Variations in mean sea-ice albedo can be explained using sea-ice concentration, surface air temperature and elapsed time from onset of melt as drivers.[/i]
I would have thought they should get a better fit if they did use FYI/MYI difference.
Likewise. The data series in Riihelä et. al. is through 2009, and the percentage of FYI jumped in 2008. In view of the albedo difference between FYI and MYI relating to melt pond formation (Figure 3 of Perovitch & Poshenski 2012, as Chris Reynolds discussed at Dosbat
http://dosbat.blogspot.co.uk/2012/09/sea-ice-albedo-changes-and-seasonal.html), you would expect some drop-off at the end of the Riihelä graphs. Maybe not in May (the blue top trace in Fig. 2) before the melt ponds amount to much, but at least in June and July. Yet those values for 2008 and 2009 are near the trendline.
It might take a few more years of data, averaging out weather effects, to show the influence of FYI/MYI on the albedo trend.