Following my previous post there has been discussion in the comments on “which graph to believe”. Sadly this becomes ideological, for some. I think the major point is that the HPS ’97 graph (the one here) just isn’t used anymore by anyone, except the septics who want to see a MWP. The graph has never been explicitly disowned, but the authors of it have published plenty more since then, all only going back 500 years (and AFAIK no-one else before or since has tried to use boreholes back that far), and showing a different timing of the cold bit. Naturally, if you’re paranoid, this is because the evil IPCC leaned on them. Frankly, thats black-helicopter stuff of the first order, but its also non-historical because MBH’98 only went back 6 centuries, so didn’t show the MWP one way or another. You have to wait for MBH’99 for the 1000 years, which was *after* the borehole folk switched to 500 years.
But if you don’t like that, then another problem is the temperature it comes out with for the LGM – about -1.5K cooler than “now” (though see next point…). This isn’t nearly cold enough – rather a hint that something is wrong.
Yet another point for those wanting a MWP is that the zero point on the graph doesn’t represent “now” – it represents start-of-20th-century (becasue the top 100m of data were not used). Given what we know from the instrumental record, then even if you believe this graph, the MWP wasn’t warmer than “now” – indeed it was a bit colder.
[Update: a suggestion from someone who has read the paper more attentively than me: It looks to me as though they have done a single inversion using an estimate of the global mean heat flow as a function of depth, as opposed to doing the inversion at each profile and then averaging as in Huang et al. 2000. Their heat flow is based on 37 profiles at 2000m, but 1001 at 100m, so there is a large change in the region being sampled. My guess is that this approach gives meaningless results -W]