Decoding Climate Change with Perl, gnuplot and Google Earth • Hacker News Board

Re: Decoding Climate Change with Perl, gnuplot and Google Earth

#31

Post by azgolfer » Tue, Dec 29, 2009, 5:58 AM UTC

Tichy wrote: post #26

patrickgzill wrote: post #21

I'm sorry, isn't the elephant in the room the admission that the HadCRUT data was shown to be corrupted/inaccurate? And that a few lines of Perl automated the error checking to such an extent that the errors just popped up? How exactly am I to have confidence in the rest of the numbers?

Have a link about the inaccuracy of HadCRUT data? Or are you referring to the small problem mentioned in the article? As it was fixable, I would assume it is not a problem anymore - it shows the scientific process worked for once.

Here is the most recent sign that that data is 'cooked'.

On Tuesday, the Moscow based Institute of Economic Analysis (IEA) issued a report claiming that the Hadley Center for Climate Change based at the headquarters of the British Meteorological Office in Exeter (Devon, England) had probably tampered with Russian-climate data... Over 40% of Russian territory was not included in global-temperature calculations for unknown reasons, rather than the lack of meteorological stations and observations, according to the IEA.

http://global-warming.accuweather.com/2009/12/was_russian_cl...

Re: Decoding Climate Change with Perl, gnuplot and Google Earth

#32

Post by Tichy » Tue, Dec 29, 2009, 10:03 AM UTC

patrickgzill wrote: post #27

Tichy wrote: post #26

Earlier quoted context omitted.

Have a link about the inaccuracy of HadCRUT data? Or are you referring to the small problem mentioned in the article? As it was fixable, I would assume it is not a problem anymore - it shows the scientific process worked for once.

Characterizing it as a small problem, and "fixable" seems to minimize the very obvious fact that a few lines of Perl caught problems that should have been caught well before the data was released.

The data is public, so anybody could have double checked it and found that error. Somebody now did. This is how it should work - it is the only way it can work. Errors can always happen, but scientists are supposed to double check.

If the error was deliberate, it would have made no sense to publish the data in such a way that the error could be caught easily.

Re: Decoding Climate Change with Perl, gnuplot and Google Earth

#33

Post by Tichy » Tue, Dec 29, 2009, 1:21 PM UTC

azgolfer wrote: post #31

Tichy wrote: post #26

Earlier quoted context omitted.

Have a link about the inaccuracy of HadCRUT data? Or are you referring to the small problem mentioned in the article? As it was fixable, I would assume it is not a problem anymore - it shows the scientific process worked for once.

Here is the most recent sign that that data is 'cooked'. On Tuesday, the Moscow based Institute of Economic Analysis (IEA) issued a report claiming that the Hadley Center for Climate Change based at the headquarters of the British Meteorological Office in Exeter (Devon, England) had probably tampered with Russian-climate data... Over 40% of Russian territory was not included in global-temperature calculations for unk…

Believe what you want to believe. Since you are jumping to conclusions, I don't think discussion would make sense.

The fact remains that the data set has been published openly. If it is cooked, everybody is free to point it out.

I hope the IEA has published the missing data from the Russian territory? Then it should be a simple matter of another Perl script to complete the data. Problem solved. If you have a link to the missing data, please provide it.

However, I don't think such a data set should just be used as a proof for anything. The perl script in the article her was an interesting play, but I don't think the person who wrote it is a climate researcher. He is just some guy visualizing some data that was floating around. The interpretation is another matter.

Re: Decoding Climate Change with Perl, gnuplot and Google Earth

#34

Post by shawndumas » Fri, Jan 01, 2010, 10:21 PM UTC

pixcavator wrote: post #30

shawndumas wrote: post #23

Earlier quoted context omitted.

Can someone run the numbers using a different time frame as the baseline? Can someone who is not an uneducated ignoramus like myself explain why it matters?

The shape of the graph should be exactly the same only shifted up or down. That's the good news. The bad news is that this is only true for a single station. The shape of the curve for the global average could be seriously affected by the fact that hundreds of new stations were being open around the world.

excellent. thank you.

Re: Decoding Climate Change with Perl, gnuplot and Google Earth

#35

Post by DanielBMarkham » Fri, Jan 01, 2010, 10:28 PM UTC

If I see another article thread that I am unable to downvote any of the comments I'm going to start thinking seriously about leaving HN.

Not a threat, simply pointing out that somebody is getting a little too heavy-handed with the social engineering for my comfort.