Another dodgy graph from E-G Beck

It just goes to show what you can get away with if you're willing to distort the axes of a graph. I find it difficult to think such a deception could occur by accident.

This was the very first kind of visual misdirection I learned about, when I was a youngster reading Larry Gonick and Wollcott Smith's Cartoon Guide to Statistics. It's a little freaky that it's so applicable.

Just to be clear: Aren't the games with the x-axis tick spacing the *least* of this graph's problems?! More pertinently, where the heck does this temperature data come from?! It shows clear peaks in 400 BC and 1200 AD that are higher than the 2000 AD datum (which note also is clearly depicted to be a *maximum* as if it's clear already temperatures have peaked and the world is cooling again).

My impression is that the consensus view on recent climate history is the following one that's summarized by a very nice set of webpages maintained by NOAA:

There are not enough records available to reconstruct global or even hemispheric mean temperature prior to about 600 years ago with a high degree of confidence. What records that do exist show is that there was no multi-century periods when global or hemispheric temperatures were the same or warmer than in the 20th century.

[Specific Source: http://www.ncdc.noaa.gov/paleo/globalwarming/medieval.html]

This does look dodgy. Good eye. Perhaps you'd also care to look into the graphic included in the IPCC AR4 and TAR in which Keith Briffa's temperature reconstruction is truncated at 1960, which is coincidentally the point in time at which his reconstruction radically diverges from the other reconstructions and also from the instrumental record.

Kevin, Why do you think he truncated it at 1960? More precisely he truncated the calibration of the proxy at 1960, so it would not include later times when he knew that the proxy was not behaving in an appropriate way. He did know that it was appropriate up till then. Geez.

Eli - how did he know that the proxy behaved appropriately for hundreds of year up to 1960, but has become "inappropriate" since then?

Because he is a dendrologist, e.g. he studies how trees grow and what effects them and their rings.

Which is why the whole woohaw about MBH 98/99 is stupid. If there are questions about the selection of tree ring records, they should have been addressed to Malcolm Hughes and not Michael Mann.

Eli,

1) A dendrologist or dendroclimatologist is not much like a plant physiologist. I have read noted dendroclimatologists specifically referencing their own lack of understanding of plant physiology, so perhaps your appeal to his authority here is misplaced?

2) Please correct me if I have it all wrong, but I understood the point of the proxy measurements was to infer prior temperature from tree ring density. Now if evidence presents itself that the proxy fails in that regard in the present, 1) that is noteworthy and should not be hidden as it was in the IPCC TAR and AR4 and 2) your reply begs the question of how exactly [please no vague appeals to his professional authority] Briffa would know that it was a good proxy at all. Sans a time machine, how could one know whatever made the ring history "suddenly inappropriate" didn't also transpire many times in the intervening millenia?

After all, the inference that historical temperature is reflected in tree rings would appear to be confounded by the fact that they failed to correlate past 1960, wouldn't it? If the proxy one claims is accurate to reflect temperature 1000 years ago fails to accurately reflect it now, how does this not present an issue?

Let us see. Biffra knowns that certain species significantly changed behavior ~1960-70. He knows that before that their growth followed temperature in a particular way. He wants to calibrate the growth vs. temperature with an instrumental record. Does he use the post 1960 data?

Eli:

"Let us see. Biffra knowns that certain species significantly changed behavior ~1960-70."

As an aside, divergence problems have arisen for a number of species and studies in the relatively recent past haven't they? The issue is not restricted to Briffa.

"He knows that before that their growth followed temperature in a particular way."

How does he know this? The evidence that a given species was an accurate proxy would have to come from it being an accurate proxy now allowing for the induction to its past performance, right? His actual evidence is that it is not an accurate proxy now. How do you get from the evidence that at least some changes can render a proxy wildly inaccurate to that proxy having always been accurate and unchanging even when it was not being directly observed and is itself the evidence for conditions at the time? Paging Kierkegaard, I think we'll be needing a leap of faith to patch this one up.

"Does he use the post 1960 data?"

Should he use the pre-1960 data is the more important question. Post-1960 data he definitely should not use if he wants to make the case that his tree species are good temp. proxies. OTOH, if he wants to objectively evaluate the worth of those species as temp. proxies, he should absolutely use post-1960 data as counter-evidence to his hypothesis.