Proxy Reconstructions Revisited

[IanC]

know the difference between precise and accurate? find out first.

many recontructions use multiple proxies, presumably to give as accurate a value as possible working under the assumption that the errors in each proxy will cancel out when you use a bunch of them. you end up with highly averaged values that show very little variance. eg Mann and Marcott. the error bars are mathematically small, probably unreasonably so.

if you want to see the variance you have to go with one proxy at a time. eg. the Greenland ice core.

this shows various stages of warming and cooling but it only for one part of the world. and polar regions 'amplify' the variability even though they are a pretty good stand in for global temps in general just not in absolute terms because it is of course much cooler up there.

the above graph was made using 18O isotope measurements. another way is to use Nitrogen and Argon isotopes.

and here is a graph of both methods plotted together. unfortunately there is no 'average' of the two methods.

there is a reasonable general agreement between the two methods but if you averaged the two then the variance would decrease and the shape would slightly change. remember, this is for the same sample but two different methods. (a modern instrumental temperature series was added for context)

here are the first 25 proxies used in Marcott

can you see how combining these wildly different proxies would absolutely trash any temperature spikes in the past? perhaps the average is accurate, albeit highly smoothed, but to say that you can pick out temperature swings in this dog's breakfast of proxies is ridiculous.

 

[Crick]

I'd say that using multiple proxies from different locations around the world was intended to give a global measure rather than a local or regional one.

Neither Marcott, Mann or Shakun (the one you left out) ever claimed that their records would show short duration spikes and when the question was brought up by others, they provided the chronological resolution of their work.

You never demonstrated your claim that the error bars in the final work were unreasonably small.

One error that continues to be made here is the assumption that the Earth's temperature is going to drop as quickly as it has risen - creating a spike. That is extraordinarily unlikely. That makes this search for a spike fishing for a red herring.

 

[skookerasbil]

My God man.........nobody cares.

A thread like this is proof that for the AGW people, this is a veritable hobby. They just like the dialogue.......but it sways nobody and I mean nobody in the bigger picture. Discussions like this might interest about 213 people in the population. Why do you think that in years, this forum has drawn zero regular posters to the AGW side. There are like 4 regular guys..............still.........

"Short duration spikes".......... wtf ...........is it no wonder why the entire planet has tuned out global warming as a concern?

 

[Crick]

No one cares what you think or what you have to say Skook.

 

[IanC]

I'd say that using multiple proxies from different locations around the world was intended to give a global measure rather than a local or regional one.

Neither Marcott, Mann or Shakun (the one you left out) ever claimed that their records would show short duration spikes and when the question was brought up by others, they provided the chronological resolution of their work.

You never demonstrated your claim that the error bars in the final work were unreasonably small.

One error that continues to be made here is the assumption that the Earth's temperature is going to drop as quickly as it has risen - creating a spike. That is extraordinarily unlikely. That makes this search for a spike fishing for a red herring.

yes, there are multiple proxies from around the world. often with changes of a few degrees, the timing of the changes are different in each one, AND SOME GO IN OPPOSITE DIRECTIONS! the Marcott graph gives error bars of +/- 0.2C for the whole 12,000 year range. I dont find that credible even if mathematical formulae spit those numbers out. crap proxies dont turn pristine just because you use 73 of them.

Marcott made the rounds with his press releases claiming that the last hundred years made up all the temperature drop of the last 5000. it was not until he was pointedly questioned that he admitted that his work was 'not robust' for the last century, and that people should not have jumped to the conclusion that it was.

ice core proxies from different areas around the world have very good resolution because there are yearly bands and volcanic markings to compare. they show large variations that do not show up in the hamburger of ground up proxies like Marcott.

now you are saying that the temperature will not drop in the next 100 or 200 years. do you have some proof of that? has natural variability magically stopped?

 

[Crick]

Marcott never claimed, suggested or insinuated that his work WAS robust for the last century. Who the fuck cares? We have thermometers for that. Do you not trust thermometers?

It will not drop anytime soon because we are not at equilibrium with the CO2 we've already added to the atmosphere. And, of course, we continue to add more. If we could drop CO2 emissions to our 1750 levels this instant, temperatures would not stop rising for a good century. The lifetime of CO2 in the atmosphere is long enough to guarantee it.