Corrected NASA GISTEMP data has been posted

After GISS’s embarrasing error with replicating September temperatures in the October analysis, the NASA GISTEMP website was down for awhile today (at least for me).

This evening, the new gridded data was posted, and I generated a world temperature anomaly map with the new data. It clearly has some changes in it from the previous erroneous version.

See below:


GISTEMP 11-12-08 – Click for larger image

You can plot your own here at this link to GISTEMP’s map maker

Now compare the above corrected version with the erroneous one below:

GISTEMP 11-11-08

I’m sorry for the small map, as I was traveling during much of this debacle, and was not able to be online much at all. This one above comes courtesy of Kate at SDA who saved one (thanks Kate).

Note the bottom scale, the top end on the erroneous one was 13.7°C, while the corrected one tops out at 8°C. That alone should have set off alarm bells at GISS. Personally, I don’t believe the 8°C anomaly either, since much of the Russian weather data is suspect to start with, and the data distribution is sparse.

So far, no mention of the new data beyond this yesterday at the NASA GISS news page:

2008-11-11: Most data posted yesterday were replaced by the data posted last month since it looks like some mishap might have occurred when NOAA updated their GHCN data. We will postpone updating this web site until we get confirmation from NOAA that their updating programs worked properly. Because today is a Federal Holiday, some pages are still showing yesterday’s data.

We live in interesting times.


newest oldest most voted
Notify of

I think you got the graphs backwards. If the one that maxes out at 8C is the new one, then your text “Now compare the above erroneous version with the corrected one below” does not match. The above and below should be switched. — John M Reynolds


Wow. That was fast. You fixed it already. You can kill my messages now. Thanks. — John M Reynolds

Mike Bryant

Mike Bryant (13:43:04) :
I’ll wager that when the corrections are made another couple of warm spots will pop up.
Hmmm looks like I was right… Australia and Canada are the most prominent.

hans rast

What happened to Australia?

Andrea Smith

As a non-scientist, I have difficulty visualizing the differences and significances since the maps are not on the same scale. Is it possible to show the new data on the same scale as the older one?

Aaron Wells

Why would the correction cause northern Canada to become much hotter and also Australia? I would have thought that the correction would only serve the purpose of cooling off Russia.

David Walton

Re” I’m sorry for the small map”
I wonder if the GISS is not embarrassed by their error. Probably not. Everyone and anyone can make a mistake but, frankly, such errors seem to be a trend with regard to AGW climate science.


I could explain Australia–it’s south of the equator. But Canada? you got me.


Isn’t it odd that the Arctic ice has refrozen at rates never seen in the satellite age, during October, yet it was warmer in Siberia? Strange indeed.
Doesn’t the mere fact that government scientists are the ones doing the tabulation, makes you just a tiny bit suspicious that the raw data collection and the computer analysis is accurate.


Have they begun to work on the previous years data as well? How would one know which numbers have been corrected/adjusted. Archive procedures and versions would make it auditable. I wonder what the surface stations look like in Russia?

Harold Ambler

The anomalously warm Arctic stands out as a bit of an eyebrow-raiser, given the impressive ice growth during the month. The 4-8-degree-Celsius anomaly over Russia still looks odd, all the more so over the part of the Arctic basin due north of Siberia, which, again, froze like it was someplace cold.
Can UAH, RSS, and Hadley tell us what their values are for Siberia for the month, so that even the new GISS data can be verified? I.E. I don’t have the technical knowledge to know if the satellites can separate a geographic region that way, or whether they do so routinely.

The engineer

The artic in October.
Must be all that ice forming
thats doing the warming !

Mike Bryant

The scales are pretty much the same except for the darkest color. It was greater than 4 and less than or equal to 13.7, now greater than 4 and less than or equal to 8.

James Pfefferle

Anthony – Can you explain the meteorological phenomena that would cause the step change in the temperature anomaly that occurs across the Bering Straight?

Rick Sharp

Amazing how the icecap keeps growing in these high temperatures.

David L. Hagen

At RC #118 gavin Says:
“12 November 2008 at 10:00 PM
The corrected data is up. Met station index = 0.68, Land-ocean index = 0.58, details here. Turns out Siberia was quite warm last month.”
GISS Surface Temperature Analysis

Latest News
2008-11-12: It seems that one of the sources sent September data rather than October data. Corrected GHCN files were created by NOAA. Due to network maintenance, we were only able to download our basic file late today. We redid the analysis – thanks to the many people who noticed and informed us of that problem.”


Interesting. It seems gray indicates no data available since it’s not on their color scale. However, the map based on erroneous data indicates no data for northern Canada and most of Australia (which they apparently didn’t question), yet the one based on corrected data shows available data and also warmer. I wonder if that’s how they’ll maintain their much higher positive anomaly.


Is there any way that the two graphs could be subtracted from each other, so that we may see the differences? It would appear that the only change of scale is for the maximum positive anomaly. All of the other scale marks appear to be identical in the two graphs, yet there are changes in the other values (ranges) also.


I fail to see where the embarrasment is to be found. Siberia is shown to be experiencing a heat wave. 2007, 2005 and 1934 still are the hottest years on record and the variance of 1934 being hotter is around 0.01, so where is the inconsistency, the rebuttal to a multitude of data or the cause for doubt in the first place, based upon real data? The northern latitudes look like they experienced quite a warming, actually from either map.
Did you look over the 2007 graphs, models, and empirical results?
REPLY: Putting identical September data for hundreds of stations in the October data analysis and then touting October is the ‘warmest on record”, based on the flawed replication of data is not an embarrassment?
No need to check 2007 models, graphs and empirical results to see this major flub, except for those that want to change the subject. We are discussing October 2008, not 2007, not 2005, not 1934. – Anthony


Interesting how the artic ice caps are atcually observed to be melting right in front of people’s eyes.


…celsius, for contiguous states.

These GISS maps show one point in time (in this case month) compared to an average of a 30-year period for Oct. It is interesting to compare the map for October 1947 here with the Oct 2008. Why Oct 1947 – it was about the peak of the previous cycle in Siberia.
Here is a graph of HadCRU October data (1900 – 2007 with anomaly base period 1961-1990) for 9 grids in Siberia link. And here (link) is the same data averaged for the 9 grids.
The current warming is similar to the 1940s.


please remember that these are ANOMALY charts!!! If the average temp for the baseline is -25C, the -17C higher temp will still freeze ocean water very nicely!!!
Of course, that does bring into question why the Arctic didn’t refreeze as quickly earlier this decade with a cooler anomaly and alledgedly cooler ocean!! 8>)
Here you can see actual measured air temps from the Arctic:

Mike Bryant

Maybe the legend should say:
“Now compare the above corrected erroneous version with the first erroneous one below:”


What’s the next number in this sequence: 1 2 4 ?
If you guessed ‘8’, go to the back of the class.
It’s 13.7.

“The Engineer” @ 21:01:29 —
“The arctic in October.
Must be all that ice forming
thats doing the warming !”
The National Snow and Ice Data Center, Nov. 10 —
“As is normal for this time of year, ice extent increased rapidly through most of October. However, this year, the increase was particularly fast, which contributed to above-average air temperatures near the surface.”

Bill Jamison

So it was “only” an error of 25%, no big deal right?
We know for a fact that some of the data originally used for the Oct anomaly was erroneous. Why do we trust the data for the other sites? Why do we trust the data going back to the beginning of the GISTEMP?
IMO once you’ve proven data errors then the rest of the data is suspect until proven accurate.


Why is Australia and surrounds different in this chart as opposed to the one with bad Russian data…. The scale is the same. Only the values above +8C have changed in the scale on both charts…… So why the different charts for the Australian area… Only Russia should have changed.
More GISS shenanigans….. They never stop.


I do prefer the polka dots breaking out all over. Because the dots can be counted, I think that map is more quantifiable. 🙂


Ok so we are discussing October 2008 and significant warming is still observed. And no, it is not an embarrasment, something like this can happen and does happen to every organization at times. The error was fixed so quickly, and the warming trends are still quite clear. So again I ask where is the major discrepancy?


Steve McIntyre’s hot on the trail of another mystery. The hot spot in the Canadian Arctic is not supported by data.

I did a post on this concurrent with Anthony’s post and noticed the same issues with northern Canada and Australia.
For my post, I downloaded the new GHCN data and compared it to Monday’s version (which I had saved) and checked for stations with Oct 2008 data today that didn’t have data on Monday. All but 2 of the stations were in Australia (the other two were SH). So there is new information justifying changes in Australia but no new information from GHCN for northern Canada.
So what changed between Monday and Wednesday that enabled NASA to color the Canadian Arctic Islands red?

Drew Latta

Re: that the icecap keeps growing at these temps
My personal, totally unproven, guess is that perhaps you need to have significant open ocean in the Arctic to start to build continental ice sheets. Consider that you need some source of moisture in the Northern Hemisphere to build continental ice sheets… What better to do that than an open water source in the Arctic? Perhaps if the stars align (the solar forcing curve from Milankovitch cycles) correctly at the start of an ice age you get more of an open ocean condition in the Arctic which drives evaporation in the Arctic into overdrive to start that feedback loop on land to build ice sheets. Otherwise Laurentia and N. Central Canada are a long way away from moisture sources to overwhelm ablations by melting and sublimation. Certainly something has to change to put more moisture over the Northern lattitudes to build ice sheets to initiate an ice age… and I haven’t seen any good explanations that define what boundary conditions would change a arctic desert into a moist place to build up miles of ice cap. That’s my $0.02, though I don’t have any huge basis for it…

April E. Coggins

I’m curious to know what Australians think of these maps. Both maps show higher than normal temps in Australia’s most southern (coldest) regions and normal everywhere else. Is snow in Sydney normal?
jcbmack: Because global warming is a lie. We are alarmed at the rate and number of falsehoods that promote the global warming scare which will cost millions of people their lives.

Drew Latta

Huh, I misread the posts. I don’t believe that anyone actually said that icecaps were growing in this discussion. I must have been misplacing the recent info about the Alaskan glaciers having a positive mass balance. But my hypothesis still stands, you still need something to drive ice buildup in the North and there has to be a source. I’d love to be pointed in the right direction of a paper or two that explains how this happens…

Neil Crafter

do you work for NASA GISS? With all your justifying of their stuff-up one can imagine you have some sort of vested interest in this.


The main discrepency was already addressed – GISS had way too high of an anomaly for October due to faulty information from the NOAA they failed to catch.
But there is still discrepencies with the new map. Satellite temperature measurements do not indicate that Russia or even most of Asia as the GISS map indicates. In addition, as others have pointed out, if the Russian Arctic was so much warmer than normal, how did the ice grow there at a record pace in October? It just doesn’t add up. Also, I was following temperatures in Arctic Russia during the last month watching the ice grow, and they did not seem to be much above normal at all.
One more note of caution regarding GISS: if you look back through their previous maps, you will notice Siberia has had way more major “heat waves” than any other place on the planet. This seems rather odd, not just because it is consistently the same area recording the biggest anomalies, but also because it is a huge area that is very sparsely populated…

Phillip Bratby

Strange how the grey areas have shrunk to be replaced by a positive anomaly. How can you have an anomaly from something that wasn’t there before???
And in the UK we’ve gone from [+1 to +2C] to [-0.5 to -1C]. I didn’t think it had been that much warmer than normal.


Steve McIntyre:
“So what changed between Monday and Wednesday that enabled NASA to color the Canadian Arctic Islands red?”
They got a new red crayon to replace the one they wore out in Russia?


The original map looks like it has the UK and Ireland too warm also.
Anyone can make a mistake but some people seem to always make the mistake in the right direction.
Off topic: reading in new scientist two articles this morning. I am no climate scientist or indeed a scientist but i am not convinced.
“Humans may have prevented super ice age ”
“The ice age that never was”

Larry Huldén

I checked on NASA’s website and created the map for October but it is still today (Oct 13 in Finland) the same as that printed at Oct 10. From where is the new map reproduced?
Also the graph showing zonal mean (anomaly) from -90 – +90 degree latitude is identical with that from Oct 10 ! I printed out them both on Oct 10 and Oct13.
Larry Huldén
Finnish Museum of Natural History


Why is latest data allowed to “adjust” previous data?


hehe it is Nasa afterall, you sure they converted degree to celsiusor visa versa?? ,, I know that was mean but I found images of Mars landers smashing into the ground flashing through my head..
“Isn’t it odd that the Arctic ice has refrozen at rates never seen in the satellite age, during October, yet it was warmer in Siberia? Strange indeed.”
Here in the NW there has been a ton of rain.. a little atmospheric jumpstart for the ice north of us? I was thinking we would be a warmer spot because we havent gotten the cold snap that usually comes with clear skys here in the winter.. good thing though or we would be snowed in.

Tim Groves

Rick: “Amazing how the icecap keeps growing in these high temperatures.”
Oh, that’s an easy one to spin. As seawater turns to ice, the thermal energy it loses is released into the surounding air where it becomes trapped by that evil gas carbon dioxide, and so the greater scale of the freeze, the warmer the atmosphere becomes, and of course this stresses out all the starving polar bears and the poor little baby seals.

[…] lot’s of material – (and this is the tip of the iceberg) Corrected NASA GISTEMP data has been posted __________________ ~Paradox "In a time of deceit telling the truth is a revolutionary […]


Since ice absorbs a lot of heat when it melts, it necessarily follows that when water freezes it emits a great deal of heat. So bare water (or thin ice) should cause the North to be warm. As to whether this would happen in Siberia or Canada, wouldn’t that depend on the direction of the winds?
As soon as the ice pack returns to normal (with normal thicknesses so as to provide the normal insulation for the air), the effect (i.e. of water freezing giving off heat to the air) will go away and the awfully cold winter that I think we have ahead of us will continue.

Stan Jones

Why would they publish the original figures with the whole of Australia missing in the first place?

[…] Two pretty pictures… colourful, anway… though they did not manage to fool “Corrected NASA GISTEMP data has been posted” Anthony Watts or “Watch the Pea” Steve […]

GISTEMP anomalies for October for the past 30 years, with the revised data:

The arctic ‘October heat wave’ is the signature of the rapid freezing of the arctic sea ice, because a lot of latent heat is freed.
Remember, we had roughly 2 million km2 less ice extent in this September than normally.
Rapid freezing occured in October, especially along the Siberian coast. At the end of October, the sea ice extent was almost back to normal.
This freezing was driven by very low arctic temperatures. Every kg of water frees 80 kcal or 330 kiloJoule of latent heat, equivalent to heating 80 kg of (liquid) water by 1degree Centigrade- or roughly 100 cubic meter of air at surface pressure by 1 degree. 1m3 of water will thus cause a 1 degree temperature increase for 100.000 cubic meters of air at sea level pressure. This is a column of air of 100 m2 which ist 1000 m high. Correct then for the air pressure, and you have a column of air extenting almost to the end of the troposphere.
When we assume for simplicity that 1 miilion km2 of ocean surface water freezes to 1 m depth (equivalent to 1000 km3), then 100 million km2 of surrounding area will be affected by a 1 degree temperature increase in the atmosphere, or the whole arctic area of 30 million km2 gets a 3 degree increase.
What do we learn:
-As long as there is excess arctic sea ice melting during summer, there will be an ‘October (latent) heat wave’ in the arctic. These two effects are surely correlated: the one causes the other.
-The satellite data do show a signature of this effect in their arctic data, but do not show an increase in their global number (500 vs 30 million km2). It will be interesting what Hadcrut tells us and what the GISS people will tell us when they have fixed their problem. We may learn more about the systematic differences between the surface based and satellite data.
-As the relatively warmer arctic air could take up more water vapor, precipitation may have increased, which would be consistent with the large snow covered areas of the arctic, seen, e.g., on the Cryosphere Today (UIUC) maps . Albedo change should follow: we may get a rather cold winter.