NOTE: Please note that part 2 is now online, please see it here.
I recently plotted all four global temperature metrics (GISS, HadCRUT, UAH, RSS) to illustrate the magnitude of the global temperature drop we’ve seen in the last 12 months. At the end of that post, I mentioned that I’d like to get all 4 metrics plotted side-by-side for comparison, rather than individually.
Of course I have more ideas than time these days to collate such things, but sympathetic reader Earle Williams voluntarily came to my rescue by collating them and providing a nice data set for me in an Excel spreadsheet last night.
The biggest problem of course is what to do with 4 different data sets that have different time spans. The simplest answer, at least for a side by side comparison is to set their time scales to be the same. Satellite Microwave Sounder Unit (MSU) data from the University of Alabama, Huntsville (UAH), and Remote Sensing Systems (RSS) of Santa Rosa, CA only go back to 1979. So the January 1979-January 2008 period is what we’ll concentrate on for this exercise as it very nearly makes up a 30 year climate period. Yes, I know some may call this an arbitrary starting point, but it the only possible point that allows comparison between land-ocean data-sets and the satellite data-sets.
Here is the first graph, the raw anomaly data as it was published this month by all the above listed sources:
Here is the source data file for this plot and subsequent plots.
I also plotted a magnified view to show the detail of the lat 12 months with notations added to illustrate the depth of the anomaly over the past 12 months.
March 2005 to January 2008, magnified view – click for larger image
I was particularly impressed with the agreement of the 4 metrics during the 1998 El Niño year as well as our current 2008 La Niña year.
I also ran a smoothed plot to eliminate some of the noise and to make the trends a bit more visible. For this I used a 1 year (12 month) average.
Again there is good agreement in 1998 and in 2008. Suggesting that all 4 metrics picked up the ENSO event quite well.
The difference between these metrics is of course the source data, but more importantly, two are measured by satellite (UAH, RSS) and two are land-ocean surface temperature measurements (GISS, HadCRUT). I have been critical of the surface temperature measurements due to the number of non-compliant weather stations I’ve discovered in the United States Historical Climatology Network (USHCN) from my www.surfacestations.org project.
One of the first comments from my last post on the 4 global temperature metrics came from Jeff in Seattle who said:
Seems like GISS is the odd man out and should be discarded as an “adjustment”.
Looking at the difference in the 4 times series graphs, differences were apparent, but I didn’t have time to study it more right then. This post today is my follow up to that examination.
Over on Climate Audit, there’s been quite a bit of discussion about the global representivity of the GISS data-set due to all of the adjustments that seem to have been applied to the data at locations that don’t seem to need any adjustments to compensate for things like urban heat islands. Places like Cedarville, CA and Tingo Maria, Peru both illustrate some of the oddities with the adjustment methodology used by NASA GISS. One of the issues being discussed is the application of city nightlights (used as a measure of urbanization near the station) as a proxy for UHI adjustments to be applied to cities in the USA. Some investigation has suggested that the method may not work as well as one might expect. There’s also been the issue of whether of not stations classified as rural are truly rural.
So with all of this discussion, and with this newly collated data-set handed to me today, it gave me an idea. I had never seen a histogram comparison done on all four data-sets simultaneously.
Doing so would show how well the cool and warm anomalies are distributed within the data. If there is a good balance to the distribution, one would expect that the measurement system is doing a good job of capturing the natural variance. If the distribution of the histogram is skewed significantly in either the negative or positive, it would provide clues into what bias issues might remain in the data.
Of course since we have a rising temperature trend since 1979, I would expect all 4 metrics to be more distributed on the positive side of the histogram as a given. But the real test is how well they match. All four metrics correlate well in the time series graphs above, so I would expect some correlation to be present in the histogram as well. The histograms you see below were created from the raw data from 1979-2008. No smoothing or adjustments of any kind were made to the data. The “unadjusted” data in this source data file were used: 4metrics_temp_anomalies.txt
First we have the satellite data-set from UAH:
University of Alabama, Huntsville (UAH) Microwave Sounder Data 1979-2008 – click for larger image
The UAH data above looks well distributed between cool and warm anomaly. A slight warm bias, but to be expected with the positive trend since 1979.
Next we have the satellite data-set from RSS:
Remote Sensing Systems (RSS) Microwave Sounder Data 1979-2008 – click for larger image
At first I was surprised at the agreement between UAH and RSS in the percentages of warm and cool, but then I realized that these data-sets both came from the same instrument on the spacecraft and the only difference is methodology in preparation by the two groups UAH and RSS. So it makes sense that there would be some agreement in the histograms.
Here we have the land-ocean surface data-set from HadCRUT:
Hadley Climate Research Unit Temperature data 1979-2008 – click for larger image
Here, we see a much more lopsided distribution in the histogram. Part of this has to do with the positive trend, but other things like UHI, microsite issues with weather station placement, and adjustments to the temperature records all figure in.
Finally we have the GISS land-ocean surface data-set:
NASA Goddard Institute for Space Studies data 1979-2008 – click for larger image
I was surprised to learn that only 5% of the GISS data-set was on the cool side of zero, while a whopping 95% was on the warm side. Even with a rising temperature trend, this seems excessive.
When the distribution of data is so lopsided, it suggests that there may be problems with it, especially since there appears to be a 50% greater distribution on the cooler side in the HadCRUT data-set.
Interestingly, like with the satellite data sets that use the same sensor on the spacecraft, both GISS and HadCRUT use many of the same temperature stations around the world. There is quite a bit of data source overlap between the two. But, to see such a difference suggests to me that in this case (unlike the satellite data) differences in preparation lead to significant differences in the final data-set.
It also suggests to me that satellite temperature data is a more representative global temperature metric than manually measured land-ocean temperature data-sets because there is a more unified and homogeneous measurement system, less potential bias, no urban heat island issues, no need of maintaining individual temperature stations, fewer final adjustments, and a much faster acquisition of the data.
One of the things that has been pointed out to me by Joe D’Aleo of ICECAP is that GISS uses a different base period than the other data-sets, The next task is to plot these with data adjusted to the same base period. That should come in a day or two.
UPDATE1: I’ve decided to make this a 3 part series, as additional interest has been generated by commenters in looking at the data in more ways. Stay tuned for parts 2 and 3 and we’ll examine this is more detail.
UPDATE2: I had mentioned that I’d be looking at this in more detail in parts 2, and 3. However it appears many have missed seeing that portion of the original post and are saying that I’ve done an incomplete job of presenting all the information. I would agree for part1, but that is what parts 2 and 3 were to be about.
Since I’m currently unable to spend more time to put parts 2 and 3 together due to travel and other obligations, I’m putting the post back on the shelf (archived) to revisit again later when I can do more work on it, including show plots for adjusted base periods.
The post will be restored then along with the next part so that people have the benefit of seeing plots and histograms done on both ways. In part 3 I’ll summarize 1 and 2.
In the meantime, poster Basil has done some work on this of interest which you can see here.
UPDATE3: Part 2 is now online, please see it here.
[…] A look at temperature anomalies for all 4 global metrics […]
I am not a scientist, but the explanation for the above seems very obvious to me.
The basic principle that James Hansen and GISS work on is that there is a strong correlation in trend between neighbouring weather stations. Neighbouring stations are defined as any within 1200km (or 745 miles). Therefore any station is going to be under suspicion if its trend does not correlate well with two “neighbouring” stations in a radius of 745 miles.
If we assume that there is a general small warming trend as shown in the RSS and UAH data this will mean that most stations will show a positive trend. The minority that show a genuine cooling trend will therefore come under suspicion and when compared to their neighbours will need to be adjusted upwards so that the “error” is removed. Of course this upward adjustment of stations creates its own positive feedback, because these adjusted stations can then be used to justify making upward corrections in other stations. And so it goes on.
But the satellite data was “adjusted” to match the surface station data, because the sat data diverged, and was thought to be wrong?
Once again, when the issue of basic climate data is examined in depth, the data pumped out by our own government, generated by our tax money, is heavily biased toward a global warming conclusion. Hansen’s NASA data is clearly out of line with these other temperature sets. Hansen needs to be constantly vetted and called into account for his data sets. As a goverment employee, he should be held to the highest standard to examine and remove any biases from his work. If he continues to stonewall requests for all his data sources AND adjustments, he should be removes and replaced with a real scientist, not an agenda driven self-promoter.
Obviously the satellite data needs to be “adjusted” more! Where’s Hansen?
Facetiousness aside, this is very interesting, great idea to use the histogram to look at this data in another way. Often looking at data using a different technique will yield interesting insights, and this appears to be one of those cases. It’s hard to imagine that much of a shift without a significant warm bias in both the raw ground data and the adjustments made to allegedly correct it for UHI effects. As is being shown with your surface stations effort, and Steve McIntyre’s efforts looking at the code for the adjustments, we’re finally seeing all the warts on the ground data, and why it should not be relied upon.
Excellent work.
Anthony:
Can you spell out for those of us who are slow on the uptake, which anomaly you are referring to. Also doesn’t the difference in the positions of the satellite data series compared to the ground data essentially anticipate what you see in the histograms?
So the task now is to identify the “good” CRU and GISS sites and see if they will give a distribution similar to UAH/RSS. That is, will the most pristine sites eliminate what certainly looks like a systematic warm bias? I know folks are well on the way to doing this, but how would one or two particularly well known good sites compare to the satellite histograms?
Perhaps one of the agencies is using new math in its calculations and 2+2 might = 5. I find it hard to understand how the basicly the same instruments
(raw data) can have such divergance with out there being some sort of adjustment/bias being introduced into the data. I would wonder how the satalites have been calibrated over the years that the one agency with it wwould seeem more control would have come more close to the median of the satalite data.
just a few thoughts (probably not coherent)
Bill
I have to admit the first comparison GISS data doesn’t look quite as bad as in your earlier post. It’s also interesting that the more smoothing or filtering done it looks progressively less an outlier from the shape of the curve.
The histogram, however, is a telling comparison. Good analysis!
Anthony,
You should make it clear what the base period is for each of the datasets. You may be comparing apples and oranges if the base periods are different.
Hey! I got my name in lights! hehe.
And of course, as has been pointed out before, choosing 1979 as an anomaly starting point is extremely arbitrary. No fault of yours, Anthony, I understand that, but really, who gets to decide what the anomaly starting point is? Choosing a cold year is cherry-picking, choosing a warm year is cherry-picking. Therefore a mean of the entire century should be taken and that used as the measuring stick for any anomaly. Still, such a thing would have very little
value, in my opinion.
REPLY: The choice was not arbitrary, it was so that all four sets macthed in time scale. Satellite data starts in 1979.
On the anomaly charts, there is a clear difference between the satellite and GISS data to the right of the 1998 El Nino event. The satellite plots feature a head in 1998 and clearly defined shoulders on either side. The shoulders on the right are a little higher than those on the left but not by much, and El Nino is clearly a head (peak). In contrast, the GISS plot also shows the distinct rise in 1998, but the shoulders to the right are sloping upward and are level with the 1998 peak, forming what I call a hunchback.
Satellite plot: head and shoulders
GISS plot: hunchback
In summary, it appears that GISS is exploiting weaknesses in the SST measuring system and that you guys (especially CA) are incrementally revealing the tricks of their trade.
That’s raw GISS data. I have the GISS-adjusted, homogenized (and probably sissified) data, and it puts 2005 as the warmest year.
1998: .71
1999: .46
2000: .42
2001: .57
2002: .69
2003: .67
2004: .60
2005: .76
2006: .66
2007: .73
Those are the last 10 years of GISS “anomolies”. Cooked.
http://data.giss.nasa.gov/gistemp/tabledata/GLB.Ts.txt
Note that 1998 clocks in at #3 on the hit parade. (2007 is #2.) They didn’t even “lose” El Nino. Just dumbed ‘im down.
On closer inspection, I see that GISS 2005 and 2007 are higher than 1998, even on your graph. (That’s also the case for NOAA data, come to think of it.)
Watt’s up with that?
REPLY: This data set is more “current” than the one from last summer.
And so it goes on.
Can’t you just see them adjusting UHI by comparing them with those CRN-4 and CRN-5 rural stations? (Thus concluding that UHI isn’t really all that much of a factor.)
fascinating. long time lurker (I read your stuff at Climate Audit) first time poster.
putting up all 4 data sources just proves to me that something weird happened over the last year.
Too bad the science is settled. I’d love to find out what it was.
(I’m being sarcastic about the science being settled.)
I mentioned that, or tried to. But is 1979 the anomaly point?
REPLY: Ah ok I see what you are getting at. I’ll put that up.
“When the distribution of data [GISS and HadCRUT] is so lopsided [compared to RSS and UAH data], it suggests that there may be problems with it”
They’re actually measuring two very different things, as I’m sure you know. And it’s possible for them both to be correct. The temperature at ~2m [GISS, HadCRUT] is likely going to be related to the integrated brightness temperature from the ground to 10km (with appropriate weightings) [RSS, UAH]. The fact that the 4 temperature measurements are so similar, in my opinion, gives weight to the fact that the rise in temperatures is not an artifact of measuring technique.
“But, to see such a difference [in the surface temperatures] suggests to me that in this case (unlike the satellite data) differences in preparation lead to significant differences in the final data-set.”
GISS has a method, right or wrong, that estimates the poles. I don’t think HadCRUT does, which could explain the difference.
Raven: “You may be comparing apples and oranges if the base periods are different.”
The base periods are different. GISS is always above HadCRUT, which is nearly always above the other two. This means that GISS has a base period where the temperature is the lowest, HadCRUT has a slightly higher temperature in its base period, and the satellite records have an even higher (but about equal) temperature during their base periods.
I forgot to mention, if you didn’t adjust the data in the 4 time series so that they had the same base period any conclusions you draw from the histogram analysis are likely to be wrong. I only mention this because it doesn’t appear that this occured. The easiest way to do this is to set 1979-2008(Jan) as the base period, and subtract the mean out of all the time series above. I suspect you’ll find the histograms in much better agreement.
REPLY: I’ve been corresponding with Joe D’Aleo at ICECAP on the very subject of the difference in base period. I have the adjusted base period data in hand, courtesy of Earle Williams, and I’ll be plotting that next and doing the same analysis. It will be interesting to see if the distribution in the histograms changes significantly. Stay tuned.
Here is a comment from Dr. Christy on CA:
Now, I have one misrepresentation to point out on Steve M.’s charts. The temperature comparisons shown are not apples to apples. All climate models indicate the global tropospheric temperature should warm at a rate of 1.2 times that of the surface (1.4 times that of the surface for the tropics – see CCSP SAP 1.1. or Douglass et al. 2007). So, to put surface temperature projections from models on a chart with observed tropospheric temperatures, one must reduce the tropospheric temperature trend by a factor of 1.2 for the comparison to be legitimate. I think the result would be of interest to the readers, and it is entirely defensible as shown in numerous publications.Adding this factor would compensate for the differences between satellite and surface measurements which Atmoz noted.
REPLY: Thanks Raven, good catch. This is all an ongoing learning experience, and I welcome this kind of input! Thanks to Earle Williams, I have the datasets adjusted to a common baseline, I’ll plot those next, then I’ll attempt what Dr. Christy suggests.
Looks like this will end up being a 3 parter then.
But they forgot the burping cows. http://www.foxnews.com/story/0,2933,332956,00.html
Atmoz beat me to the punch. Since GISS uses a cooler base period than the others, more of it’s anomalies will be positive rather than negative. They need to be compared using the same “normal” period or else it is not a valid comparison. Also, just so I am clear, I believe you indicated you were using the GISS data prior to homogenization. Is this correct? It would be especially interesting to see a comparison of both the un-homogenized and homogenized data sets to the other metrics to see just how much of a difference the homogenization protocols really make.
Thanks
Bob
Is this dataset available for download anywhere?
I had the same thought as D’Aleo about the difference in base period, so it will be interesting to adjust for that and see what we have.
Either way, I’ve some regressions I’d love to run on the data if it is available.
REPLY: Thanks See this original post which started it all. There are links to the datasets under each graph. I’ll be happy to post whatever your analysis brings, as I am interested too.
http://wattsupwiththat.wordpress.com/2008/02/19/january-2008-4-sources-say-globally-cooler-in-the-past-12-months/
I took the data provided in the original post and subtracted the mean from each time series, and plotted the 4 new time series and histograms. I don’t have a quick way to make the cool histogram plots that Anthony does, but they are much more similar. I look forward to seeing your analysis.
http://atmoz.org/blog/2008/02/27/4-global-temperature-anomalies-say-the-same-thing/
REPLY: Thanks for doing that, the real question though is why does GISS use a different base period than the other data sets?
More on this later as I promised.
Anthony: In the 12-month Smoothed 1979 to 2008 graph, a major difference between sets shows up in the HADCRUT data after the 97-98 El Nino. For each of the others, their respective minimum temperatures in the troughs preceding and following the El Nino are approximately the same value. Not the HADCRUT. The minimum temperature in the trough that follows is about 0.15 deg C higher than the minimum temperature of the preceding trough. I’ve noticed it in their data for years, but I didn’t know it didn’t exist in the others. What did Hadley change around then?
REPLY: Don’t know he answer to that? Anyone?