That plot is based on reanalysis data, and covers just one region. Since I am now regularly integrating current NCEP/NCAR reanalysis data (here), I thought I could do the same for that region, and the comparison would be a check. However, in some ways that high Arctic is not where melting is currently concentrated, so I thought I could also give a better spread of areas. Naturally, that means a gadget, below the fold.
The DMI plot shows an average from 1958-2002. That isn't without cost - it requires melding a number of different reanalyses to cover the range. It could be argued too that it misses recent Arctic warming, and so may show modern data as unusually warm, when by modern standards it isn't.
Anyway, I decided to stick with NCEP/NCAR data from 1994 to present, over which time it seems fairly reliable. And of course I use NCEP for current. There are some discrepancies. DMI aims to show the true surface temp 2m above the ground. I show what is basically the bottom grid level.
Here are the comparative plots for 2015. I've tried to stick to the DMI style here:
|DMI||NCEP/NCAR vai Moyhu|
I have however used Celsius - I can't see the point of Kelvin here, especially with freezing being prominent. The plots are broadly similar. The DMI average has a longer period above zero, but to a smaller extent. This may reflect the slight difference in levels. DMI shows a warmer looking winter - reflecting I think the more ancient reference average.
So below the fold is the gadget. It lets you choose arbitrary rectangles on a 5×15° grid in the region above 60°N. I'll place it just below, with further description below that.