The engineering of a physically realistic profile for western Baffin BayWe began by optimally interpolating just the data available to us. This "beta" PHC 2.1 field exhibited the problem discussed above: all of Baffin Bay was characterized by the West Greenland current. We then chose a western Baffin Bay profile from this beta field. One characteristic of this profile was a fairly shallow mixed layer of about 75m, a West Greenland Current trait. Realistically, western Baffin Bay probably has a deeper mixed layer in winter, so for our new profile, we extended the 75m value down to 200m. Next, we obtained a summer profile from PHC 2.0 at the same location. (Summer PHC 2.0 is well represented since there are plently of summer data, so we trust these data.) We don't expect the difference between winter and summer data to be too great at depth, so to finish off our profile, we used the summer profile data beginning at 250m (which is the next depth level available in PHC 2.1 after 200m). Using a 200m mixed layer for our new profile gave the smoothest transition to the summer data profile beginning at 250m, giving us confidence that our strategy was reasonable.