Sea ice volume variability and water temperature in the Greenland Sea

This study explores a link between the long-term variations in the integral sea ice volume (SIV) in the Greenland Sea and oceanic processes. Using Pan-Arctic Ice Ocean Modelling and Assimilation System (PIOMAS, 1979-2016), we show that the negative tendencies in SIVgo ::: loss :: of :::: sea ::: ice :::::: volume ::::: (SIV) ::: in ::: the ::::: region ::::: goes in parallel with the increasing ice ::: sea :: ice ::::::: volume : flux through the Fram Strait. The overall SIV loss in the Greenland Sea comprises 113 km per decade, while the total SIV import through the Fram strait is increasing ::::: Strait :::::::: increases by 115 km per decade. An analysis of the ocean 5 temperature and the mixed layer depth (MLD) in the marginal sea ice zone (MIZ), based on ARMOR data-set (1993-2016), revealed doubling of the amount of the upper ocean heat content available for the :: sea : ice melt in the MIZ. This increase :: in ::: the ::::: upper ::::: ocean :::: heat ::::::: content : over the 24-year period can solely explain the SIV loss in the Greenland Sea, even when accounting for the increasing SIV flux from the Arctic. The increase in the ocean heat content is found to be linked to an increase in the temperature of the Atlantic water ::::: Water : in the Nordic seas :::: Seas, following an increase of ocean heat flux form 10 the subtropical North Atlantic. We argue that the predominantly positive North Atlantic Oscillation (NAO) index during the four recent decades, together with the intensification of the deep convection in the Greenland Sea, are responsible for the overall intensification of the circulation in the Nordic seas :::: Seas, which explains the observed long-term variations of the SIV. Copyright statement. TEXT

The ocean clearly plays an important role in the sea ice :::::::: formation ::: and : melt in the region. In particular, it is speculated that 15 the oceanic convection in the region favours a more intensive warm water flux from the south, affecting the air temperature and the sea ice extent (Visbeck et al., 1995). However, presently there is a lack of investigation linking oceanic processes with the sea ice variability in the Greenland Sea (Comiso et al., 2001;Kern et al., 2010).

10
The gate and the methodology used here was :::: were : adopted from Ricker et al. (2018), while in :: the : other two studies somewhat different methodologies and gates locations ( Fig. 1a) were used. Each of the studies also is : is :::: also : based on different data-set ::::::: data-sets of sea ice concentration (SIC), thickness (SIT) and drift (SID) ( Table 1).

Assessment of PIOMAS-derived ice volume flux through the Fram Strait and sea ice volume in the Greenland Sea
In order to assess the quality of the PIOMAS data in the region, PIOMAS monthly effective sea ice thickness in the Greenland
Interannual variations of water characteristics, averaged over the upper 200-m and over : in : the MIZ area, are shown in Figure   5 6. From 1993 , an overall year mean increase of :: an :::::: overall ::::::: increase :: of :::::: annual ::::: mean : temperature in the MIZ is observed, suggesting an increasing intensity of the sea ice melt. The temperature increases during all seasons, but the strongest increase is detected in autumn (by 0.5 and 0.6 • C over the 24 years). The winter convection efficiently uplifts heat to the sea surface. The heat is released to the atmosphere and goes tothe sea ice melt, decreasing the interannual trends to insignificant (see Table 3).
Several studies show that during the positive NAO phase , the intensity of oceanic heat flux to the Nordic seas :::: Seas increases by 50%, and the NwASC intensifies along the Scandinavian coast (Skagseth et al., 2004;Raj et al., 2018). On the other hand, the positive NAO phase drives a higher ice drift through the Fram Strait, proved to be the main driver for interannual variations 25 of SIF to the Greenland Sea (Ricker et al., 2018). It is also noted that the positive NAO phase increases of the intensity of the EGC (Blindheim et al., 2000;Kwok, 2000). Finally, the link between the Atlantic water ::: AW transport by the WSC and the cyclonic circulation in the Greenland Sea, related to NAO phase, is obtained from observations and numerical models (Walczowski, 2010;Chatterjee et al., 2018).
From the beginning of 1970s the winter NAO index is growing. From 1979 to 2016 it is mostly positive (Fig. 7), although 15 an overall winter trend can be separated into an increase from 1979 to 1994, a rapid drop from 1995 to 1996 and an increase from 1996 to 2016. The NAO index drop : in : 1995-1996 is observed as :::::::: coincides :::: with a drop in SIV loss and ::::::: regional ::: sea ::: ice :::::: volume :::: loss ::: and :: a decrease in the WSC water temperature ( Fig.4 (b,e)), and :::: b,d). :::: This : can be related to the minimum heat flux through the Svinoy ::: Svin : ø : y : section in 1994 ( Fig. 4 (f) : ,d). The time needed for water properties to propagate from Svinoy :::: Svin : ø : y to the Fram Strait with the NwAC is of order of 1.5-2 years (Walczowski, 2010). 20 Summer NAO index does not govern the interannual variations of the atmospheric system, as well as in the oceanic ones (circulation in the Nordic seas :::: Seas intensifies in winter and is thought to bring more Atlantic Water ::: AW to the recirculation region compared to than ::: that in summer). Consistent with other studies of seasonal interannual variations of current intensity in the region, our results suggest that these are winter variations of the Atlantic water ::: AW transport that bring up the interannual variations of the subsurface water temperature in the MIZ of the Greenland Sea. The decreasing summer NAO index from 25 1979, may be responsible for a somewhat stronger tendency in the SIV decrease ::: loss : in winter, compared to summer (Fig. the additional freshwater input from the ice melt. Oppositely, during freshening of the upper Greenland Sea, the Great salinity anomaly 1966-1972, more ice is :::: was observed in the MIZ region -::: the Odden ice tongue was pronounced (Rogers and Hung, 2008). This confirms the reverse relation between the sea ice content ::::: extent and the MIZ salinity in the Greenland Sea and their dependence on interannual variations of the intensity of the Atlantic Water ::: AW : advection.

25
Our analysis of the upper ocean water properties in the marginal sea ice ::::: (MIZ) : zone of the EGC, shows a notable increase of the Atlantic Water ::::: (AW) temperature below the pycnocline, as well as of winter mixed layer depth from 1993 to 2016. These changes result in a higher sea-surface heat release, providing twice the value of additional heat needed for bringing up the observed SIV loss. Therefore ::: This :::::::: suggests ::: that, the long-term variations of the heat flux entering the Nordic Seas, advected northwards with the NwAC as the Atlantic Water ::: AW and, further on, with the WSC into the MIZ , are found to govern :::::: largely 30 :::::::: contribute the corresponding long-term SIV variations in the Greenland Sea. The analysis of marginal sea ice zone (MIZ) ocean parameters showed an increase in mixed layer depth (MID ::::: MLD) and its temperature from 1993 to 2016. The estimated amount of additional oceanic heat released from 1993 to 2016 surplus the amount of ::: heat : necessary for bringing up the observed SIV loss. Therefore, we state that the Atlantic Water ::: AW advection into the MIZ largely contributes to the SIV loss.
The long-term variations of the Atlantic water transport all the way through the Froe-Shetland ridge, with the WSC and to the MIZ zone. Interannual variations between the parameters, though, do not have high correlations, governed by variations in the local forcing. 5 We also showed ::: We :::::: suggest : that the simultaneous tendencies in the long-term increase of SIF and of the Atlantic water ::: AW transport are both linked to a higher intensity of atmospheric circulation during the positive NAO phase, and, possibly, to the intensity of deep convection :::::: positive ::::: AMO :::::: phase, :::: often :::::: linked :: to ::: the :::::::::::: intensification :: of ::: the ::::::: AMOC :::: since ::: the :::::: 1980s. Not being independent, both mechanisms finally lead to a decrease of SIV in the western Greenland Sea.