This website uses cookies, which are small text files that are used to make websites work more effectively. In order to continue using this website, you will need to accept the use of cookies.

The thickness distribution is one of the main state variables for the Arctic sea ice. It is a key information used, for instance, to estimate the total sea ice volume over the Arctic ocean and its respective seasonal and interannual variability. Despite the continuous efforts for compiling former and recent datasets from a range of observational platforms and the advent of new satellites capable of observing this variable, in situ and remote observations of sea ice thickness and related parameters such as draft and freeboard are still very sparse both in time and space.

Due to this lack of observations, the ocean–sea ice reanalyses remain the main sources of Arctic sea ice thickness. A reanalysis product consists of models’ outputs, which are generated over a certain time span by the same model, configurations, assimilated data, and procedures, and so are distributed onto regular grids, evenly stepped in time.

In a recently published manuscript in The Cryosphere, Ponsoni et al. compared 14 different reanalyses in order to evaluate the impact that the assimilation of sea ice concentration data by the reanalyses has on the resulting sea ice thickness values, in terms of reproducibility and variability.

By comparing the reanalyses against different available observations, the results suggest that systems which take advantage of sea ice data assimilation do not necessarily perform better in terms of sea ice thickness compared with reanalyses which do not assimilate sea ice concentration. This results reflect the fact that few reanalysis, and its respective assimilation method, do not reflect the covariances between sea ice concentration and sea ice thickness fields.

However, sea ice data assimilation does has a clear impact on two important aspects of the sea ice thickness variability: the timescales (or persistence) and length scales. The first aspect is related to how long anomalies in the sea ice thickness field persist over time, while the second reveals how these anomalies spread in space. The results show that reanalyses built with sea ice data assimilation present shorter timescales (2.5–5.0 months) and length scales (337.0–735.5 km) compared to the reanalyses built without sea ice data assimilation (4.9–7.8 months and 846.7–935.7 km, respectively) . The main reason why the assimilation of sea ice concentration data impacts the timescales and length scales is linked to the fact that when a reanalysis assimilates sea ice information, the system is forced towards the assimilated conditions, different from what occurs with free-running models. Eventually, data assimilation introduces thickness increments that are not necessarily physical and therefore contributes to an attenuation in the correlation of this variable at a certain grid cell both in time, with their future estimations, and in space, with the neighboring grid points.

The findings of this study support the design of an optimal observing system for the Arctic sea ice thickness, a milestone of APPLICATE’s Work Package 4.