By Anthony Watts, Watts Up With That

There’s a lot of buzz about regarding the Kaufman et al paper published today in Science which claims a recent reversal on a long term Arctic cooling trend and “found that the cooling trend reversed in the mid-1990s.” In the NOAA internal newsletter I cited yesterday, NOAA claims that the “According to the most recent Arctic Report Card, the Arctic Ocean continues to warm”. OK fair enough, we’ll have a look.

NOAA based this on Hadley’s CRU dataset, which of course Hadley refuses to show any raw data for or methodology despite repeated FOI requests, making verification impossible. (Read more here)

Arctic-wide annual averaged surface air temperature anomalies (60-90N) based on land stations north of 60N relative to the 1961-90 mean. From the CRUTEM 3v dataset, (available online here). Note this curve does not include ship observations.

Note the trend from 1980 to present. Note also that there are few weather stations above 60N and even fewer on the Arctic Ice itself. The data is relatively sparse and interpolation/gridding/averaging is employed to come up with the coverage all the way to 90N. We’ll get back to this.

From the abstract:

The temperature history of the first millennium C.E. is sparsely documented, especially in the Arctic. We present a synthesis of decadally resolved proxy temperature records from poleward of 60N covering the past 2000 years, which indicates that a pervasive cooling in progress 2000 years ago continued through the Middle Ages and into the Little Ice Age. A 2000-year transient climate simulation with the Community Climate System Model shows the same temperature sensitivity to changes in insolation as does our proxy reconstruction, supporting the inference that this long-term trend was caused by the steady orbitally driven reduction in summer insolation. The cooling trend was reversed during the 20th century, with four of the five warmest decades of our 2000-year-long reconstruction occurring between 1950 and 2000.

Here’s the press release from EurekAlert. The research team reconstructed past temperatures on a decade-by-decade basis during the past 2,000 years using information gleaned from ancient lake sediments, ice cores, tree rings and other samples. As part of the study, the decade-by-decade climate data reconstruction was compared with sophisticated climate model simulations run by NCAR researchers.

My first thought is that the recent claims seem to be similar to what we saw happen with the flawed Steig et al Antarctic paper. We have a few weather stations, sparse coverage, some Mannomatic style math, and some bold claims from the results.

WUWT commenter Dave Middleton pointed out one problem with this comment: The BBC did a better job of reporting the story and included a graph from the paper. This paper supposedly ties all of the 20th century Arctic warming to greenhouse gas emissions. There’s one tiny problem with the paper’s claim.
All of the anomalous warming occurred in one “step shift” before 1950; while most of the increase in atmospheric CO2 has occurred since 1950. And here is that graph Dave annotated:

OK that’s one problem. Here’s another. Both Kaufman et al and NOAA claim recent Arctic warming. In the case of the Kaufman paper, they specifically claim they “found that the cooling trend reversed in the mid-1990s.”

Since we can’t really look at the Hadley CRU data since it is held under lock and key despite the repeated FOI requests so that analysis and verification can be performed, we can’t really analyze it pertaining to NOAA’s claim of warming. Since NOAA and HadleyCRU use many of the same stations above 60N (they’d have to since there are so few) it seems reasonable to assume they share similar data in the Kaufman et al paper.

Fortunately there is another Arctic temperature data source available we can look at to compare against. And that is from the Danish Meteorological Institute (DMI). Like NOAA, they offer a dataset that shows temperature in the high latitudes. Here is what they say about that dataset and how it is obtained.

The daily mean temperature of the Arctic area north of the 80th northern parallel is estimated from the average of the 00z and 12z analysis for all model grid points inside that area. The ERA40 reanalysis data set from ECMWF, has been applied to calculate daily mean temperatures for the period from 1958 to 2002, from 2002 to 2006 data from the global NWP model T511 is used and from 2006 to present the T799 model data are used.

The ERA40 reanalysis data, has been applied to calculation of daily climate values that are plotted along with the daily analysis values in all plots. The data used to determine climate values is the full ERA40 data set, from 1958 to 2002.

Here is the most recent DMI graph of Arctic temperature:

Note that the blue line represents the “melting point” of ice in Kelvin or 0C/32F The green line represents the average climate from 1958 to 2002, i.e. the “baseline”

I don’t have time to get into a detailed analysis of the raw DMI data this morning as I have other duties, but I do have time to do a visual check that is just as telling.

Kaufman et al claims they “found that the cooling trend reversed in the mid-1990s.” That should easy to spot in the DMI graphs if it exists. So I animated the entire set of DMI graphs from 1958 to 2009. See if you can spot the temperature spikes or the “cooling trend reversed in the mid-1990s.” signature.
Click here for full sized animation

Watch the top of the bell curve above the blue line. See any big changes? I don’t. Note that in the animation above, due to a slight change in DMI’s graphical presentation for 2008 and 2009, I had to graphically fit 2008 and 2009 to match the rest of the animation framework so that there would not be a distracting jump at the end. The data is unchanged in doing this.

One of the most common claims of alarmists is that the Arctic is “melting” and that implies a temperature cause in their statements. But as we see, during the critical melt window, the DMI data seems to hold right along the climatic normal.

One thing about DMI, if you go to their main web page, you don’t find any alarming pronouncements about Arctic melting or temperature reversal like you do at NOAA. Others like NASA say the wind pattern changes is more of an issue, blowing the sea ice southward. Perhaps NOAA and Kaufman should look more closely at before making grand claims.

Read full post here.

Icecap Note: See also this plot of arctic temperatures by Willie Soon from Polyakov of U of Alaska vs the Total Solar Irradiance and CO2. Note the temperatures peaked here about 1940. I agree with Dr. Singer’s response to the story, “I trust Willie Soon’s data more” than the hockey stick crew who have perfected the ability to cherry pick sources of observation and proxy data. See climate audit for examples how this happened again with Kaufman and the media and Science is letting them get away with it.

NOAA’s GHCN data shows a similar graph for the arctic that Polyakov showed, with recent temperatures no higher than the levels around 1940.

Also finally here is the latest daily JAXA IJIS based plots for July through September 9 for 2007, 2008, 2009 for arctic ice extent. Note as the melt season draws near an end, we are running 20.8% above the record low levels of 2007, continuing the rebound started in 2008.

See how scientists predictions have made out to date here. Bloggers had a much better record.