As part of a set of papers we are working on, Emily Gill of the University of Colorado has analyzed the NCEP/NCAR lower tropospheric temperature trends from latitude 60N and 70N to the North Pole for June,July and August. This is shown below for two time periods; the top figure from the time period when satellite coverage becames global and the bottom figure since the large ENSO event in 1998.
These plots are provided as part of the examination of the reasons for the greater sea ice melt in recent years, which I discussed in the post
These two figures address the issue raised in that post to perform
“…analyses of lower tropospheric and surface temperature anomalies by season for the Arctic sea ice regions.”
It is clear there has been warming over the period of record. However, it is relatively small. Using a linear regression, the June, July and August warming since 1979 was +1.0 C, and since 1998 +0.5 to +0.6 C in the region from 60N and from 70N North Pole. There is quite bit of interannual variability such that a linear trend does not explain a majority of the variations over this time period.
Emily Gill has also provided the global June, July and August analyses. The global linear regression change for 1979 to 2012 is +0.73C. For the period 1998 to 2012 and for 1999 to 2012 the linear regression change is +0.43 C and +0.57 C, respectively (the different start years were to include the 1998 large positive value associated with the large ENSO event). Interestingly, there is not much of an Arctic amplification of warming.
It is not clear how this modest lower tropospheric warming would have resulted in such large Arctic sea ice melting unless
i) the warmth was accompanied by less cloudiness than average,
ii) the sea ice was always very marginally close to melting.