I've had an informative email exchange with Paul Gettings. For some reason he's unable to copy his messages to this mailing list, but he gave me permission to repost his messages here. Below I've pasted together our three-message exchange. (Paul's first message to me was in reply to a message I sent to the list.) Those of you who aren't "sick and tired" of this topic may find Paul's messages interesting (I did). If your blood-pressure elevates on sight of further discussion of this topic, you should probably hit "delete" right now. :-) Chris ######################################################################## Paul's first message: ######################################################################## Date: Wed, 8 Nov 2006 13:02:12 -0700 From: Paul Gettings <Gettings@earth.utah.edu> To: slas@2nerds.com Subject: Re: [Utah-astronomy] Climate chaos? On 08-Nov-06 11:30, slas@2nerds.com wrote:
It's unfortunate that you don't have time to review the article posted at the beginning of this thread. One of the theses of that article is that the supposed recent unprecedented rise in global temperature simply does not exist. The author of the article makes the case that
This is incorrect. The recent rise in temperatures do exist, as shown by the instrumental record. The "hockey stick" has not been used as a good indicator of past climate for at least 3 years, at least in the climate science community. More recent proxy reconsctructions, using techniques that explicitly seek to maintain long-period signals, show a MWP, Little Ice Age, and significant decadal structure for the past 1000 years. The temperature excursions of the MWP and LIA do not exceed the recent (<100 years) warming. The new proxy results compare very well with the results of borehole temperature inversions for a pre-observational mean temperature, or centennial ramp fits for the past 1000 years. Borehole measurements have been shown to track surface air temperatures to much better than 0.1C on annual and longer time scales, using a variety of experiments. Hence, when the borehole results show local or global warming/cooling events in the past, there is extremely strong evidence that such changes occured. The UN IPCC reports do not mention borehole reconstructions at all, and only the most recent NAS reports do so. This despite the simple (but accurate) physics of the reconstructions - no fancy math, just simple physics - and the use of boreholes to reconstruct past average climate for nearly 20 years. At least the new reports are finally taking the borehole results into account. Note that when the newest proxy reconstructions are correctly compared to the borehole temperature-depth data (ignore anything by Mann), they agree extremely well. This does not "prove" the proxies, as borehole reconstructions have excellent temperature precision, but very poor temporal precision, due to the diffusion of temperature changes in the ground. However, it does show that the proxy reconstructions are consistent with the borehole temperature record, which is a completely independant line of evidence. For the record, my research group is actively involved in the use of boreholes for past climate reconstruction. Anyone wanting more references should feel free to email me, and I'll pass them along. -- That which does not kill you, may leave you with a nasty limp. ######################################################################## Chris replied thusly: ######################################################################## Date: Wed, 8 Nov 2006 14:36:15 -0700 (MST) From: slas@2nerds.com To: Paul Gettings <Gettings@earth.utah.edu> Subject: Re: [Utah-astronomy] Climate chaos? Hi Paul, Many thanks for your reply. [I received your message but did not receive a separate copy via the mailing list. Did you send a copy to the mailing list? I think the group would appreciate hearing from someone who knows what they're talking about. In any case, I'm happy to carry on privately.] On Wed, 8 Nov 2006, Paul Gettings wrote:
This is incorrect. The recent rise in temperatures do exist, as shown by the instrumental record.
What resources can I consult to learn more of this particular instrumental record?
The "hockey stick" has not been used as a good indicator of past climate for at least 3 years, at least in the climate science community.
Are you saying that the "hockey stick" graph used in the old IPCC report is no-longer considered to be a reliable reduction of the data? My reading of the original article suggests that there is much room for debate about the methods used to produce that particular "hockey stick" temperature graph. I'd be happy to know how opinions of that particular graph (which was widely used for political effect) have changed over time, if in fact they've changed at all. If opinions of that graph have changed, what led to the change?
More recent proxy reconsctructions, using techniques that explicitly seek to maintain long-period signals, show a MWP, Little Ice Age, and significant decadal structure for the past 1000 years. The temperature excursions of the MWP and LIA do not exceed the recent (<100 years) warming.
What do you make of those who argue otherwise--that the recent warming is well in line with long-term historical data? Are they lackies for the oil industry? Are they interpreting the data incorrectly? This topic fascinates me: how does one square the "consensus" regarding human-caused climate change with the fact that many respected climate scientists have wildly differing opinions? In short, who can we trust and why should we trust them?
The new proxy results compare very well with the results of borehole temperature inversions for a pre-observational mean temperature, or centennial ramp fits for the past 1000 years. Borehole measurements have been shown to track surface air temperatures to much better than 0.1C on annual and longer time scales, using a variety of experiments. Hence, when the borehole results show local or global warming/cooling events in the past, there is extremely strong evidence that such changes occured.
I'm sceptical about proxy studies, probably because I don't understand them well. (I suspect that the coupling mechanisms between the proxies and "reality" are loose and open to wide interpretation.) Anyway, because of my scepticism regarding the proxy stuff, I'm hopeful that more direct methods can be found to reconstruct global temperature histories. It sounds like the borehole measurements might be one avenue leading towards that goal.
For the record, my research group is actively involved in the use of boreholes for past climate reconstruction. Anyone wanting more references should feel free to email me, and I'll pass them along.
Please pass along those references to bolehole-related reconstructions. I'll award bonus points if you also pass along references to materials which contradict the studies obtained by your own research group. :-) By the way, what methods do you use to remove or subdue the temporal uncertainty of borehole measurements? In the absence of any other data, how do you align borehole material with a putative timeline? Aside from the crude "depth" of material in a core, what properties of borehole material provide clues about the passage of time? Thanks, Chris ######################################################################## And Paul replied thusly: ######################################################################## Date: Wed, 8 Nov 2006 17:27:45 -0700 From: Paul Gettings <Gettings@earth.utah.edu> To: slas@2nerds.com Subject: Re: [Utah-astronomy] Climate chaos? On 08-Nov-06 14:36, slas@2nerds.com wrote:
[I received your message but did not receive a separate copy via the mailing list. Did you send a copy to the mailing list? I think the group would appreciate hearing from someone who knows what they're talking about. In any case, I'm happy to carry on privately.]
I can't post to the list, or I would do so. :) Feel free to repost my emails to the list....
What resources can I consult to learn more of this particular instrumental record?
The best instrumental record of surface air temperature is the World Historical Climatology Network (HCN). The measurements have been adjusted to reflect changes in thermometers, heat island effects, etc. to try and get the best surface air temperature measurements over the largest possible geographic distribution, for the longest time. There are also regional/national networks that do the same thing; e.g. the U.S. HCN. To do large scale studies, researchers use weighted averages (or some other interpolation-like scheme) for large areas (typically 1 degree or 5 degree boxes). This removes much of the "micro-climate" effects - local variations in temperature can swamp large-scale, long-term trends in the data. The accuracy of the averages and small samples for regional work, can be best addressed using spatial sampling arguments (recall your time-series analysis from college), which basically boil down to a result that the HCN stations are a sufficient sampling.
The "hockey stick" has not been used as a good indicator of past climate for at least 3 years, at least in the climate science community. Are you saying that the "hockey stick" graph used in the old IPCC report is no-longer considered to be a reliable reduction of the data?
The original hockey stick graph may well have had problems with the analysis technique. Subsequent work has used different techniques to map proxy data to temperatures. These newer techniques take special care to maintain long-period signals, which are mostly lost in the original proxy techniques. Hence, the "hockey stick" is very flat before the present because it loses the decadal and longer frequencies. The new techniques maintain those frequencies, which leads to a result with the MWP, LIA, etc.
What do you make of those who argue otherwise--that the recent warming is well in line with long-term historical data? Are they lackies for
Those who argue that recent warming is in line with historical values are incorrect. The warming is in line with pre-historical (geologic) records. The Earth has not seen current or future levels of CO2 for, likely, millions of years. The range of predictions for the climate models indicates that with increasing CO2 concentrations, climate will warm, which will cause more severe storms, droughts, floods, etc. This will make life more expensive, but not impossible.
topic fascinates me: how does one square the "consensus" regarding human-caused climate change with the fact that many respected climate scientists have wildly differing opinions? In short, who can we trust and why should we trust them?
From the science meetings I have attended, the opinion differences are regarding exact values of forcings, feedbacks, and perturbations, not whether humans can or are affecting climate. I know of many scientists, our group among them, that do not consider climate change to be an existential threat to humans - it will make life expensive, not impossible.
I'm sceptical about proxy studies, probably because I don't understand them well. (I suspect that the coupling mechanisms between the proxies and "reality" are loose and open to wide interpretation.)
Proxies measure temperature to a degree. Most have temperature as a secondary effect, but the primary control on the measure is also linked to temperature. Hence, they are a convenient way to measure temperature over very long timescales, with excellent temporal resolution (annual or decadal). Their main problems come from difficulties with long-period signals and possible decoupling from a temperature signal. The long-period signals can be maintained or restored using appropriate analysis techniques or coupling proxy results with borehole results. Temperature decoupling is unlikely over the timescales being studied, as proxy studies generally use multiple types of proxies, some of which are firmly rooted in basic physics (e.g. isotope fractionation, such as delta-Oxygen-18 studies).
hopeful that more direct methods can be found to reconstruct global temperature histories. It sounds like the borehole measurements might be one avenue leading towards that goal.
Borehole reconstructions can directly reconstruct the average temperature histories over their locations. There is good sampling of the extra-tropical northern hemisphere, which provides a large area for calibration of global climate models.
Please pass along those references to bolehole-related reconstructions. I'll award bonus points if you also pass along
Harris, R.N., and Chapman, D.S. 2005. Borehole temperatures and tree rings: Seasonality and estimates of extratropical Northern Hemispheric warming. Journal of Geophysical Research, vol 110, doi:10.1029/2005JF000303. Also see references therein. The paper, in book form, should be available at the U of U library (ask the science reference desk where they are currently keeping the old journals). If you have net access, and are at the U, you should also be able to get the paper electronically as a PDF. I think I would violate copyright to email you a copy, or I'd just attach the PDF. :)
references to materials which contradict the studies obtained by your own research group. :-)
When we find some work that contradicts ours, we can pass them along. :) Mann tried to argue that boreholes were not in agreement with proxies due to snow cover, but Bartlett et al (2004 and 2005) show that ground and air temperatures track extremely well over annual and longer timescales, and the snow effect is not significant. Arguments over the causes of the current warming have more vigor, as the climate system is so complex we don't have a complete description of it yet. There are forcings and feedbacks that are ignored or poorly known, and we will likely need to be able to do global models at regional resolutions to accurately predict the climate in 2200. However, every improvement of our climate models (in resolution and complexity) has acted to reduce the spread of results, not remove the possibility of significant warming. Note that we use climate models to simulate where the climate might go with a doubling of CO2, or changes in solar irradiance, or ..... Since we can't build a new planet, tweak something, and let it run for a few thousand years, the only way we can do better than random guessing is a computer simulation. Hence, we don't really care if the models can accurately reproduce the temperatures of the planet in the 20th century, we care if the regional or global averages are close, and the weather patterns look similar to Earth's. Then, the extrapolation of the model to future times gives some insight into where the climate might go in our future. Run this for many, many realizations, with different forcings, starting points, etc. and you have a mapping of the possible futures of climate. We might not be able to assign good probabilities to any given outcome, but we can be pretty confident that our real climate will be somewhere in the range; if this range only has bad outcomes, we have a serious problem. :)
By the way, what methods do you use to remove or subdue the temporal uncertainty of borehole measurements? In the absence of any other
OK, I see this work every day, and forget that not everyone on the planet knows exactly how we do these measurements and calculations. :) Borehole temperature reconstructions are done from temperature-depth data; that is, we drop a (very accurate) thermometer (digital, auto-logging, etc.) down a borehole and measure temperatures at different depths. The ground, to a very high degree, is a tape recorder of surface temperatures. Changes at the surface diffuse into the ground at a well-defined rate, and the mathematics of this diffusion is very well understood (analytic solutions, no less!). Hence, if I have a record of temperatures vs. depth in a borehole, and some knowledge of the rock type (which I get from the drilling logs and geology), I can reconstruct the past surface temperatures very accurately, with enormous confidence. However, because the ground is diffusive, changes in the ground tend to smear over time. Hence, I cannot reconstruct a year-by-year record of temperature change, but only an averaged value for some interval. The interval is roughly the same as the time in the past, so I can get the average temperature for a 1000 year window centered at 1000 years ago, provided I have a deep enough borehole with sufficiently accurate temperature measurements. In practice, we use boreholes that are deep enough to get the last 200-500 years, and we use an analysis method that looks for a pre-instrumental mean temperature. For the northern hemisphere, this indicates roughly 1C of warming in the last 200 years. Our more recent work (such as the paper referenced above) has been on reconciling the difference between the proxy results, which often show 0.5-0.6 C of warming, with our results. It appears that the proxies are likely missing about half the warming of the globe, due to the seasonality of the proxy critters/trees, which is captured by the boreholes. -- MUNUC: The most fun you can have without taking hostages. ########################################################################