be mathematically equivalent in estimating air and skin temperatures if the variations in the error contribution of surface emissivity were fully represented within the R in the reference approach, which may require significant technical changes to the assimilation system. However, additional benefits of the proposed approach could potentially be realized in a 4D-EnVar assimilation system by propagating the updated surface emissivity and within the outer loop of an incremental formulation. Additionally, it is demonstrated that well-represented background error correlations between air and skin temperatures in B can further improve the error reductions.

The sea and lake ice concentration pan-Arctic analysis system at Environment and Climate Change Canada (ECCC) initializes both the short-range Arctic sea ice forecasting models and numerical weather prediction tools. In this study, our previously developed approach for deriving ice concentration from RADARSAT-2 was adjusted to become applicable to the RADARSAT Constellation Mission (RCM) data. This algorithm adaptation was done by incorporating RCM beam-specific wind speed retrieval models and enhancing the quality control procedure. We found an excellent agreement (correlation = 0.996) between ice concentrations derived from nearly 4000 RCM images and ice concentrations from the image analyses manually produced by the Canadian Ice Service (CIS) over a one-year period between August 1, 2020 and July 31, 2021. Then, for data assimilation experiments in the ECCC ice analysis system, we extracted ice concentrations at 1.6 km resolution from >65,000 RCM images acquired over the same test time period and over the whole Arctic. We demonstrated that assimilation of RCM-derived observations significantly improves ice analyses especially over lakes, in the vicinity to the coasts, and marginal ice zones. Accuracy of the ice analyses in the areas that are not covered by the CIS charts such as Alaska, High Arctic, and Eurasian Arctic substantially increased after the assimilation of RCM data. Visual assessment of the ice analyses against optical data showed that high-resolution details in ice concentration fields were successfully introduced by the RCM retrievals. Operational implementation of near-real-time RCM data assimilation in the ice concentration analysis system that initializes the ECCC Regional Ice-Ocean Prediction System is underway.

A new global daily sea-surface temperature (SST) analysis system has been developed at Environment and Climate Change Canada (ECCC). All components of the new SST analysis system are implemented within the Modular and Integrated Data Assimilation System (MIDAS) software. MIDAS is already used for the data assimilation component of the main operational numerical weather prediction (NWP) systems at ECCC. The new SST analysis system, integrated together with the global sea-ice analysis, will be part of the combined ocean surface analysis used for all operational prediction systems at ECCC. The data assimilation method used to compute the new SST analyses is two-dimensional variational method with a diffusion operator for representing the horizontal background-error correlations. A new algorithm for satellite data bias estimation has also been developed employing gridded bias estimates computed from a spatial averaging of the differences between collocated satellite and in-situ data. New algorithms for quality control and thinning of satellite data have also been implemented, making each type of observational dataset more evenly distributed over the globe. The performance of the new SST system is examined relative to the current operational SST system by using independent data. The impact of using the new SST analysis within NWP and ocean prediction systems is also evaluated. When compared with the operational system currently in use, the experiments employing the new SST analysis system produce a nearly neutral impact on the NWP and ocean prediction systems. This validation of the new system is an important first step towards the ability to use MIDAS to perform ensemble-based three-dimensional ocean and coupled ocean-ice-atmosphere data assimilation. A new global daily sea-surface temperature analysis system has been developed at Environment and Climate Change Canada (ECCC) to replace the existing operational system. All components of the new system are implemented within the Modular and Integrated Data Assimilation System (MIDAS) software. MIDAS is already used for the data assimilation component of the main operational numerical weather prediction systems. Developing a future coupled data assimilation system at ECCC will be facilitated by implementing all components within a common framework. image

ABSTRACT This is a review article invited by Atmosphere-Ocean to document the contributions of Recherche en Prévision Numérique (RPN) to Numerical Weather Prediction (NWP). It is structured as a historical review and documents RPN’s contributions to numerical methods, numerical modelling, data assimilation, and ensemble systems, with a look ahead to potential future systems. Through this review, we highlight the evolution of RPN’s contributions. We begin with early NWP efforts and continue through to environmental predictions with a broad range of applications. This synthesis is intended to be a helpful reference, consolidating developments and generating broader interest for future work on NWP in Canada.

The Global Ensemble Prediction System (GEPS) of Environment and Climate Change Canada was recently upgraded to a coupled atmosphere, ocean, and sea‐ice version from an uncoupled atmosphere‐only system. This has been operational since July 2019, with over a year of forecasts now available to evaluate the system throughout all seasons. Using metrics that score the forecast error in ice‐edge position, the spatial probability score and the integrated ice‐edge error, we investigate the spread–error relationship in probabilistic Arctic sea‐ice forecasts from the system and compare this with the skill of the system relative to persistence and a companion Global Deterministic Prediction System (GDPS). Within this ensemble framework, we explore the advantages of having a probabilistic forecast and probe its usefulness in addressing the errors in the system. Both the ensemble GEPS and the deterministic GDPS systems show enhanced sea‐ice prediction over persistence in all months except May and June, when significant biases exist in the systems in shallow‐sea and shelf regions. We attribute a significant portion of these biases to problems modelling landfast ice, but other sources of bias, including significant uncertainties in initializing and verifying sea‐ice analysis, also contribute. The lowest errors in the systems are found during September and continue at reasonably low levels through much of the boreal winter. The minimum and maximum extent periods, along with the early freeze‐up period, are shown to be periods for which the ensemble system offers enhanced benefits over a single deterministic forecast. For these periods, the errors are low and strongly correlated spatially with the ensemble spread. Nevertheless, we find that the ensemble system would likely still benefit from further improvement of the spread/error relationship in the system, currently hampered due to ensemble perturbations that are produced solely in the atmospheric component.

In an ensemble Kalman filter, when the analysis update of an ensemble member is computed using error statistics estimated from an ensemble that includes the background of the member being updated, the spread of the resulting ensemble systematically underestimates the uncertainty of the ensemble mean analysis. This problem can largely be avoided by applying cross validation: using an independent subset of ensemble members for updating each member. However, in some circumstances cross validation can lead to the divergence of one or more ensemble members from observations. This can culminate in catastrophic filter divergence in which the analyzed or forecast states become unrealistic in the diverging members. So far, such instabilities have been reported only in the context of highly nonlinear low-dimensional models. The first known manifestation of catastrophic filter divergence caused by the use of cross validation in an NWP context is reported here. To reduce the risk of such filter divergence, a modification to the traditional cross-validation approach is proposed. Instead of always assigning the ensemble members to the same subensembles, the members forming each subensemble are randomly chosen at every analysis step. It is shown that this new approach can prevent filter divergence and also brings a cycling ensemble data assimilation system containing divergent members back to a state consistent with Gaussianity. The randomized subensemble approach was implemented in the operational global ensemble prediction system at Environment and Climate Change Canada on 1 December 2021.

New techniques for automated retrieval of ocean surface wind speed from the RADARSAT Constellation mission (RCM) without input wind direction have been developed and tested. We collected a large number of collocated and coincided ocean buoy wind measurements and RCM co- and cross-polarization observations acquired over the West and East coasts of Canada from February 1, 2020 to July 25, 2021: 1190 data points for RCM ScanSAR 50 m (SC50M) VV–VH, 3819 points for SC50M HH–HV, and 397 points for ScanSAR Low Noise (SCLN) HH–HV images. The observations captured a wide range of wind conditions with the maximum wind speeds of 22.3, 23.8, and 19.7 m/s for SC50M VV–VH, SC50M HH–HV, and SCLN HH–HV, respectively. For these three types of RCM data, the ocean surface wind speed was modeled as a function of the co-polarization (VV or HH) and cross-polarization backscatter (VH or HV) as well as the noise floor and the incidence angle. The models were tested on completely independent subsets of data, and their performance was compared against CMOD5.N (for VV) and CMODH (for HH) models which require input wind direction. The root-mean square errors (RMSEs) for the proposed models in the testing subsets are 1.20, 1.52, and 1.26 m/s, while CMOD models showed 1.37, 1.91, and 3.06 m/s for SC50M VV–VH, SC50M HH–HV, and SCLN HH–HV, respectively. The proposed models are being integrated in various Environment and Climate Change Canada applications including ice and wind data assimilation systems and the National SAR Winds program.

The approach of applying different amounts of horizontal localization to different ranges of background-error covariance horizontal scales as proposed by Buehner and Shlyaeva was recently implemented in the four-dimensional ensemble-variational (4DEnVar) data assimilation scheme of the global deterministic prediction system (GDPS) at Environment and Climate Change Canada operations. To maximize the benefits from this approach to reduce the sampling noise in the ensemble-derived background-error covariances, it was necessary to adopt a new weighting between the climatological and flow-dependent covariances that increases significantly the role of the latter. Thus, in December 2021 the GDPS became the first operational global deterministic medium-range weather forecasting system to rely completely on flow-dependent covariances in the troposphere and the lower stratosphere. The experiments that led to the adoption of these two related changes and their impacts on the forecasts up to 7 days for various regions of the globe during the boreal summer of 2019 and winter of 2020 are presented here. It is also illustrated that relying more on ensemble-derived covariances amplifies the positive impacts on the GDPS when the background ensemble generation strategy is improved.

The all-sky assimilation of radiances from microwave instruments is developed in the 4D-EnVar analysis system at Environment and Climate Change Canada (ECCC). Assimilation of cloud-affected radiances from Advanced Microwave Sounding Unit A (AMSUA) temperature sounding channels 4 and 5 for non-precipitating scenes over the ocean surface is the focus of this study. Cloud-affected radiances are discarded in the ECCC operational data assimilation system due to the limitations of forecast model physics, radiative transfer models, and the strong non-linearity of the observation operator. In addition to using symmetric estimate of innovation standard deviation for quality control, a state-dependent observation error inflation is employed at the analysis stage. The background state clouds are scaled by a factor of 0.5 to compensate for a systematic overestimation by the forecast model, before being used in the observation operator. The changes in the fit of the background state to observations show mixed results. The number of AMSUA channels 4 and 5 assimilated observations in the all-sky experiment is 5-12% higher than in the operational system. The all-sky approach improves temperature analysis when verified against ECMWF operational analysis in the areas where the extra cloud-affected observations were assimilated. Statistically significant reductions in error standard deviation by 1-4% for the analysis and forecasts of temperature, specific humidity, and horizontal wind speed up to maximum 4 days were achieved in the all-sky experiment in the lower troposphere. These improvements result mainly from the use of cloud information for computing the observation-minus-background departures. The operational implementation of all-sky assimilation is planned for Fall 2021.

A new technique for automated retrieval of ice concentration from RADARSAT-2 dual-polarization HH-HV ScanSAR Wide images for subsequent assimilation in ice numerical models is presented. First, we extended our previously introduced ice and water detection approach operating at a 2.05 km x 2.05 km spatial scale to a set of 19 different spatial scales ranging from 2.05 km (41 pixels) down to 0.25 km (5 pixels). As the spatial resolution was increased, the overall accuracy of ice and water detection stayed at a very high level across all scales (between 99.5% and 99.8%), but the number of water retrievals substantially dropped. Second, we designed an approach for estimating ice concentration in a 2 km × 2 km (40 × 40 pixels) area consisting of 64 5 × 5 pixel blocks. The 5 × 5 pixel blocks which are initially classified as unknowns are iteratively combined in clusters with effective spatial scales larger than 5 pixels. The clusters are further classified as ice or water using the ice probability model corresponding to the effective spatial scale. The 40 x 40 pixel area becomes populated with high-resolution (5 x 5 pixels) ice and water retrievals, and the ice concentration is estimated based on the number of ice and water retrievals. The proposed approach produces a much better agreement with the Canadian Ice Service Image Analysis ice concentrations (rootmean-square error (RMSE) = 2.2%) compared to the original 2-km ice/water detection approach (RMSE = 19.9%). The developed technique will be adapted to the RADARSAT Constellation Mission data for data assimilation in Environment and Climate Change Canada Regional Ice-Ocean Prediction System.

In this study, our technique for automated extraction of sea ice concentration from RADARSAT-2 data was adapted to the recently launched RADARSAT Constellation mission (RCM). Ice concentrations were derived from more than 2,600 RCM images acquired between August 1, 2020 and March 31, 2021. The retrieved ice concentrations were compared against the Canadian Ice Service (CIS) Image Analysis products. A very good agreement between RCM-derived and CIS Image Analysis ice concentrations was found with the root-mean square error of 2.2% and R-2 of 0.997. Our results suggest that automatically derived ice concentrations from RCM are well suited for further assimilation in sea ice numerical prediction systems such as Environment and Climate Change Canada Regional Ice-Ocean Prediction System.