If you look at Chapter 6 of NAS, you’ll see heavy reliance on Thompson’s tropical ice cores. There’s much to consider here and this is a very first look. Here’s an interesting statement:
"A quantitative assessment of temperature change from north Tibetan cores [Dunde and Guliya], using typical [whatver that is in tropical ice cores] isotopic sensitivity, is preliminary, but both suggest warming over the last 150 years of at least 1 deg C."
Since Dunde was only drilled in 1987, it is no wonder that the quantitative assessment is only "preliminary". We at climateaudit have also done a "preliminary assessment" using temperature data rather than "typical isotopic sensitivty" and got different results.
Look at Figure 6-1 showing the plots of 6 Thompson cores (Kilimanjaro is not illustrated – anyone want to bet that it doesn’t have an elevated modern dO18 graph? Didn’t think so. I’ve plotted Kilimanjaro last fall,) In the Andes, 2 of the 3 series have higher MWp than modern, but one series, Huascaran, has very elevated values with the rise occurring in the 18th century and continuing to the present. In the Tibetan cores, the only series with a marked modern increase is Dasuopu, but again, the increase seems to occur throughout the millennium. In the original publication, Dasuopu was said to be a precipitation proxy, but is nonetheless averaged with the other series (dare one say because it has HS), yielding an HS Himalyan composite in Figure 6-2 and tropical composite in Figure 6-3. (Also, by the way, the digital versions of the series in Figure 6-1 are archived as a result of my efforts at Climatic Change. Prior to that, nothing had been archived from Dunde in over 17 years and the file is still limited to decadal dO18).
But that’s not the topic here. The NAS panel says that they used the "four available ice cores from Tibet". OK, but there’s an ice core from Mt Everest; yeah, it’s in Nepal but maybe it would be interesting to compare it to Dasuopu, the most southerly Tibetan site. I think that the following plot is interesting on a number of levels. The bottom panel reproduces the archived data from 1000-1997, which is archived only in 10-year steps, showing that the series has HS.
The top panel zooms in on the 1800-2000 period (with a bigger vertical range) and shows the Dasuopu series in black (with a more compressed look due to the change in scale.) The red series is the archived Everest series from Mayewski, only going back to the mid-19th century, but bringing much information for that period.
First, one notices just how much dO18 information is removed by only reporting decadal information. Mayewski has archived annual (and even sub-annual data). Why should Thompson be permitted to archive only 10-year averages. Second, the scale of variation in Thompson’s decadal series is obviously much smaller than the annual ranges, which are very noisy. Third, the series have different appearances on any scale. I haven’t done a correlation, but I’ll bet that the two series, apples and apples, have negligible correlation. The Everest data seems to have maxed in the 1930s (at about the same time as Greenland temperatures, or for that matter, an earlier U.S. temperature peak – to this day, the record USHCN annual temperature is 1934, not 1998).
What does it mean? It sure doesn’t seem very consistent.