We decided to take a fresh new look at the historical hydrographic data from the Environmental Working Group-atlas (EWG) in light of the recent dramatic changes, first and foremost to shed light on the cause of the large change in sea ice extent in the 21st century.
Large changes in the hydrographic properties in the Arctic Ocean after the 1990s are evident, especially in depths and regions corresponding to Arctic Atlantic and Arctic Pacific waters. The changes started in the 1990s and continued in the beginning of the 21st century.
The increase in Atlantic temperature in the inflow region is the most evident hydrographic change that has taken place. The Atlantic temperature in the region north of Spitsbergen is 3.5 +/- 0.8 degrees higher in the 21st century compared to the historical data, whereas the changes are less in the Eurasian and Canadian Basins, with temperatures up to 0.5 +\- 0.4 degrees higher than the EWG climatology. The temperature increase is accompanied by salinification in the Atlantic Layer. The salinity north of Spitsbergen is 0.4 +\- 0.2 units higher in the 21st century compared to the historical data. The salinity in the other regions in the Arctic Ocean is higher with 0.2-0.3 +\- 0.1 units.
The Pacific Layer in the Canadian Basin is 0.1-0.6 +\- 0.425 degrees warmer in the 21stcentury compared to the historical data. The Pacific temperature increase in the Canadian Basin is accompanied by a freshening in the 21st century. The Pacific Water is up to 1 +\- 0.125 unit fresher than in previous decades. The temperature increase and salinity decrease led to decreasing densities in the Pacific Layer.
The observations imply that the sea ice cover in the Canadian Basin is closely linked to events of warm Pacific Water. The waters were especially warm in 2007, when a warm pulse that lasted for almost a year coincided with the sea ice extent minimum in 2007. In 2009 there is so far no such event of warm Pacific Water evident, and the outlook for the sea ice so far is not really hinting at an extreme year.
The interface between the cold Halocline Water and the underlying Atlantic Water has been lifted in the 21st century, with up to 60 meters. The ascent started in the 1990s. Traditionally it has not been argued in favour of a close relationship between the sea ice and the Atlantic Layer because of the thick halocline layer, a cold vertically stable layer lying above the warm Atlantic Layer. However the observations show that this layer is thinning, meaning that the Atlantic influence on the sea ice will be more important in the future.
As an additional part of the thesis I have analysed model results from the NCAR and HADLEY models. The models are two of the global climate models used by the Intergovernmental Panel on Climate Changes (IPCC) in their assessment of the status of understanding the climate changes. Specifically, I have compared the NCAR and HADLEY models to observations, to see how well they simulate the changes seen in recent decades, both in ocean potential temperature and sea ice concentration.
According to the models larger changes are predicted in the 21st century. The NCAR model predicts ice free summers by the end of this century. The loss of sea ice in both models is closely related to ascent and warming of the Atlantic water masses (though in the HADLEY model this process does not kick in until post the year 2000). Neither change occur in model runs without continued atmospheric forcing. I therefore argue that the ascent of the Atlantic Layer, which started in the 1990s is a key climate indicator which is important to continue monitor. Presently the warm Atlantic Water cannot affect the Arctic sea ice significantly because the warm water is covered by a thick layer of cold water (Halocline Water). As the analysis shows, this thick layer is thinning, and is projected to vanish within the 21st century. The Arctic is a vulnerable region, and if this is a development which cannot be reversed we are facing large challenges in our immediate future.