Intercomparison of Salinity Products in the Beaufort Gyre and Arctic Ocean

Abstract

Salinity is the primary determinant of the Arctic Ocean’s density structure. Freshwater accumulation and distribution in the Arctic Ocean have varied significantly in recent decades and certainly in the Beaufort Gyre (BG). In this study, we analyze salinity variations in the BG region between 2012 and 2017. We use in situ salinity observations from the Seasonal Ice Zone Reconnaissance Surveys (SIZRS), CTD casts from the Beaufort Gyre Exploration Project (BGP), and the EN4 data to validate and compare with satellite observations from Soil Moisture Active Passive (SMAP), Soil Moisture and Ocean Salinity (SMOS), and Aquarius Optimally Interpolated Sea Surface Salinity (OISSS), and Arctic Ocean models: ECCO, MIZMAS, HYCOM, ORAS5, and GLORYS12. Overall, satellite observations are restricted to ice-free regions in the BG area, and models tend to overestimate sea surface salinity (SSS). Freshwater Content (FWC), an important component of the BG, is computed for EN4 and most models. ORAS5 provides the strongest positive SSS correlation coefficient (0.612) and lowest bias to in situ observations compared to the other products. ORAS5 subsurface salinity and FWC compare well with the EN4 data. Discrepancies between models and SIZRS data are highest in GLORYS12 and ECCO. These comparisons identify dissimilarities between salinity products and extend challenges to observations applicable to other areas of the Arctic Ocean.

The Role of the Russian Shelf in Seasonal and Interannual Variability of Arctic Sea Surface Salinity and Freshwater Content

“…Large, anomalous input from these freshwater sources can impose physical components that lead to differing salinity values between products (Fournier & Lee, 2021). Hall et al (2022) also found similar discrepancies between satellites and ocean models, including reanalysis products. However, their study highlighted the correlation of ORAS5 to observations both at the surface and at depth using in situ measurements.…”

“…Hall et al. (2022) also found similar discrepancies between satellites and ocean models, including reanalysis products. However, their study highlighted the correlation of ORAS5 to observations both at the surface and at depth using in situ measurements.…”

“…Compared to other ocean models commonly utilized to estimate Arctic Ocean salinity, Hall et al. (2022) found ORAS5 to have the largest positive correlation and smallest bias to satellite and in situ observations in the BG area, both at the surface and at depth.…”

Sea Ice Meltwater in the Beaufort Gyre: A Comprehensive Analysis Using Sea Surface Salinity Data from SMOS

“…3a), so a larger number of satellite SSS observations can be retrieved. Therefore, for all Septembers from 2011 to 2019, we first compare the SSS retrieved from SMOS against that solved by two model reanalysis: TOPAZ4b, which assimilates SMOS SSS data (Xie et al, 2023), and ORAS5, which has been reported as the best SSS performer in the BG (Hall et al, 2021). Both ORAS5 and TOPAZ4b models exhibit significant deviations from observed SMOS SSS, failing to capture the freshest waters detected by the satellite (monthly mean SSS of 19.3 psu) (Fig.…”

“…Despite lower sensitivity in polar oceans due to low sea surface temperatures (SST), recent advances in retrieval algorithms have led to tailored Arctic products (Martínez et al, 2022), facilitating the integration of SSS data into Arctic studies (e.g. Fournier et al (2019); Hall et al (2021); Umbert et al (2021Umbert et al ( , 2023).…”