lundi 20 décembre 2010

Comment on Watts up with that

This blog is a comment on, todays turn up from Watts up with that.

Well, if you really want to compare with the Dalton minimum, isn't it really essential to use data from that period? And compare that data set with the current one? Yet Archibald ends with data sets ranging from 1976, 1966,1930,2006 and then tries to connect it to 1776-1820 ish. Without using and raw data from 1776-1820 other than sunspot observations. Am I the only one who thinks this a bit strange?
What's needed is a flux measure for those dates in the Dalton or at worst a proxy from ice core data.Then you can compare.

Go back to the data and try again before making predictions based on what at best could be described as a dodgy data set.

This kind of science really gets me because I think that Archibald may have a valid point but it's rather blunted because the scientific argument isn't cutting enough, if you'll excuse the mixed metaphore.

Look at this data set, it goes back 11,000 years, until 50 years before now.

Then add the observations made from 1955 until 2010 (using raw data )and then , perhaps then you'll get an interesting graph. Perhaps one could comment that proxies aren't as reliable as the 10,7 flux raw data, but it's the best we've got.And that mixing the two data sets is also dodgy. But hey,lets do it any way.

look here
or here

Run the data through PAST
and paste your graph and hey presto, then, perhaps then you've got data you can compare. Then you could look for similarities with the rider 'the past doesn't guarantee the present'

Basically the current argument runs like 'its a Dalton rerun because I say so, because I predicted it'( based on data you can't take seriously), and that's just not good enough.

You also forget that most of the Dalton cooling was volcanic, go check the sulphate core records ( and compare them for the time span you want to consider . It's not enough to wag your finger and say 'Ah but solar minimum causes volcanic eruptions because there is a correlation'. Go prove it.Correlation doesn't mean cause.

I spend my days telling students how important it is to respect the scientific method, but sometimes when I see this kind of argument, it just looks like pseudoscience. Intreresting, though provoking, bloggable ,but ultimatly dismissable because it lacks vigour.

There's no null hypothesis, no paper, no findings, no evidence, no method.

Depressingly ,everything here can and will be dismissed, and really that's a shame because we really need to keep an open mind.

1 commentaire:

Rickj a dit…

new ideas here