Effect of Assimilation on Oceanographic Profile Prediction in the Frontal Zone North of Svalbard

Underwater acoustic modeling relies on realistic description of vertical sound speed profiles, usually derived from temperature (T) and salinity (S). Since in-situ measurements in the ocean are sparse, especially in the Arctic, numerical ocean models are a commonly used source for (T,S) profiles due to their high spatiotemporal resolution and large geographical coverage. Ocean models still have challenges in reproducing observed T and S fields to sufficient accuracy, especially in dynamical areas with fronts and eddies. Data assimilation of measured vertical (T,S) profiles is one solution to reduce the error in ocean models. In this study we present results from a measurement campaign north of Svalbard in September 2022. Vertical (T,S) profiles were collected and the majority of the profiles (78%) were used for real time assimilation in an operational ocean model covering the Norwegian Arctic waters. The remaining profiles were used as reference data. A transect along the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$20^{\circ}\ \mathrm{E}$</tex> meridian across the continental slope (from approximately 81 to 82.5 N) was measured twice during the cruise. This gave an excellent opportunity to study the ocean model's ability to preserve the adjustment from the assimilation. The assimilation improved the ocean model outputs, with reduced bias and root-mean-square error. The most prominent improvements were found in the upper 40 meters for salinity and between 30 and 100 meters for temperature. In the frontal zone along <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$20^{\circ}\ \mathrm{E}$</tex> the effect of assimilation clearly improved the water mass distribution and results showed that the ocean model preserved the effect for at least five days.