Novel Arctic sea ice data assimilation combining ensemble Kalman filter with a Lagrangian sea ice model

Abstract

Abstract. Advanced data assimilation (DA) methods, widely used in geophysical and climate studies to merge observations with numerical models, can improve the state estimates and consequent forecasts. We interface the deterministic Ensemble Kalman filter (DEnKF) to the Lagrangian sea ice model, neXtSIM. The ensemble is generated by perturbing the atmospheric and oceanic forcing throughout the simulations and randomly initialized ice cohesion. Our ensemble-DA system assimilates sea ice concentration (SIC) from the Ocean and Sea Ice Satellite Application Facility (OSI-SAF) and sea ice thickness (SIT) from the merged CryoSat-2 and SMOS datasets (CS2SMOS). Because neXtSIM is computationally solved on a time-dependent evolving mesh, it is a challenging application for ensemble DA. As a solution, we perform the DEnKF analysis on a fixed reference mesh, where model variables are interpolated before the DA and then back to each member's mesh after the DA. We evaluate the impact of assimilating different types of sea-ice observations on the model's forecast skills of the Arctic sea ice by comparing against satellite observations and a free-run ensemble in an Arctic winter period, 2019–2020. Significant improvements in modeled SIT indicate the importance of assimilating weekly CS2SMOS SIT, while the improvement of SIC and ice extent are moderate but benefit from daily ingestion of the OSI-SAF SIC. In contrast, the bivariate improvements between SIC and SIT are unobvious. Our ensemble-DA system based on the stand-alone sea ice model is computationally efficient and demonstrates comparable skills to operational forecasting models that use DA.

Tailoring data assimilation to discontinuous Galerkin models

“…The MITgcm model has been paired with a local error subspace transform Kalman filter (Mu et al ., 2018). Other examples, such as the TOPAZ4 (Sakov et al ., 2012; Xie et al ., 2018), the NEMO‐LIM2 (Massonnet et al ., 2013), and the neXtSIM model (Cheng et al ., 2023; Richter et al ., 2023; Sampson et al ., 2021), all use different flavours of the ensemble Kalman filter (EnKF) (Evensen et al ., 2022). It is planned that DA capabilities will also be added to neXtSIM$$$ {}_{\mathrm{DG}} $$$.…”

Bivariate sea-ice assimilation for global-ocean analysis–reanalysis

“…The main correction comes from the use of CryoSat-2 data; the assimilation of SMOS reduced the error in the thin ice about 11 % and 22 % in March and in November respectively, without degradation in the other variables. Yang et al (2014) and Mu et al (2018b) tested the localized singular evolutive interpolated Kalman filter to integrate thickness data and showed an overall error that is similar to the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS) (Zhang and Rothrock, 2003) when compared to independent in situ measurements.Finally, Cheng et al (2023) recently showed, in the standalone Lagrangian sea-ice model neXtSIM interfaced to a deterministic ensemble Kalman filter (EnKF) scheme in a multivariate manner, that improvements in SIT estimates indicate the importance of assimilating weekly CS2SMOS SIT, while the improvements of sea-ice concentration (SIC) and ice extent are moderate but benefit from daily correction from OSI-SAF SIC. In this study, we extend an operational 3DVar data assimilation (DA) scheme, OceanVar, employed in the production of global and regional ocean reanalysis and forecasts (Storto et al, 2019a;Escudier et al, 2021;Lima et al, 2021;Ciliberti et al, 2022), to treat sea-ice concentration (SIC) and thickness (SIT) data.…”