![image42[1]](http://wattsupwiththat.files.wordpress.com/2014/05/image421.png?quality=75)
By Paul Homewood A guest post earlier today by David Dohbro, comparing satellite and surface temperature datasets, appears to have attracted a certain amount of criticism, not least because it has tried to compare anomalies based on different baselines. This is an update to that analysis.
To get around this issue, I am posting a comparison of GISS surface temperatures and RSS satellite numbers, both based on a baseline of 1979-98, which is the one used by RSS. (GISS still use 1951-80).
This change of baseline means that all GISS temperatures anomalies are reduced by 0.28C.
The graph below shows the GISS minus RSS temperature anomalies from January 1979, the start of the satellite record, to April 2014. I have not shown individual months, as monthly variations can be quite large, and would simply serve to muddy the waters. Instead, I have shown the 12-Month average of the differences.
There are some marked ups and downs in the record up to about 1995. This may be due to early inaccuracies in the satellite dataset, though this would be pure speculation on my part.
Since then, though, certain patterns emerge:
1) During the major El Ninos of 1998 and 2010, RSS temperatures spike to higher levels than GISS. In other words, GISS minus RSS goes negative.
2) There was a period of a few years after 1998, when the two datasets ran closely together.
3) From around 2004, though, there has been a distinct and growing divergence, with GISS showing relatively more warming.
4) Over the full period, as the graph shows, there is a trend is towards greater warming on GISS.
It is also worth noting that the April 2014 temperature anomalies, based on the same 1979-1998 baseline, are as follows:
| GISS | RSS |
| 0.45C | 0.25C |
In other words, GISS are currently showing 0.20C more warming than RSS. This is a significant number, being nearly half of the observed warming, as measured against the 1979-98 average.
The purpose of this post is not to point the finger, and say that one or the other dataset is wrong. Nevertheless, I believe that the differences observed raise serious questions about the accuracy of our global temperature measurements.
See the original Dohbro post
Sources
1) GISS –
http://data.giss.nasa.gov/gistemp/tabledata_v3/GLB.Ts+dSST.txt
2) RSS
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
In order to compare recent divergence, I matched the earliest part of each series, Nick.
In my humble opinion, the whole point of the post of David Dohbro was the comparison based on land data. This shift towards global data misses the mark.
Sometimes I get up in the morning and I can’t tell if it’s 0.2 degrees too warm or 0.4 degrees too warm
Now, this sort of discussion is why I started reading this site in the first place. I make not pretenses of being a mathematician, so I value greatly the input of people who have these technical skills.
Same with the recent discussion about why Goddard’s recent graph and statements needed “correcting” (to be polite about it). A blind man could see that something funny was going on, but without the work of some of the people with valuable experience in the field, I’d never be able to explain to someone exactly what the problem was.
So, thanks, and keep it up.
We are all victims of this insanity, rolling through a period of benign climate arguing about trivial changes. I for one would like to frac a few thousand gas wells in Africa and improve some lives.
@Nick Stokes
“This just gets dumb. After all the fuss of the predecessor of this plot, you’ve shown a plot of RSS
vs UAH with dikfferent baselines.
RSS ’79-’97, UAH 1981-2010”
Here’s a hint, I wrote “simple eye ball test” for a reason.
Although I’m familiar with the game being played here. In order to duck your original deception, you latch onto any irrelevancy you can find and run with that… Oh well.
Zeke:
The point is there is an issue with one of the series, which you (and Nick) apparently totally wiffed on.
Actually the real issue is “should GISTEMP even agree with UAH or RSS?” They don’t measure the same thing… one measures 1-m temperature the other lower tropospheric temperature.
You need a model to decide if the difference matters. Do you have a reliable one? Because if you don’t you can’t make an argument whether the direct comparison is valid or not.
It is a bigger issue that UAH is running much “warmer” than RSS, because they are in principle measuring the same thing.
Nick Stokes:
The point of aligning on one end is to you can visually compare the slopes (trend comparison).
If you align in the middle, you’re comparing the “wiggles” (correlation).
The difference between the series is more interesting, though.
It does seem like something is going on with RSS.
It’d be interesting to compare the trends by zonal (latitudinal) average, but I don’t have time to play with that.
Will Nitschke says: May 19, 2014 at 11:57 pm
“I wrote “simple eye ball test” for a reason”
What you wrote was:
<i."As you can see using a simple eye ball test, RSS runs warmer for most of this period. (One exception is a few months around Dec ’12 into ’13.)"
The reason why RSS “runs warmer” is that the different anomaly base raises it by 0.1°C. Get that right, as WFT does, and it looks quite different. If I took your judgmental approach, I would call that misleading.
Carrick,
“Actually the real issue is “should GISTEMP even agree with UAH or RSS?” They don’t measure the same thing… one measures 1-m temperature the other lower tropospheric temperature.”
That’s one to take up with the author of this post (I agree).
“The point of aligning on one end is to you can visually compare the slopes (trend comparison).”
That plot wasn’t aligned anywhere. It showed just two years and said, look, RSS and UAH are about the same.
“It does seem like something is going on with RSS.”
Your plot is similar to Roy Spencer’s, though he has aligned the anomaly bases. And as I quoted above, he does have views on what is going on with RSS.
Carrick:
In your post at May 20, 2014 at 12:17 am which is addressed to Zeke:You say
Yes! None of the various global and hemispheric temperature products (i.e. RSS, UAH, GISS, HadCRUTn, etc.) actually measures the “same thing”. The teams which produce these products each uses a unique definition of global average temperature (GAT) and each team frequently changes the definition it uses: these changes provide the “adjustments” which Anth0ny Watts says at May 19, 2014 at 8:10 pm means
(His witticism is a truism – see http://jonova.s3.amazonaws.com/graphs/giss/hansen-giss-1940-1980.gif – and is also true of other GAT data sets.)
However, the sophistries of e.g. Nick Stokes are enabled by the lack of a definition of GAT. Anybody can change both meaning and emphasis of any statement about a parameter which is whatever somebody wants it to be at any time.
And there are two basic understandings of what GAT may be; viz.
(i) GAT is a physical parameter that – at least in principle – can be measured;
or
(ii) GAT is a ‘statistic’; i.e. an indicator derived from physical measurements.
These two understandings derive from alternative considerations of the nature of GAT.
If the GAT is assumed to be the mean temperature of the volume of air near the Earth’s surface over a period of time, then GAT is a physical parameter indicated by the temperature sensors (i.e. thermometers mostly at weather stations or the microwave sounding units mounted on satellites) that is calculated using the method of mixtures (assuming unity volume, specific heat, density etc.).
Alternatively:
If the the temperature sensors are each considered to indicate the air temperature at each measurement site and time, then GAT is a statistic that is computed as being an average of the total number of sensor indications.
Either of these understandings provides problems.
(1) If GAT is considered as a physical parameter that is measured, then the data sets of GAT are functions of their construction. Attributing AGW – or anything else – to a change that is a function of the construction of GAT is inadmissable. Furthermore, the compilers of the data sets each frequently alter their definition – and, therefore, their method of calculation – of GAT. Therefore, the obtained values of GAT cannot be validly compared to anything because it is not known what is being compared.
And
(2) If GAT is considered to be a statistic then it can be computed in several ways to provide a variety of results, each of different use to climatologists. (In such a way, the GAT is similar in nature to a Retail Price Index, which is a statistic that can be computed in different ways to provide a variety of results, each of which has proved useful to economists.) If GAT is considered to be a statistic of this type, then GAT is a form of average. In which case, the word ‘mean’ in ‘mean global temperature’ is a misnomer, because although there are many types of average, a set of measurements can only have one mean. Importantly, if GAT is considered to be an indicative statistic then the differences between the values and trends of the data sets from different teams indicate that the teams are monitoring different climate effects. But if the teams are each monitoring different climate effects then each should provide a unique title for their data set that is indicative of what is being monitored. Also, each team should state explicitly what its data set of GAT purports to be monitoring.
In summation, nobody knows what a value of global average temperature (GAT) represents, and what it represents is frequently altered, so it is not possible to validly compare a value of GAT to any other datum.
Richard
MaxLD
An inquiry. You indicate you have used a 12-month running average. Is this correct? This is a lagging indicating with respect to the time. Is that what you want to show? This is commonly used in stock analysis.
That’s right. The 12 months ending April 14, is shown on the graph at April 14.
I realise some people prefer centred averages.
Interesting discussion. 🙂
Extending the analysis to UAH, and still using 1979-98 baselines, we get these anomalies for April 2014
GISS +0.45C
RSS +0.25C
UAH +0.29C
So, UAH runs slightly warmer than RSS, but both are currently much colder than GISS.
Paul Homewood says: May 20, 2014 at 3:22 am
“So, UAH runs slightly warmer than RSS, but both are currently much colder than GISS.”
Your numbers are close to the scaling on the plot here. You can see that UAH and RSS are unusually close in April. Over the last year, RSS has been coolest, except in February, and mostly UAH closer to GISS.
A cynical person might suspect that the big negative around 1998 was a deliberate effort to avoid the Bob Beamon effect. Can’t have global warming if you’re not breaking records.
You’re complaining that we’re not ideological regurgitards?
Hi Paul – Excellent post.
On the divergence of the surface and lower tropospheric trends, see our papers
Klotzbach, P.J., R.A. Pielke Sr., R.A. Pielke Jr., J.R. Christy, and R.T. McNider, 2009: An alternative explanation for differential temperature trends at the surface and in the lower troposphere. J. Geophys. Res., 114, D21102, doi:10.1029/2009JD011841. http://pielkeclimatesci.wordpress.com/files/2009/11/r-345.pdf
Klotzbach, P.J., R.A. Pielke Sr., R.A. Pielke Jr., J.R. Christy, and R.T. McNider, 2010: Correction to: “An alternative explanation for differential temperature trends at the surface and in the lower troposphere. J. Geophys. Res., 114, D21102, doi:10.1029/2009JD011841”, J. Geophys. Res., 115, D1, doi:10.1029/2009JD013655. http://pielkeclimatesci.wordpress.com/files/2010/03/r-345a.pdf
McNider, R.T., G.J. Steeneveld, B. Holtslag, R. Pielke Sr, S. Mackaro, A. Pour Biazar, J.T. Walters, U.S. Nair, and J.R. Christy, 2012: Response and sensitivity of the nocturnal boundary layer over land to added longwave radiative forcing. J. Geophys. Res., 117, D14106, doi:10.1029/2012JD017578. Copyright (2012) American Geophysical Union. http://pielkeclimatesci.files.wordpress.com/2013/02/r-371.pdf
Roger
richardscourtney says (May 20, 2014 at 2:18 am): “Yes! None of the various global and hemispheric temperature products (i.e. RSS, UAH, GISS, HadCRUTn, etc.) actually measures the “same thing”. The teams which produce these products each uses a unique definition of global average temperature (GAT) and each team frequently changes the definition it uses…”
Thank you for yet another reminder. When referring to so-called global average temperature, I try to include the words “so-called”, but often forget. Keep reminding us. 🙂
Any chance you could expand your comment into a full article some time?
Nick,
a climate model, which does not quite match reality.
No kidding.
If one has a large number of good thermometers in one spot and averages them
FIFY
Gary Hladik:
At May 20, 2014 at 11:59 am you ask me
Although it lacks the illustrations, I think what you want is Appendix B of this
http://www.publications.parliament.uk/pa/cm200910/cmselect/cmsctech/memo/climatedata/uc0102.htm
Richard
Carrick,
I think both Nick and I agree that the divergence between UAH and RSS is the interesting comparison to make. GISS doesn’t diverge at all from other surface records (NCDC and Hadley) once you account for differences in spatial interpolation methods. Both Berkeley and C&W produce similar (or greater) warming to GISS in recent years because they spatially interpolate their fields globally.
There is also the difference between satellite records in and surface records in general, though the cause is much more difficult to determine given that they are measuring different things and are each subject to different potential biases or sources of error.
The “what happened to UAH” question was a subtle dig at the tendency of folks to cherrypick the series that shows the results they want to see. Hence the preference of GISS in some quarters, or the switch from showing UAH to showing RSS once RSS started reporting cooler temperatures.
Nick wallows in minor math, as if.
What smoke!
Surface data are heterogeneous. To understand, we must compare the components.
http://img215.imageshack.us/img215/5149/plusuah.png