By Steve Goddard and Anthony Watts
Some of the excellent readers of the last piece we posted on WUWT gave me an idea, which we are following up on here. The exercise here is to compare GISS and satellite data (UAH and RSS) since the start of 2003, and then propose one possible source of divergence between the GISS and satellite data. The reason that the start of 2003 was chosen, is because satellite data shows a rapid decline in temperatures starting then, and GISS data does not. The only exception to the downward trend was an El Nino at the start of 2007, which caused a short but steep spike. Remembering back a couple of years, Dr. Hansen had in fact suggested that El Nino might turn into a “Super El Nino” which would cause 2007 to be the “hottest year ever.”
The last six years (2003-2008) show a steep temperature drop in the satellite record, which is not present in the GISS data. Prior to 2003, the three trends were all close enough to be considered reasonably consistent, but over the last six years is when a large divergence has become very apparent both visually and mathematically.
Since the beginning of 2003, RSS has been dropping at 3.60C/century, UAH has been dropping at 2.84C/century, and GISS has been dropping at 0.96C/century. All calculations are done in a Google Spreadsheet here:
The divergence between GISS and RSS is shown below. Since the start of 2003, GISS has been diverging from RSS at 2.64C/century, and GISS has been diverging from UAH at 1.87C/century. RSS has been diverging from UAH at minus 0.76C/century, indicating that RSS temperatures have been falling a little faster than UAH over the last six years, as can also be seen in the graph above.
Below is a 250km map of GISS trends from 2003-2008. One thing which stands out is that GISS has large areas with sparse or no coverage. Notably in Africa, Antarctica, Greenland, Canada, Brazil, and a few other places.
Click for larger image
Many of the GISS holes seem to be in blue regions on the map. Here is a post and video of the GHCN station loss over the past several years globally, created by WUWT contributor John Goetz:
Here are two images showing the difference between GISS global coverage in 1978 and 2008:
Click for a larger image
Click for a larger image
There is a tremendous amount of station dropout in 30 years. Dropout is worst in the high northern latitudes, most all of Canada, and about half of Africa. Of particular note is the red band at the southernmost latitude, which “seems” to indicate a continuous coverage there. Of course we know that is not true, given the paucity of stations in the Antarctic interior. Read more here.
By contrast, while it doesn’t hit both poles (neither does GISS) UAH has much broader global coverage as seen below. Could this be part of the explanation for the divergence between GISS and satellite data? What do the readers think?
![[Image]](https://i0.wp.com/discover.itsc.uah.edu/amsutemps/browse/AMSU_A_15.latest.a_04.png?resize=520%2C278&quality=75)
Click for larger image
![[Image]](https://i0.wp.com/discover.itsc.uah.edu/amsutemps/browse/AMSU_A_15.latest.d_04.png?quality=75)
Click for larger image
How different would the GISS graph appear, if it showed a -3.6C/century cooling trend over the last six years? For reference, the steep GISS warming trend from 1980 to 2002 was about 0.4 degrees.
John Finn (08:48:07) :
” Jan 2009 looks to be comfortably above average – and I stick by a prediction I made on another blog that 2009 will be warmer than 2008.”
What planet are you on? Comfortably above average? Do you mean all time coldest in Illinois and Maine? You have a very strange idea of comfortable. Seriously though you have about a 50% chance of being right but even though it is the warmest here it has been in 3 weeks I am not going to celebrate. (-15C)
I am pretty certain that November and December were the coldest in at least a generation around here so it could be warmer than 2008 and still darn cold. Growing season in 2008 was about 3 weeks shorter than typical, a bit more cooling on that stat and we could all get mighty skinny.
I keep hoping for some global warming to get here, I really do, maybe I should surrender and move to Yuma?
Jeff Alberts,
I agree but to free the prisoner, and help him escape his chains, sometimes you have to start by finding a way over the barbed wire fence.
Ya know what I mean? 😉
Andrew ♫
Wondering aloud: he was talking about the global temperature because, believe it or not, Maine and Illinois do not form the entire world. Not even the US!
Go check satellite temperatures at UAH website: http://discover.itsc.uah.edu/amsutemps/amsutemps.html
The closest-to-surface temperature are at 1 km.
“We already know from the data we have, however poor it may be, that GW isn’t global by any stretch of the imagination.”
I’ll that a step further and say we don’t know exactly what we know, because the data has been adjusted.
Andrew ♫
Taking the change from the 1970’s to the 1990’s and extrapolating a 100 year trend isn’t really much different than taking the change from 6am to noon at my house and extrapolating that at the current rate of temperature change, the Earth will be hot enough to melt steel in 100 years. The only thing different is the duration of the cycles.
Not taking into account natural temperature cycles having nothing to do with human activity is about as silly as not taking into account the natural diurnal cycle.
“If you only have one station, how would you possibly know if anything global is happening? Sheesh!”
Because if it is “global” then it is “global” and if it is happening at one place it is happening at them all. In other words, if I have data over, say, 200 or 300 years, that is probably enough to show if there is a trend either positive or negative. Taking data for less than 100 years is likely useless. You need at least one cycle of data just to establish some frame of reference. That takes somewhere along the lines of 50 to 70 years. You then need to measure over a second cycle to see if it has changed from the first. So now we are at somewhere between 100 and 140 years. And then you need a third cycle to see of there is really any trend … to see if the cycles are topping out at higher or bottoming out at lower values over time … and so here were are at about 1.5 to 2 centuries before any kind of real “signal” can be resolved and a real trend line would take, say, 4 or 5 such cycles to resolve.
In other words, attempting to show “man made global warming” during a period when we were in a period of a natural cycle that has rising temperatures anyway is actually quite silly.
Anthony,
You asked, “How different would the GISS graph appear, if it showed a -3.6C/century cooling trend over the last six years?”
It would look exactly like this:
http://i81.photobucket.com/albums/j237/hausfath/Picture12.png
REPLY: Actually that was Steve Goddard’s question but thank you for answering it. – Anthony
Steve Goddard:
As a follow up, I remembered a post I had done in May about the GISTEMP 1200km versus 250km smoothing.
http://bobtisdale.blogspot.com/2008/05/giss-temp-smoothing-radius-comparison.html
I had already endured the trouble of retrieving the GISTEMP data with 250 smoothing and wanted to see if it would add anything to the discussion. I went back and cleaned up the delta T (1200km smoothing minus 250km smoothing) graph and took out the poly trend line so not to have to endure the wrath of others who comment here:
http://i39.tinypic.com/120qkpg.jpg
Then I shortened the time span to January 2003 to May 2008. Aha, there’s a positive trend in the 5+ years of data.
http://i42.tinypic.com/5pldon.jpg
So I went back and tried to update the graph by tacking on the data since May, without looking for any updates there may have been to the earlier data. Here’s the delta-T graph with the update:
http://i42.tinypic.com/mcexe9.jpg
And the short-term version:
http://i41.tinypic.com/2qixs75.jpg
That cluster in recent months concerns me, as does the shift in the trend line, so I’m not sure if that really helped.
It is no answer to my challenge to point to periodic variations in the recent rate of global temperature increase. Unless these are explicable (volcanoes, aerosols, ice melt, redistribution of heat, solar variation) in concrete physical terms they are simply inconsistent with CO2 forcing and current climate models.
Five years is ample time to confirm these measurements. Even one year should suffice if the data is as accurate and comprehensive as claimed.
The evidence is clear that natural variability is capable of outweighing CO2 forcing. Since we don’t understand why we cannot predict the future.
For Jim Norvell:
Last year I analyzed the temp histories (from GISS) of only those stations in existence before 1940. This date was the mid-century peak in temp, as well as the start of CO2 increase. I identified over 400 stations (excl US) whose net increase from 1940 to 2008 was but 0.2 deg C on avg (identical to US graph). This amount was considerably less than the GISS Global Mean (+0.6 deg C). The GISS analysis was corrupted (in my opinion) by the addition of thousands of stations after 1950.
Sam
Nobody sensible should expect that the 4 major global mean temperature series, 2 satellite, 2 surface based, will ever show 100% correlation. Calculating an estimate of the lower tropospheric temperature from the MSU readings is a complex task and the two agencies that publish such an estimate, while they use the same source data, apply different assumptions about orbital decay, satellite calibration etc. Similarly the two major surface based estimates also share many readings of sea temperatures and land stations in common but make different homogeneity adjustments and notably, GISS extrapolates over the Arctic region from the nearest stations, while HADCRUT does not. The surface and satellite estimates will always differ because they are measuring a different quantity, the near surface temperature and the lower tropospheric temperature respectively, and while the coverage of the satellites is good, it does not extend to the extreme high latitudes. Given these sources of divergence, the actual convergence is, arguably, renmarkable
Getting onto trends and back on topic, it seems that the satellite indices respond more strongly than the surface based estimates to ENSO events, warming more than the surface based series during an El Nino and cooling more in a La Nina. For example, here is RSS warming more than GISS during the 1998 Super El Nino and here cooling more during the 1985 La Nina.
To quantify this source of divergence I plotted the ENSO Multivariate index against the difference between RSS and GISS and UAH and GISS respectively. By eyeball the chart shows a striking correlation in some periods, but the match breaks down in others, e.g. the 1998 El Nino gave a large divergence while the 1983 peak did not. Calculating the correlation coefficient for UAH-GISS from 1979 onwards (using all the data, that is) gives an unimpressive result of +0.08 using the raw numbers, however if I time shift the temperature response to 6 months later than the ENSO data, this increases the correlation to a more significant +0.32 , reinforcing the visual impression of a lag between the two. Here is the timeshifted data:
http://spreadsheets.google.com/ccc?key=pFFm87cBy5_qaPiWsXRWCEg#
For Mr Goddard’s carefully selected period of 2003 onwards the correlation coefficient is +0.73 for UAH-GISS and +0.83 for RSS-GISS, however I do not assign huge significance to a dataset of just 72 points. However the proposal that the ENSO factor in the divergence is null is not supported by the data, clearly.
My little spreadsheet is no substitute for a rigourous analysis of course, just me scratching an itch, I suspected the ENSO had a large role in the satellite/surface differences, and I have now quantified it, to my own satisfaction at least. As to why the ENSO should affect the satellite estimates more than the surface station results, two speculations occur… it could be that the troposphere reacts more strongly than the surface to ENSO, possibly driven by changes in evaporation rates in the tropics, or it could be as simple as the satellite analyses overweighting the tropics, compared to Hadley and GISS.
Who knows? None of this sheds any light on the real puzzle: which is why the trend in UAH diverges so much from the other three major series over the satellIte era.
“Five years is ample time to confirm these measurements. ”
It absolutely is not. We had 30 years of cooling going in to the 1970’s. We had thirty years after that. Five years of data would have shown absolutely nothing about what was going on with “global temperature”.
As the most recent peak in temperatures did not reach the level of the 1930’s peak, it appears that we are in an overall period of cooling shorter term variations of warming and cooling but overall, temperatures appear to be cooling. And the problem with even the “modern” record is that the early part of the record simply shows recovery from the Little Ice Age.
The fastest rate of rise in the temperature record is during the 1700’s. There was no significant CO2 impact from humans at that time. What Hansen has documented is natural cyclic warming from a positive PDO cycle and extrapolated that over time as if it would continue forever.
You can absolutely not gain any significant information over a 5 year period in the Northern Hemisphere except to learn what the current natural cycles are doing. Anyone who claims otherwise is practicing voodoo, not science.
Can you look at the rate of this morning’s temperature change and use that information to predict what tomorrow’s temperature will be? No. Can you look at the rate of change in daily high temperatures in April and predict based on just that information what October’s high temperatures will be? No. But if you have a history of the entire annual cycle of temperature over an entire year, you can get an idea of what the range of summer is likely to be. And it is the same with cycles that are longer than a year.
We are talking about cycles of natural variation that are several decades long. You can not take readings over a portion of it and extrapolate from that what the temperature will be a century hence.
Said: We had 30 years of cooling going in to the 1970’s. We had thirty years after that.
Meant: We had 30 years of cooling going in to the 1970’s. We had thirty years of warming after that.
I would like to sincerely thank everyone involved for a fascinating blog that is doing exactly what scientists should be doing with this and other issues – examining it from all sides until no doubt remains.
My interest in this far exceeds my knowledge – as I count myself more a philosopher than physical philosopher.
As to those who rudely interupt this discussion with comments designed to stifle debate – I do wish they would go back and re-read Descartes. As the father of modern scientific method [my assertion – debate at your peril 🙂 ] his methods have been followed in outline from the time he proposed them. We abandon them now at our peril.
After all – you can’t do any science without using his co-ordinate system.
Congratulations people – Descartes would be proud.
And sorry for being slightly off thread.
Peter Jones (20:36:02) : “from Yahoo story: Americans giving Obama extraordinary support: polls ‘A survey conducted by The New York Times and CBS News found a US public eager to give the president-elect a wide berth…'”
LMAO! Give him a wide berth!!! I tried, guys, I tried!
John Philip:
Looks like the two of us did the exact same analysis independently. You can see my spreadsheets here if you want to play around with them: http://drop.io/hausfath/asset/temp-data-12-2008-xls
I’m just curious what the mechanism could be for the 6 month lag, as I’m somewhat loathe to arbitrarily shift the data simply because I eyeball a better fit.
If my boss asked me to program a report(as part of my job) to take the issue seriously wether AGW is true or not, I would have to know what the claim is.
If one assumes the face value (in “science” you aren’t supposed to assume, so you have to ask what the claim actually is EVERY TIME, which my experience with people tells me that assumption is already fraught with problems, -my boss and I always fight about what the assumptions are, but for conversation’s sake, let’s say) one has to show the A, the G, and the W for AGW to be true.
Let’s start with the W, because the A and the G are types of the W. We can break out the A and the G later (personal preference?). Which one should I go after first? Is that a yes or no question?
Is it warming period? Yes or no? Can the question be answered with a yes or a no? It depends on what the ‘it’ is. Which location should I use as my first ‘it’? Where do I start? Can that question be answered with a yes or a no? I suppose it depends on your preference again? Bad question.
If ‘it’ is the globe, then the question I should actually start with is ‘Is there Global Warming’? Surely I can answer that with Yes or no? It depends on if ‘it’ (the globe) is warmer now than any time in the past (I can’t assume anything), if I can say yes or no.
Is there a temperature record for the entire globe? Yes or no? Well, what would be ‘good enough’ to measure warming for the entire globe? Is that a yes or no question? Which personal preference do I use again?
As you can see, I am stuck with a question ‘Is there a way to measure the temperature of the entire globe right now’ with which the answer is a personal preference. Yes or no? Is there a way that can be accepted as the measure of the temperature of the entire globe right now that someone knows? Yes or no?
I would say to my boss, ‘I need more information to trouble-shoot’. Here’s a question that ALWAYS gets results- What does the client with the problem actually WANT?
😉
Andrew ♫
Crosspatch, you entirely miss my point. Five years is quite enough to determine the measurements – ie the temperature is not increasing over that period. So the question immediately is why?
Yes, this is natural variation. But it must be explained. Exactly where is the heat going? If it correlates with PDO, so what? Where is the heat going or being stored?
There is not that much difference in how the different temp series respond to the ENSO. GISS had an unusual response to the 1997-98 El Nino but over the whole GISS record back to 1880, it is not that much different.
Here are the regression coefficients on the ENSO for the major temp series. (Note ENSO region anomalies can vary up to +/- 3.0C which is a big number compared to the global temp record.)
UAH – 0.063
RSS – 0.059
Hadcrut3 – 0.076
GISS – 0.053
And the lag for the ENSO should be 3 months (not 6 months).
What explains the 3 month lag is the lag between ocean temperatures and surface temperatures. Basically, the ocean lags the surface by about 50 to 80 days. Land temperatures peak about 30 days after the summer solstice (July 21st and the coldest part of the year is 30 days after the winter solstice).
Ocean temps, however, peak about 80 days after the summer solstice (Sept 12th) – The peak polar ice melt is Sept 12th – the peak Hurricane season is Sept 12th – the actual ocean temps peak on Sept 12th. And the opposite timeline occurs in the winter.
When the ocean gives back temperatures to the atmosphere, there is also a 50 to 80 day lag until all the ocean heat/cold in the ENSO region is dumped back into the atmosphere (globally of course, Bob Tisdale has shown the ENSO has a long-term lasting impact in the Western Pacific.)
Bill Illis: you got ‘er nailed. And it’s all about heat STORAGE on this planet, not magical radiation cartoons. The radiation is a “symptom” or “measurement” of what is happening.
“Yes, this is natural variation. But it must be explained. Exactly where is the heat going? If it correlates with PDO, so what? Where is the heat going or being stored?”
The only place heat is stored is in the ocean as atmospheric heat is eventually radiated into space. What changes is where the warm and cold spots are, not really the total amount of heat. When we have changes in wind currents, we have changes in the surface temperatures at various places. When the surface temperatures change, the air temperature changes.
This is why I would measure ocean temperature at the sea bottom, that would give a better indication, I think, of the heat content of the ocean. If the water near the bottom of the ocean warms a half-degree, we have a large increase in the amount of heat in the ocean. If the surface warms half a degree, it might mean that the trade winds have simply slowed down a little.
Changes in wind speed and direction can change surface temperatures and cause upwellings to change. Some of these weather patterns are cyclical in long term oscillations. You can just about correlate the sardine catch off of California with changes in the PDO, for example.
It isn’t so much that the total heat is changing, it is that the locations of the warm and cool spots are changing. This influences weather patterns and can be self-reinforcing. But it could well be that heat builds up until the system switches state, then this heat is released into the atmosphere where it is dissipated into space. Once enough heat has been lost, the system “flips” to the other state and begins to absorbing heat again.
But the only place that “stores” heat is going to be the ocean and the only place it is going to be “lost” is to the atmosphere and into space.
“Nobody sensible should expect that the 4 major global mean temperature series, 2 satellite, 2 surface based, will ever show 100% correlation.”
I agree with that. But by the same token, they should show the same basic general direction. They might vary in absolute value but one wouldn’t expect to see a situation where the satellite and two ground-based networks (HADCrut and GISS) show cooling and a third (NOAA) shows warming over the same period. It reminds me of an old Sesame Street song that went something like “one of these things doesn’t belong”.
Once the CRN gets up and going, it will put a lot of these issues to bed because there will no longer be any need for “adjustments”. This is a network designed from the start to be a climate monitoring network rather than a weather network pressed into service for climate monitoring. It is going to take a considerable amount of time, though, to collect enough data from these stations to get any idea of long term trends. Besides that, the network covers only the US, we still have considerable problems with stations in the rest of the world (ROW) as those tend to have the least assurance in quality of equipment and observations.
@EMH (12:38:29) :
What you wrote, I am sincerely with you 100% until you get to here.
“Reason must prevail, so that humanities’ resources can used to make wise decisions about environmental challenges from weather and natural, sometimes catastrophic, earth events, WHILE cleaning up poisons going into our air, land, and water, addressing poverty, health, starvation, wars, conserving all resources, and doing our best to live in harmony with all life on the planet. We need human and spiritual wisdom, and we desperately need respect for one another!”
To that, all I can say is “kumbayah.” I’m a bit cynical about the current global political climate. I share your wish but I believe herding 10,000 cats is easier than what you’re hoping for.
Crosspatch, the problem is that the ocean temperature doesn’t show that heat is being stored there over the last five years. Both temperature and sea level rises stalled. The onus is on the AGW proponents to show where the heat is being stored.
The evidence so far as far as I can see is that it is not being stored anywhere on earth. Therefore their forcing assumptions are false.
“ocean analysis chosen as HadI/Reyn_v2, you will see that the ocean coverage is in fact quite complete”
Don’t Reynolds temperatures include satellite info, thus should be excluded for comparison with satellite temperatures?