Satellite sea surface salinity in response to freshwater processes in the Hudson Bay

Abstract

Hudson Bay (HB) is the largest inland sea in Northern Hemisphere. HB is covered by sea ice and snow in winter while complete open water in summer. It is known that two processes dominate HB freshwater cycle: runoff from surrounding rivers, and low salinity water released (trapped) when ice melts (freezes). This study examines the inter annual anomaly of satellite sea surface salinity (SSS) in the context of HB freshwater cycle. We analyze 4 years (5/2015-4/2019) of SSS data from SMAP (JPL V4.2) and SMOS (BEC V3.0) in conjunction of relevant data, including daily discharge rate from rivers surrounding HB from Canada, daily sea ice concentration from National Snow and Ice Data Center (NSIDC), precipitation from NOAA (CMAP), and evaporation from OAflux.

Dividing the Hudson Bay system into sub-regions, we examine local freshwater contents in each sub-domain. We found river discharge and seasonal sea ice changes indeed dominate SSS inter-annual variability, but at different time and locations; while the surface forcing (P-E) plays a minor role. Particularly, in area along the path of river discharge such as the James and eastern Hudson Bay, the abnormally low SSS patches shown in August 2015 could be linked with anomalous positive river discharge for the season, with 2-3 months delay. On the other hand, in area of interior Hudson Bay, the SSS response to sea ice changes seems immediate and strong. For example, when sea ice melts abnormally early in June 2017, SSS shows extremely fresh signature in the area more than one month earlier than other years. We estimate yields, i.e. the contribution from runoff, sea ice and P-E respectively in terms of the HB freshwater contents. We also report SSS variability at two HB gateways: one connects HB with Foxe Basin, where Arctic Ocean water transports into HB; other at entrance to the Hudson Strait, which is the corridor between HB and the North Atlantic Ocean.

This study demonstrates the feasibility of using SSS to gain quantitative understanding of freshwater components and contributions. Moreover, it serves as an alternative validation approach for satellite SSS retrieval in the high latitudes regions, such as HB, where very limited in situ salinity observations during the SMAP period are available for direct comparison. We will also discuss issues regarding future retrieval algorithm improvement