Transport changes
The Arctic freshwater content and transports through the key straits remain different among different Arctic ocean models or Arctic ocean reanalyses (
Aksenov et al., 2016;
Uotila et al., 2019). In contrast to
Koldunov et al. (2017), our results show that the assimilation changes the ocean circulation in the pan-Arctic Ocean (Figure 8a). We also compare examined the transport changes in detail. Table 3 summarizes mean fluxes of volume, liquid freshwater, sea ice volume, and heat through the Fram Strait, Davis Strait, and the Barents Sea Opening in INTAROS-ctrl and INTAROS-opt. As can be seen, changes in the fluxes are much more significant than those shown by
Koldunov et al. (2017). Sea ice volume transports are reduced by ~26.4% through the Barents Sea Opening while increased by 30.9% through the Fram Strait. Net southward volume flux through the Fram Strait is reduced by 16.8% due to an enhancement of the Norwegian North Atlantic Current and a weakening of the Arctic outflow, resulting in increased heat fluxes to the Arctic Ocean and decreased freshwater flux from the Arctic Ocean. The southward net volume flux through the Davis Strait is increased, accompanied by more southward freshwater transport. Data assimilation seems not always to bring the oceanic transports closer to observations, as the reduced liquid freshwater flux through the Fram Strait shows. Unfortunately, direct observations of the transports through the key straits cover different periods, and particular methods are used to fill the spatial gaps, resulting in significant uncertainties in the observational estimates. Moreover, model studies (
Köhl & Serra, 2014) also show substantial decadal variability of the transports. Because of this, direct comparisons are complicated, and conclusions drawn could be ambiguous.
To further examine transport changes after data assimilation and compare with the TOPAZ4 reanalysis, we show the volume, liquid freshwater, and sea ice volume transports from INTAROS-ctrl, INTAROS-opt, and TOPAZ4 (Figure 11). Mean freshwater transports through the Fram Strait and the Davis Strait are substantially changed, but the variability changes are small. A better match between INTAROS-opt and the TOPAZ4 reanalysis can be observed for the sea ice volume transport (Figure 11c,f) than for the liquid freshwater (Figure 11b,e) and volume transports (Figure 11a,d), especially after the year 2012. Regardless of the biases, the variability of volume and liquid freshwater transport through the Davis Strait in TOPAZ4 match well with the optimization (Figure 11d,e). However, no clear correlations are observed in volume and liquid freshwater transports through the Fram Strait (Figure 11a,b).
Overall, the optimization changes the mean freshwater transports through the Fram Strait and the Davis Strait. However, the mean volume and liquid freshwater fluxes through the key straits remain different between our reanalysis and the TOPAZ4 reanalysis.