There’s some really interesting things going on with global temperature. On one hand we have UAH and RSS which show Global Temperature anomalies near zero, while NCDC/NOAA and GISS (which derives from NCDC data with their own adjustments added) show large positive anomalies.
Joe D’Aleo at ICECAP writes:
Last month, NOAA had May 2009 to be the 4th warmest on record globally. Meanwhile NASA UAH MSU satellite assessment showed it was the 15th coldest May in the 31 years of its record. This divergence is not new and has been growing. Just a year ago, NOAA proclaimed June 2008 to be the 8th warmest for the globe in 129 years of record keeping. Meanwhile NASA satellites showed it was the 9th coldest June in the 30 years of its record.
Of course the obvious question “who’s right” will be the subject of many posts to come, but I wanted to get this out there for discussion. There’s some interesting things going on with the NCDC data.
The combined global land and ocean surface temperature was the second warmest on record in June, behind 2005, and tied with 2004 as the fifth warmest on record for the year-to-date (January-June) period. The global ocean had the warmest June on record. The ranks found in the tables below are based on records that began in 1880.
What is truly interesting about June (besides the wide discrepancy between global data sets) is the time period with which the onset of the warming occurred. Some say it has to do with El Nino developing in the Pacific. Perhaps, but the El Nino conditions we see now are not comparable to what we saw in 1998, yet we have global temperatures being reported that are comparable.
It is an interesting mystery, and it will be interesting to see how it plays out and what is discovered. Stay tuned for more on this topic.
Addendum: I should point out that there is a lag between surface and lower troposphere, so we’ll see what July says as LT is already shaping up a bit warmer at UAH. – Anthony
Addendum: I should point out that there is a lag between surface and lower troposphere, ….. – Anthony
Question: If there is warming due to CO2 does it follow that the surface should warm before the troposphere?
Flanagan (23:41:10) :
………………………….
Also: there’s no need in going into conspiracy theories and the like. I mean: what are you going to say next month when the satellite anomalies will break records as well?
…………………………
Can I propose a guest post next month? Serious, I for one appreciate a warmer who is at least trying to put forward a measurable science based argument.
Ian B
July 17, 2009 5:44 am
With regard to RBateman’s early comment, June was warm and dry in London, particularlty the later part of the month.
The effect was obviously anthropogenic – the new roof on Wimbledon centre court precluded rain for two weeks.
We had a couple of days above 30 deg C (first in about 3 years) and several in the upper 20s
July on the other hand has been ordinary to a bit poor so far – high teens to low 20s, quite cloudy and showery.
All in all, close to a typical British summer.
Where I live the ‘normal’ high is 82 right about now. We’ve had a few days above that but today it’s supposed to be 70 and continue in the 70’s for days. It’s been a chilly summer. BUT like normal NOAA and GISS will tell us that we had a above normal July or a barely below normal July. Yet when we actually do barely have above normal they tell us we were drastically above normal. I think they need to work on their fudging of the numbers to become more accurate.
Pofarmer
July 17, 2009 6:20 am
I was farming in 1998. This ain’t no 1998.
Jim
July 17, 2009 6:38 am
Does anyone know where a map of the temperature sensors used for the global measurements is? I have searched the web and can’t find it. I guessing we are seeing the same kind of “smearing” of high temperatures from some stations just as in Steig’s Antarctic trend analysis. I wonder what Jeff ID and Ryan O could make of the data. On that note, is the raw data even available?
Jim
July 17, 2009 6:42 am
AJStrata (05:16:00) : One approach would be to just average together the actual readings and not try to “impute” the missing data. It’s not perfect but I bet it’s a better approach than whatever statistical fairy dust NASA and NOAA use.
Pofarmer
July 17, 2009 6:43 am
I agree the century long record has big error bars around it, but more recent measurements do seem to be in the ballpark. The ~0.3C rise in SST between 1993 and 2003 is consistent with a rise of around 14×10^22J in ocean heat content, as determined for the sea level rise due to thermal expansion measured by satellite altimetry. SST’s have risen recently because the ocean is emitting the heat it gained and stored during the run of high amplitude-short minimum cycles. Now the sun is quiet, the ocean is emitting heat. This is confirmed by the outgoing longwave radiation record which shows OLR increased by 4W/m^2 from June 2000 to 2006. It has stayed at that higher level since because of the protracted solar minimum.
So SST’s are rising, but the ocean is cooling. The emission of heat will warm the atmosphere over the next months, until the air temperature suppresses the ocean’s emittance. There will be lag time.
If true, that is an elegant mechanism.
JP
July 17, 2009 7:04 am
“And indeed, observing record positive anomalies with (the beginning of) an average El-Nino event IS a sign a warming took place. I already said that when we had the 7th or 8th warmest years during long and deep la ninas, but nobody was listening. I just hope the Nino won’t be too strong this time.”
Based on what, Flanagan? NASA? GISS now has June 2007 as being the warmest or second warmest month on record. For some reason 1998 cooled, according to thier datasets. But NASA get its surface data from primairily from NOAA, which in turn has a database heavily weighted to Norh American stations. And the majority of those stations are from aerodromes, which are listed as rural (ie little to no UHI adjustment). To make things worse (or more complicated) 75% of the global weather reporting stations have closed since 1960. The majority of those stations were rural. What NOAA and GISS are doing is comparing apples and oranges. Comparing the years 1934 to 2008 when in 2008 there were less reporting stations, and more urban enviorments (ie huge concrete tarmacs) will result in a very predictive trend (marked warming).
As others have noted, some of the regions (or gridcells) that have the highest positive anomalies (East Asia, Siberia, Artic) do not have a single reporting station. They may have many decades ago, so all we can surmise is that the folks at NASA and NOAA are splicing thier own extrapolated trends on to old data. This isn’t science.
Mark
July 17, 2009 7:08 am
Timing is everything.
This warming is coming at the right time for the AGW crowd… A few months before cap-and-trade is to be voted on in the Senate followed by Copenhagen.
I read somewhere that a massive six thousand page climate report is coming out in about a month.
urederra
July 17, 2009 7:10 am
Well, I know that weather is not climate and all but I am watching Le Tour de France right now, live on TV, they are climbing the Alps, and IT IS SNOWING!!!!
Well, either that or fine hail, it does like something in between snow and hail, actually.
Go, Contador, go!!!
Ron de Haan
July 17, 2009 7:12 am
AJStrata (21:06:22) :
The truth? The truth is the alarmists have been smearing bad ground based data over more precise and temporally linked satellite data to obscure the cooling. It’s easy to do statistically as I show in this post: http://strata-sphere.com/blog/index.php/archives/9852
The point is a satellite operating for 3-6 years uses a single, calibrated and verifiable sensor to cover the globe making thousands of measurement points. A single sensor which can be checked against ground sensors which only measure the temperature for a distance of a few feet at best.
When you compare this measurement system in space to the compounding errors from thousands of uncalibrated, unverified, lower precision sensors spread thinly across the Earth’s surface it is clear what is going on. Ground based sensors cannot produce a global measurement with any accuracy under a degree if the error budget was being done correctly. Each sensor has an unmeasured error, which combines to create a huge error in ‘global temperature’.
A satellite sensor has a fixed error source that can be calibrated (self checked) many ways. There is only one error source, not thousands.
I mean, does anyone still do basic math anymore?
AJStrata,
Thanks for posting this interesting theory.
The big question is, “HOW DID THEY DO IT?”
Only if this question is answered do we get somewhere.
And better let us do it quickly, preferably before September 8th.
I was just in Jacksonville Florida, touring a 1700’s sugar cane plantation. We were told there and in St. Augustine that the climate there was now too cold for sugar cane and oranges, so those crops have moved to southern Florida.
If the world’s climate is “warmer than it’s ever been”, well, WUWT?
It’s friggen 59 F at 9:40 CDT in Minneapolis.
Almost had frost up in the “Arrowhead” last night.
I’ve been predicting “hammer hard winter” for about 9 months.
We’ll see. Had to turn on HEAT this morning to bring house from 68 F
to 72 F, barely comfortable.
Nogw
July 17, 2009 7:42 am
Have the satellite date been adjusted with respect to ground stations with the errors Anthony has shown us these have?
Could it be possible to adjust them back according to surfacestations.org and see what truth is?
Flanagan
July 17, 2009 7:45 am
Well Nick, I can already say that during the early stage of an El Nino, the Walker convective cell is weakened which slows down the mixing of the different atmospheric layers.
But there is also the seasonal problem of UAH, which shows oscillations around the anomalies, with a minimum in May/June. Strange, in my sense.
Flanagan
July 17, 2009 7:49 am
uruderra: they are in the Vosges, not in the Alps and it’s raining, not snowing.
Nogw
July 17, 2009 8:09 am
From the link you give above: If El Niño conditions continue to mature as projected by NOAA, global temperatures are likely to continue to threaten previous record highs.
Have they PROJECTED global temperatures?….
Jared
July 17, 2009 8:16 am
GISS is a phony operation. Starting with the release of their December 2008 data I’ve been tracking what they’ve been doing with the data between January 1998 and November 2008.
Well here’s what they are doing.
0.0105 increase in temp per year with the December 2008 release for Jan’98-Nov’08
0.0107 increase in temp per year with their April 2009 release for Jan’98-Nov’08
0.0109 increase in temp per year with their July 2009 release for Jan’98-Nov’08
^^^^ That’s all from the same January 1998 thru November 2008 time frame. I’m not including the new data afterwards. It’s the same 131 months. Yet somehow GISS has managed to make the increase in global warming in that time frame grow and grow with each months new release. Simply amazing.
Ron de Haan
July 17, 2009 8:17 am
I know, it’s just weather:
Remarkably Cool, we are talking about frosts and freezes occurring in July!!!
Updated: Friday, July 17, 2009 7:32 AM
Temperatures currently in the nation’s heartland are more typical of April or October than July across many locations. Highs will struggle to reach 60 Friday across the Upper Peninsula of Michigan and the Lake Superior region of Minnesota. The high in Chicago will briefly touch 70, Detroit will struggle to get past 70, and Minneapolis will never reach 70. Many places will see highs a good 10 to 18 degrees below normal.
What is going on? Well, a northerly breeze, cloud cover, showers and the jet stream being unusually far south for this time of year will all play a varying role in the chilly air. Look for another cool day in the same locations Saturday after a very chilly night overnight. Some areas in northern lower Michigan have seen temperatures drop to near freezing (We are talking about frosts and freezes occurring in July!!!) recently, and while temperatures will not drop that low thanks to a breeze and some cloud cover, it will still be chilly Saturday night.
Story by AccuWeather.com Meteorologist Mark Paquette. http://www.accuweather.com/news-story.asp?partner=rss&article=4
rickM
July 17, 2009 8:43 am
I think what bothers me is the base period that 2009 was measure against – 1961 to 1990, which would have been “cold” biased, and would make this year seem much warmer. Wy is there no standard? Why is just a 30 year average used vice a much longer term time frame that actaully may show a cyclical pattern?
Advocacy in science is very corrupt and I do accuse them of advocacy
Paul (woodfortrees):
Regarding the possible lag between surface and satellite temp response to El Nino, we were having an interesting discussion over at Lucia’s place: http://rankexploits.com/musings/2009/giss-updates-increase-recent-historic-trends-slightly/
There appears to be some lag between the two, but its hard to tease out because ENSOs vary so much in magnitude and temperature is so noisy. There definitely was a faster surface response than satellite response at the start of the last major El Nino back in ’97.
If you want to remove the ground station mess from the satellite data you begin with just the satellite data itself. The challenge there is how to handle the change over from one satellite to another, and the aging (drift) if the sensor.
My proposal would be to identify 3-6 well documented ground stations around the globe which have good performance and span all 31 years of the satellites. These are the statistical anchor points for the satellite drift and transitions. If, on average, these 6 stations indicate a life time change in Sat1’s sensor from +.5°C to +1.1°C we can adjust for sensor aging. If the same 6 stations detect Sat2’s bias dropping to +.3°C when it took over for Sat1 we can not only adjust for drift but transition.
The key is to only use a small number of ground reference points so you don’t add mountains of sensor errors from the ground.
Once you compute a satellite only plot for the last 31 years, then you can not only compare to NOAA and GISS, you can estimate their errors and biases – even regionally. You can show how each surface sensor performed against the satellites (one Sat or across satellite)
Turn the tables, use the satellites to question the ground sensor errors and calibration. It is, in fact, the proper method scientifically and statistically. Single spacecraft sensor with known and measured performance is the only way to check the performance of thousands of independent sensors with varying precision and health.
When you do this, the NOAA-GISS anomalies are no longer global warming, but data showing how bad the ground sensors are performing! Which is the how this should have been computed in the first place.
Addendum: I should point out that there is a lag between surface and lower troposphere, ….. – Anthony
Question: If there is warming due to CO2 does it follow that the surface should warm before the troposphere?
On the other hand, looking at the frequency (actually, harmonic number) spectrum:
http://www.woodfortrees.org/plot/uah/from:1979/to:2009/window/fourier/magnitude/from:10/to:60/plot/gistemp/from:1979/to:2009/window/fourier/magnitude/from:10/to:60/plot/rss/from:1979/to:2009/window/fourier/magnitude/from:10/to:60/plot/hadcrut3vgl/from:1979/to:2009/window/fourier/magnitude/from:10/to:60
GISTEMP seems to have the most annual (harmonic 30) energy, although it’s pretty small.
OK, I’m confused!
Flanagan (23:41:10) :
………………………….
Also: there’s no need in going into conspiracy theories and the like. I mean: what are you going to say next month when the satellite anomalies will break records as well?
…………………………
Can I propose a guest post next month? Serious, I for one appreciate a warmer who is at least trying to put forward a measurable science based argument.
With regard to RBateman’s early comment, June was warm and dry in London, particularlty the later part of the month.
The effect was obviously anthropogenic – the new roof on Wimbledon centre court precluded rain for two weeks.
We had a couple of days above 30 deg C (first in about 3 years) and several in the upper 20s
July on the other hand has been ordinary to a bit poor so far – high teens to low 20s, quite cloudy and showery.
All in all, close to a typical British summer.
Ah, hang on… Looking at the last decade only, which is the period “Lamont” is claiming an increase for, UAH has markedly more energy at 12 months (harmonic 10).
http://www.woodfortrees.org/plot/uah/from:1999/to:2009/window/fourier/magnitude/from:2/to:20/plot/gistemp/from:1999/to:2009/window/fourier/magnitude/from:2/to:20/plot/rss/from:1999/to:2009/window/fourier/magnitude/from:2/to:20/plot/hadcrut3vgl/from:1999/to:2009/window/fourier/magnitude/from:2/to:20
Note there’s also that interesting roughly bi-annual cycle in there as well, which you can see if you smooth out the inter-annual changes:
http://www.woodfortrees.org/plot/uah/from:1999/to:2009/mean:12/plot/gistemp/from:1999/to:2009/mean:12/plot/rss/from:1999/to:2009/mean:12/plot/hadcrut3vgl/from:1999/to:2009/mean:12
I’ve often wondered about that, but I’ve no idea where it might come from (it only seems to be recent)
Where I live the ‘normal’ high is 82 right about now. We’ve had a few days above that but today it’s supposed to be 70 and continue in the 70’s for days. It’s been a chilly summer. BUT like normal NOAA and GISS will tell us that we had a above normal July or a barely below normal July. Yet when we actually do barely have above normal they tell us we were drastically above normal. I think they need to work on their fudging of the numbers to become more accurate.
I was farming in 1998. This ain’t no 1998.
Does anyone know where a map of the temperature sensors used for the global measurements is? I have searched the web and can’t find it. I guessing we are seeing the same kind of “smearing” of high temperatures from some stations just as in Steig’s Antarctic trend analysis. I wonder what Jeff ID and Ryan O could make of the data. On that note, is the raw data even available?
AJStrata (05:16:00) : One approach would be to just average together the actual readings and not try to “impute” the missing data. It’s not perfect but I bet it’s a better approach than whatever statistical fairy dust NASA and NOAA use.
I agree the century long record has big error bars around it, but more recent measurements do seem to be in the ballpark. The ~0.3C rise in SST between 1993 and 2003 is consistent with a rise of around 14×10^22J in ocean heat content, as determined for the sea level rise due to thermal expansion measured by satellite altimetry.
SST’s have risen recently because the ocean is emitting the heat it gained and stored during the run of high amplitude-short minimum cycles. Now the sun is quiet, the ocean is emitting heat. This is confirmed by the outgoing longwave radiation record which shows OLR increased by 4W/m^2 from June 2000 to 2006. It has stayed at that higher level since because of the protracted solar minimum.
So SST’s are rising, but the ocean is cooling. The emission of heat will warm the atmosphere over the next months, until the air temperature suppresses the ocean’s emittance. There will be lag time.
If true, that is an elegant mechanism.
“And indeed, observing record positive anomalies with (the beginning of) an average El-Nino event IS a sign a warming took place. I already said that when we had the 7th or 8th warmest years during long and deep la ninas, but nobody was listening. I just hope the Nino won’t be too strong this time.”
Based on what, Flanagan? NASA? GISS now has June 2007 as being the warmest or second warmest month on record. For some reason 1998 cooled, according to thier datasets. But NASA get its surface data from primairily from NOAA, which in turn has a database heavily weighted to Norh American stations. And the majority of those stations are from aerodromes, which are listed as rural (ie little to no UHI adjustment). To make things worse (or more complicated) 75% of the global weather reporting stations have closed since 1960. The majority of those stations were rural. What NOAA and GISS are doing is comparing apples and oranges. Comparing the years 1934 to 2008 when in 2008 there were less reporting stations, and more urban enviorments (ie huge concrete tarmacs) will result in a very predictive trend (marked warming).
As others have noted, some of the regions (or gridcells) that have the highest positive anomalies (East Asia, Siberia, Artic) do not have a single reporting station. They may have many decades ago, so all we can surmise is that the folks at NASA and NOAA are splicing thier own extrapolated trends on to old data. This isn’t science.
Timing is everything.
This warming is coming at the right time for the AGW crowd… A few months before cap-and-trade is to be voted on in the Senate followed by Copenhagen.
I read somewhere that a massive six thousand page climate report is coming out in about a month.
Well, I know that weather is not climate and all but I am watching Le Tour de France right now, live on TV, they are climbing the Alps, and IT IS SNOWING!!!!
Well, either that or fine hail, it does like something in between snow and hail, actually.
Go, Contador, go!!!
AJStrata (21:06:22) :
The truth? The truth is the alarmists have been smearing bad ground based data over more precise and temporally linked satellite data to obscure the cooling. It’s easy to do statistically as I show in this post:
http://strata-sphere.com/blog/index.php/archives/9852
The point is a satellite operating for 3-6 years uses a single, calibrated and verifiable sensor to cover the globe making thousands of measurement points. A single sensor which can be checked against ground sensors which only measure the temperature for a distance of a few feet at best.
When you compare this measurement system in space to the compounding errors from thousands of uncalibrated, unverified, lower precision sensors spread thinly across the Earth’s surface it is clear what is going on. Ground based sensors cannot produce a global measurement with any accuracy under a degree if the error budget was being done correctly. Each sensor has an unmeasured error, which combines to create a huge error in ‘global temperature’.
A satellite sensor has a fixed error source that can be calibrated (self checked) many ways. There is only one error source, not thousands.
I mean, does anyone still do basic math anymore?
AJStrata,
Thanks for posting this interesting theory.
The big question is, “HOW DID THEY DO IT?”
Only if this question is answered do we get somewhere.
And better let us do it quickly, preferably before September 8th.
I was just in Jacksonville Florida, touring a 1700’s sugar cane plantation. We were told there and in St. Augustine that the climate there was now too cold for sugar cane and oranges, so those crops have moved to southern Florida.
If the world’s climate is “warmer than it’s ever been”, well, WUWT?
It’s friggen 59 F at 9:40 CDT in Minneapolis.
Almost had frost up in the “Arrowhead” last night.
I’ve been predicting “hammer hard winter” for about 9 months.
We’ll see. Had to turn on HEAT this morning to bring house from 68 F
to 72 F, barely comfortable.
Have the satellite date been adjusted with respect to ground stations with the errors Anthony has shown us these have?
Could it be possible to adjust them back according to surfacestations.org and see what truth is?
Well Nick, I can already say that during the early stage of an El Nino, the Walker convective cell is weakened which slows down the mixing of the different atmospheric layers.
But there is also the seasonal problem of UAH, which shows oscillations around the anomalies, with a minimum in May/June. Strange, in my sense.
uruderra: they are in the Vosges, not in the Alps and it’s raining, not snowing.
From the link you give above:
If El Niño conditions continue to mature as projected by NOAA, global temperatures are likely to continue to threaten previous record highs.
Have they PROJECTED global temperatures?….
GISS is a phony operation. Starting with the release of their December 2008 data I’ve been tracking what they’ve been doing with the data between January 1998 and November 2008.
Well here’s what they are doing.
0.0105 increase in temp per year with the December 2008 release for Jan’98-Nov’08
0.0107 increase in temp per year with their April 2009 release for Jan’98-Nov’08
0.0109 increase in temp per year with their July 2009 release for Jan’98-Nov’08
^^^^ That’s all from the same January 1998 thru November 2008 time frame. I’m not including the new data afterwards. It’s the same 131 months. Yet somehow GISS has managed to make the increase in global warming in that time frame grow and grow with each months new release. Simply amazing.
I know, it’s just weather:
Remarkably Cool, we are talking about frosts and freezes occurring in July!!!
Updated: Friday, July 17, 2009 7:32 AM
Temperatures currently in the nation’s heartland are more typical of April or October than July across many locations. Highs will struggle to reach 60 Friday across the Upper Peninsula of Michigan and the Lake Superior region of Minnesota. The high in Chicago will briefly touch 70, Detroit will struggle to get past 70, and Minneapolis will never reach 70. Many places will see highs a good 10 to 18 degrees below normal.
What is going on? Well, a northerly breeze, cloud cover, showers and the jet stream being unusually far south for this time of year will all play a varying role in the chilly air. Look for another cool day in the same locations Saturday after a very chilly night overnight. Some areas in northern lower Michigan have seen temperatures drop to near freezing (We are talking about frosts and freezes occurring in July!!!) recently, and while temperatures will not drop that low thanks to a breeze and some cloud cover, it will still be chilly Saturday night.
Story by AccuWeather.com Meteorologist Mark Paquette.
http://www.accuweather.com/news-story.asp?partner=rss&article=4
I think what bothers me is the base period that 2009 was measure against – 1961 to 1990, which would have been “cold” biased, and would make this year seem much warmer. Wy is there no standard? Why is just a 30 year average used vice a much longer term time frame that actaully may show a cyclical pattern?
Advocacy in science is very corrupt and I do accuse them of advocacy
Paul (woodfortrees):
Regarding the possible lag between surface and satellite temp response to El Nino, we were having an interesting discussion over at Lucia’s place: http://rankexploits.com/musings/2009/giss-updates-increase-recent-historic-trends-slightly/
There appears to be some lag between the two, but its hard to tease out because ENSOs vary so much in magnitude and temperature is so noisy. There definitely was a faster surface response than satellite response at the start of the last major El Nino back in ’97.
If you want to remove the ground station mess from the satellite data you begin with just the satellite data itself. The challenge there is how to handle the change over from one satellite to another, and the aging (drift) if the sensor.
My proposal would be to identify 3-6 well documented ground stations around the globe which have good performance and span all 31 years of the satellites. These are the statistical anchor points for the satellite drift and transitions. If, on average, these 6 stations indicate a life time change in Sat1’s sensor from +.5°C to +1.1°C we can adjust for sensor aging. If the same 6 stations detect Sat2’s bias dropping to +.3°C when it took over for Sat1 we can not only adjust for drift but transition.
The key is to only use a small number of ground reference points so you don’t add mountains of sensor errors from the ground.
Once you compute a satellite only plot for the last 31 years, then you can not only compare to NOAA and GISS, you can estimate their errors and biases – even regionally. You can show how each surface sensor performed against the satellites (one Sat or across satellite)
Turn the tables, use the satellites to question the ground sensor errors and calibration. It is, in fact, the proper method scientifically and statistically. Single spacecraft sensor with known and measured performance is the only way to check the performance of thousands of independent sensors with varying precision and health.
When you do this, the NOAA-GISS anomalies are no longer global warming, but data showing how bad the ground sensors are performing! Which is the how this should have been computed in the first place.