Print 119 comment(s) - last by Andy35W.. on Nov 27 at 6:30 AM

GISS's October Data. The large reddish-brown area in Russia is actually September readings.
Amateur team finds NASA error similar to one they discovered a year ago.

NASA'S Goddard Institute for Space Studies (GISS) is one of the world's primary sources for climate data. GISS issues regular updates on world temperatures based on their analysis of temperature readings from thousands of monitoring stations over the globe.

GISS’ most recent data release originally reported last October as being extraordinarily warm-- a full 0.78C above normal. This would have made it the warmest October on record; a huge increase over the previous month's data.

Those results set off alarm bells with Steve McIntyre and his gang of Baker Street irregulars at They noted that NASA's data didn't agree at all with the satellite temperature record, which showed October to be very mild, continuing the same trend of slight cooling that has persisted since 1998. So they dug a little deeper.

McIntyre, the same man who found errors last year in GISS's US temperature record, quickly noted that most of the temperature increase was coming from Russia. A chart of world temperatures showed that in October, most of Russia, the largest nation on Earth, was not only registering hot, but literally off the scale. Yet anecdotal reports were suggesting that worldwide, October was actually slightly colder than normal. Could there be another error in GISS's data?

An alert reader on McIntyre's blog revealed that there was a very large problem. Looking at the actual readings from individual stations in Russia showed a curious anomaly. The locations had all been assigned the exact temperatures from a month earlier-- the much warmer month of September. Russia cools very rapidly in the fall months, so recycling the data from the earlier month had led to a massive temperature increase.

A few locations in Ireland were also found to be using September data.

Steve McIntyre informed GISS of the error by email. According to McIntyre, there was no response, but within "about an hour", GISS pulled down the erroneous data, citing a "mishap" and pointing the finger of blame upstream to the National Oceanic and Atmospheric Adminstration (NOAA).

NOAA's Deputy Director of Communications, Scott Smullens, tells DailyTech that NOAA is responsible only for temperature readings in the US, not those in other nations.

The error not only affected October data, but due to the complex algorithm GISS uses to convert actual temperature readings into their output results, altered the previously published values for several other months as well. The values for August 2008, for instance, changed by 0.11C and the global anomaly as far back as 2005 increased by a hundredth of a degree.

GISS is run by Dr. James Hansen, a strident global warming advocate who has accused oil companies of "crimes against humanity".  Hansen recently made headlines when he travelled to London to testify on behalf of a group of environmentalists who had damaged a coal plant in protest against global warming. Hansen also serves as science advisor to Al Gore.

Dr. Hansen could not be reached for comment.

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

RE: What??? Say it ain't so Jim!!!
By Ringold on 11/11/2008 9:18:48 PM , Rating: 2
Not quite what I was getting at, but you're close. Lets say you're working with an equation, and you plug in numbers. You should have a vague idea of what should pop out, right? For example, if you're trying to model household income, and out pops a value of a trillion dollars, or even 200k, you know something has been screwed up. For all of Russia, temperatures spiked huge versus what they should've been for October, a similar red flag.

Difference is, when I do my own work and get silly results, I painstakingly double check every step. GISS, apparently, says "Yay, it got warmer! PUBLISH!" As for economic data in general, do you ever see wildly incorrect data get published in the US only to later suffer massive revision? Not really. You cite the job numbers for October, but it was September that got a sizable revision in the Oct jobs report. The revision was still in the same ball park however, and not even entirely unexpected. Russia showing up as a giant fireball in that picture in the article? A little unexpected, I'd say.

Maybe they're not entirely to blame, but I'm a little surprised at their incompetence in not catching it immediately, unless they're so under staffed that they automatically kick out data and charts with very little human interaction beforehand.

By foolsgambit11 on 11/12/2008 8:01:28 PM , Rating: 2
Actually, I was referring to the fact that when the November numbers come out, they'll also revise October. Because they revise the numbers every time.

But you're absolutely right, they should have been curious when they got such a striking anomaly.

UPDATE: They released the October data again. See

Notice, there's still a big heating area over Russia. I don't know what that's about, but the anomaly number has come down. It's at .65 instead of, what was it, .78? Something like that. It looks like parts of Russia are slightly cooler than the previous map displayed at the top of this article, but most of it is still pretty hot, versus the average for October.

“Then they pop up and say ‘Hello, surprise! Give us your money or we will shut you down!' Screw them. Seriously, screw them. You can quote me on that.” -- Newegg Chief Legal Officer Lee Cheng referencing patent trolls
Related Articles

Copyright 2016 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki