Gavin was kind enough to point out to me that GISTemp provides spatially-explicit output tables as well as graphing capabilities, which allows for a much more nuanced examination of regional trends. I downloaded the data and plotted it in a wonderful little program called Panoply to create the following images. You can create your own maps here. I’m using a Mollweide projection to ensure that areas of the earth are depicted accurately according to their area.
First up, 1900-1940 warming:
Next, 1940-1970 cooling:
And finally, the modern 1970-2010 warming:
Some interesting things to note:
- Warming during the 1900-1940 period was not dominated by the Northern Atlantic per se, though the Atlantic Ocean as a whole warmed up more than most other parts of the world.
- Northern hemispheric land regions cooled much more during 1940-1970 than southern hemisphere land regions. Oceans are less clear-cut, with warming in the Northern Pacific but cooling in the Northern Atlantic, and cooling in the Southern Pacific/Indian Ocean with warming in the Southern Atlantic.
- The modern warming period, while more pronounced in northern latitudes, seems considerably more uniform than the 1900-1940 warming period.
Update
I’m also attaching variants of these graphs using a 250km smooth instead of a 1200 km smooth.
1900-1940:
1940-1970:
1970-2010:
1970-2010 Looks like a massive fireball. Nothing has changed about climate science whatsoever since it first started. They draw pictures of globes that are colored in by 2nd graders and then they put in values at the bottom that don’t even make sense.
By the looks of the 1970-2010 map, it looks like Russia has warmed by over 1.6, yet we never hear them talking about the warming in Russia. Secondly, what are the temperature scales in regards to? If we used historic GAT like we should, that map should be completely blue, with lots of dark blue marks.
Shoosh,
The temperature scales aren’t in regards to anything, they are simply trends for the period in question for each gridcell (using a 1200 km smooth; feel free to use the 250 km one as well if you want, as it is instructive to see how land coverage has changed over time).
First thing I noticed is that the Warming isn’t Global.
Andrew
Also, who is this twerp Tamino? Just like real climate, he wont’ let me post anything. The cowardice on the warmist’s side is unrivaled. Gavin is paid by taxpayers to run realclimate…the only people who go there are his fanboys…it is not a site for the public to be informed.
Hey guys, my name is Jay. I have a real jerk of a brother named Ray who makes a lot of ridiculous claims. Feel free to ignore any comments you might read from Ray.
Andrew_KY
Andrew, how bout how 1979-2010 looks like a fireball? If I didn’t know that the planet in question was earth I would assume it was Venus or something and uninhabitable.
Zeke, I like you, your a nice person that is trying to point out new things all the time but I expect much more from NASA. This is the work of a sad clown with a lopsided rubber nose, a flower that doesn’t squirt celtzer water and a banana cream pie without the cream. The “supposed” trends for the period 1970-2010 seem to be completely wrong in my estimation. My interpretation of that map is that the world is cooking and that is far from the case.
Zeke… let’s get to the punchline and show how many stations were reporting from 1900 – 1940
Shoosh,
I’m producing this particular map, not NASA. Is there is scale and color scheme you would prefer to see? I’d be happy to make it for you.
MikeC:
1500 in 1900, 2500 in 1930. Increases pretty linearly between the two dates: http://rankexploits.com/musings/wp-content/uploads/2010/09/Screen-shot-2010-09-17-at-3.32.09-PM.png
RE: #67033
How about the number of stations that reported for all of 1900-1940? You can’t compute a trend for a station that came online, say, in 1937. Or do you?
Problem is the paper we are discussing uses data sources DIFFERENT THAN GISS.
That’s part of the point.
So, to address the question by looking at a subset of the data used in the study you miss the point.
For example, the study included Bouy data, and I believe some additional data from Norway and Russia not used by GISS.
I havent had time to do a complete cross check but that is where people should start.
Merely throwing up GISS doesnt address the interesting issue in the paper.
Although you do want to look at both. That’s also the point. Talking to peter, his biggest concern is the lack of a detailed investigation of this.
After reading Ar4 on this, I would have to agree. The other point is that their is a tendency to shut down investigations of natural variability or more charitably, they are not encouraged.
Mosh,
The buoy data isn’t included in Reynolds/HadISST? In that case, I’d imagine it will be in HadISST3 when that comes out, as they tend to include any data they can get their hands on.
My reading of the GISS data is that it tends to support Webster’s claim that 1900-1930 warming was not focused on the Nothern Atlantic. In general, I’d be rather surprised of additional data significantly changed regional trends given the relatively high spatial correlation, apart from perhaps more pre-1950s data in the southern hemisphere where coverage is sparse.
Oh sorry Zeke. Instead of using a new color scheme, can you post a globle labeled “last 10,500 years” use the paint bucket tool and dump red into the entire thing. Then that would be fair to me.
Another problem I have is that the map necessarily implies sustained warming when we know the ocean temperatures have not been steadily rising. Zeke I commend you for doing some of your own work, it is really hard to make temperature pictures of the globe because they just seem so generalized.
MikeC,
The number of stations only goes to the uncertainty of the estimation.
As I said there are several explantions for it. ONE being the result is
random.
That would amount to saying this.
in 1900 there are a many places we could sample the temperature.
Lets say 1 million.
I select 100 of them at random where the population is less than 1 person
per square km.
Then 40 years later I look at those 100 stations.
50 of them have had no population growth.
50 of them have grown to urban densities ( more than 100 people/ sq km)
IF you think that UHI ALWAYS HAPPENS when population goes up then what is you expectation about those two data sets.
What’s your expectation? mine was that the urban sites would show more warming.
Even if I picked 2 sites my expectation would be that the urban one would warm more.
Put another way, if the rural sites show less warming, I doubt you would even consider the issue of the number of sites.
In anycase, one explanation for this reversal is that the UHI effect is small.
AND its limited to Tmin. And as chance would have it the sites selected were unlucky.
Hmm.
In anycase Zeke might be able to do a more detailed better job of this by focusing on the US only as I think the population data is better there.
But, go ahead and put a hypothesis out there.
Which will show more warming. 50 rural stations that stay rural or 50 rural stations that become urban.
note ( most of this stuff is pre tobs) all of it is pre MMTS?
Which will show more warming. 100 rural stations that stay rural or 100 rural stations that become urban.
Which will show more warming. 200 rural stations that stay rural or 200 rural stations that become urban?
inothers words what N size do you have to have before the signal becomes visible? well that depends on the effect size and variance?
What if zeke uses a paired approach?
Anyway, the point is I think people need to put some hypothesis out there and then see what is what
Re: Zeke (Jan 24 12:59),
“Mosh,
The buoy data isn’t included in Reynolds/HadISST? In that case, I’d imagine it will be in HadISST3 when that comes out, as they tend to include any data they can get their hands on.”
No they are arctic bouy’s from my reading of the paper. I had a look at it a couple weeks ago when peter and I were discussing the paper, but didn’t delve into it. I think these are the bouys that float around on the ice.
Could be wrong. My main point to peter was I thought the paper was interesting because of the fact that the dataset for northern latitudes was
different than that normally used by the standard guys. So basically, it was interesting and something that I thought worthy of a deeper look.
Basically anytime you get outside GHCN my ears perk up ( like with Ron’s GSOD) because of the sampling criticisms of GHCN.
ops sorry MikeC I thought thsi was the urban/rural thread. Disregard its relavance to this thread.
Sorry Zeke, I did say number of stations didn’t I… How about the geographical distribution of the stations?
Mosher,
What I would expect to find is what is rampant in the literature, that UHI will show up in Tmin. It’s what I found when doing pairwise comparisons. As for you guys, when you have time, the next time you look for UHI, do it with Tmin… once that’s done we can look at Tmax for hydrology/farming/etc issues. I’m not here to argue with you, I want to see how it goes.
Hasn’t the data been modified back to 1850?
Mosh:
Changes in geographical distribution can also introduce a bias in the global mean (from the effect of changing distribution of missing cells on the average). This is easy to see as there clearly is a latitudinal effect in the data.
stephen richards:
Data, metadata or offset used to compare against other data?
What data do you think were modified?
To be fair, the GISTemp data shown here has adjustments for UHI (Step 2), which slightly decreases the trend vis-a-vis raw data. It also uses adjusted USHCN data for the U.S., and the smoothing (250km or 1200km) reduces the spatial variability.
ya carrick, left that part out.
One of the areas that GISS always highlights as a ‘hotspot’ is Siberia.
Why then do Russian scientists seem unimpressed with the evidence for AGW?
Zeke:
Actually that does more than just reduce spatial variability. It is necessary in order to prevent oversampling of regions with a high density of stations. It also reduces the influence of the shifting geographical distribution of stations over time.
Re: MikeC (Jan 24 13:38),
Mike in those pairwise tests what effect size did you see?
Zeke,
It’s nice to see you’ve discovered Panoply. I recognized it immediately. I’ve used it from time-to-time, but it’s pretty slow isn’t it?
Chad,
Maybe I just have a fast machine (Macbook pro i7 processer), but it renders pretty much instantly for me.
Zeke, thanks for the inputs on making these colorful graphs. They can be important tools in attempting to understand the regional variations in a changing climate.
Zeke, do you have a map showing the geographic distribution of stations in the early period?
MikeC,
I can easily get a list of the GHCNv3 stations available each year, but I’m not sure how to map it per se. I’m unfortunately rather bad at using GIS (pre-packaged products like the one in this post notwithstanding).
Mosher,
It depended on the surrounding urbanization. These were the CRN 1-2 stations you did the Google Earth thing with. I used pairs when there was good geographical/climatological zone agreement and which were fairly close to each other. The pairs consisted of a rural and an urban and I took note of when the urbanization took place. Difference in Tmin was roughly between .75 and 4 and probably averaged about 2.0. And remember, these numbers were AFTER they were fed through the USHCN v2 algorithm.
Thanks Zeke, There is that page where you can make the little maps, I have a minute, let me see what I can cook up.
Re: MikeC (Jan 24 16:43),
simple mike. list the stations.
Re: Zeke (Jan 24 16:41),
If I wasnt in the airport now Zeke I would write a program for you to do the mapping.
Piece of cake in R.
Zeke you can also use google fusion tables to map the stations.
Google fusion tables are straight forward.
just upload a csv and then it has the tools you need
Here ya go Zeke… I’m not so sure how someone can get trends worth a doggy doo with this distribution… aside from a great deal of creativity
grrrrr…..
http://roqtock.com/id13.html
Sure Mosh… will go back n look up some of the pairs
right off the top of my head… Kingston n Providence RI
Corvalis OR n Vancouver WA… Let me do a little looking after dinner
MikeC,
Having a bit of a problem trying to get the data reformatted to list station_id by year. This is the best I can do at the moment: http://www.mediafire.com/?6kic9tczktb3t7b
Port Gibson n St Joseph… I think those are in the LA MS area…
Nandan and Dickerson… I think ND
Morris n Milan
Anyways, Mosh, you should be able to pull those right up… they’re on your google program too if you need more station ID
Data source was the CDIAC page
Well as the 10 year trend often comes up in this blog, I looked at the regional trend for 10 years. Very much a sandwich of warm Arctic, cool mids and warm equator near the Atlantic and cool equator in the Pacific. We have heard a lot about Arctic amplification/rapid ice loss and negative AOs/NH snow so not overly surprising. But the fact that the pattern is roughly mirrored in the SH was a bit suprising to me.
How did you choose the time periods?
Aren’t the relevant “warming eras” (at least according to GISS) 1910-1945 and 1975-2010?
Indeed, 1900 appears to be a warmish period.
http://data.giss.nasa.gov/gistemp/graphs/Fig.A2.lrg.gif
Zeke, i’d like to see +/- 0.25 as a neutral white, it would be less biased colorwise.
Zeke,
Looks to me like the trends north versus south during the 1900 to 1940 warming period was almost opposite the 1940 to 1970 cooling period, but with an overall positive trend everywhere. Looks like the poles are oscillating with a 60-70 year period, with an underlying overall positive trend.
ad,
Here you go:
http://i81.photobucket.com/albums/j237/hausfath/trend1900-1940white.png
http://i81.photobucket.com/albums/j237/hausfath/trend1940-1970white.png
http://i81.photobucket.com/albums/j237/hausfath/trend1970-2010white.png
SteveF,
Its hard to say with only those snapshots. You might be better off plotting the temperatures for high and low latitudes separately and seeing how they change.
The near-zero-is-white graphs are clearer.
How do you even know the temperatures for the whole state of Alaska in 1900-1940? (which wasn’t even a state yet)?
Zeke,
.
Do you have any comments on the observations that:
1) aerosol effects tend to be localized (typical atmospheric lifetimes of <10-14 days)
2) man-made aerosols are mainly produced in the northern hemisphere
3) man-made aerosols are assumed by many climate models (eg GISS Model E) to be currently off-setting a large fraction of the total GHG forcing (on the order of ~1.3 to ~1.6 watts per square meter off-set, vs ~ 1 watt per square meter of assumed net GHG forcing), and
4) warming is far stronger in the north, where off-setting aerosols are expected to be far stronger.
.
I simply see no way to square the above with the temperature trends you have graphically displayed. If anything, the trends are exactly the opposite of what would be expected if huge aerosol effects are real. I think your graphics cast doubt on the large aerosol off-set explanation for high climate sensitivity along with modest (~0.8C) temperature rises through 2010.
Liza, There were stations there including Nome, Anchorage, Fairbanks and a bunch of others… but the smoothing is 1200km for the first set of diagrams and 250 for the second set… none were there for the entire period ao you know there is a lot of guesswork going into the GISS data… but in general, they show trends that I would expect from natural variation.
as you know…
… on second look, there were zero stations in Alaska open in 1900 so I don’t know how GISS can show data for that year on their LOTI site… smells a little fishy
This “Russian scientists don’t believe in global warming” (see #67024, #67052) is an interesting(*) psuedoskeptic myth, haven’t encountered it before.
Anybody know where this fiction originated?
*Factual nonsense, as the Russian Academy of Sciences is issuing statements like these:
MikeC, I think there might have been a small “scientific presence” looking at thermometers sometime in Fairbanks Alaska in the 40’s during or around WWII but I doubt the entire state’s temps are known down to tenths of a degree for that time and before. At least not to color that graph a glaring RED like that. 😉 Same for Canada and the Arctic probably -that is a huge piece of the Earth.
Hi Zeke
Would it be possible to do a set of graphs (with the 250K) where the temperatures around 0 are left neutral. It seems to me to be at best problematic, at worst misleading to have a change within the likely measurement error to be bright yellow if slightly positive and blue if negative.
I don’t know what the range should be – and you are more likely to have a good guess – but I would have thought something like +/- 0.2 being white (with grey meaning missing) would be sensible.
Thanks
Margaret
Unfortunately UHI for december was +0.18C and looks like January is again going negative so much for this B###. Anyway i dont beleive any of the data adjusted by NASA, HADCRUIT or now the WMO. My father was a WMO expert just to let you know
I’m waiting for Zeke’s reply to SteveF about the off-setting aerosols. I agree the maps seem to caste doubt on the the role of aerosols.
What I keep harping on, and no one has criticized my thesis (does that mean it’s probably right or, rather, so unimportant as to be ignored) is that we have a 300 year long term trend of about 0.7 C warming per century (for reasons poorly understood) leading up to the age of significant carbon emissions which the “consensus scientists” (note, I did not shorten it to the “con scientists”) say began around WWII or later. Is there any reason whatsoever to believe that this 0.7 C trend of 300 years is not continuing? If we have no reason to think it is not continuing, then any anthropogenic global warming MUST be measured against that trend. Of course, we can’t infer what the current long term trend is from short time periods, but is there anything in the trend since WWII that looks different from the past 15-30 year warming and cooling periods within the overall trend of recent centuries? I’m so tired of the mischief of the warmists who manipulate the mostly ignorant public and journalists by fear mongering the “hotest year” and “hotest decade” when that’s exactly what has happened regularly over the past 300+ years and exactly what we should expect with a 0.7C warming trend caused by so-called natural variability- forcings poorly understood, if understood at all. Zeke, your three maps are very helpful in understanding the distribution of warming and cooling during those three periods.
Wonderful. Warmer temperatures at higher latitudes should make the planet much more livable. But I doubt that we would be that lucky. Where did the actual data come from? Was it the raw data from real stations measuring real temperatures or value added sets that have seen an artificial warming signal added to the actual measurements? Was the UHI effect adjustment made? And did the data use the official NZ record that the government has backed away from?
Liza, The whole area has been a boom town since the 1800’s for gold… there has been a large fishing industry there since the late 1800’s and the military has been active there because of it’s strategic importance (it’s why it became a state in the first place along with Hawwaii. The thermometers in the Aleutians have been there since the late 1800’s but none of the others like Nome, Anchorage or Fairbanks were there in 1900 when Zeke’s first time period started.
– w
Doug Allen, I’m not sure how accurate your numbers are… but of course, a good portion of the warming in the past 100 years has been from emergence from the little ice age
Mike C.
What numbers are you referring to? Yes, the 300+ year warming trend is from the LIA which also is the beginning of instrumental temperature records. Those early temperature records are not global- few in the SH, etc., but the instrumental data is certainly better than proxy data or is there disagreement on that? And that instrumental data shows about a 0.7C per century warming since the LIA. Is there any reason we should assume that that 0.7C trend has stopped- none that I know. In the absence of evidence that the 0.7C trend has stopped, shouldn’t calculations of AGW be based on that trend- warming greater than the 0.7C trend may well be AGW, warming of about 0.7C is probably a continuation of the trend (and CO2 forcing doesn’t amount to much), and warming less than 0.7C might suggest negative feedbacks, as Dr. Spencer speculates, or may suggest that the 300+ year 0.7C trend is changing. There is a lot of unavoidable ambiguity, and that is part of what I am saying, as opposed to the IPCC statements which minimize ambiguity and uncertainty for reasons we all understand.
Hi Doug,
My question about your numbers would have been that portion of the numbers from between 100 and 300 years ago. I agree that the LIA ended about 100 years ago. I’m just not certain if it was much colder 200 years ago than it was 100 years ago.
The reason I make this point is that there were opposing warming and cooling forces going on at the time. Solar was low but ocean was warm. In the past 100 years, solar has come back with a vengence while ocean activity is on a long term cooling trend.
That’s all, I just wonder if that .7 didnt start around the turn of the century or if really went back the whole 300 years.
Polar amplification (both directions – warm and cold) seems to be a real phenomenon. It shows up in every historical record. For both poles.
Arctic temperatures seems to swing by twice as much as the rest of the globe. If you look at the 1940s warming in the Arctic, it was twice as high as the global average. It cooled twice as much up to the 1970s.
In 2010, it warmed about twice as much again.
Is that a global warming signal or a cyclic Earth atmospheric thermodynamic signal?
If we’re entering into a warm climate cycle then it might be the norm for the next several centuries.
Bill Illis (Comment#67099) January 24th, 2011 at 11:57 pm
That is why they went looking for ‘fingerprints’. Warming at the poles, but cooling in the stratosphere, for example.
Basic question.
Is this data post “homogenisation”?
Mosher, what you ignore is that population density may increase *because* it’s warmer.
Zeke: Warming during the 1900-1940 period was not dominated by the Northern Atlantic per se
.
Er… What’s your definition of “dominate”? In your first graph, North Atlantic warming look pretty “dominant” to me!
.
Admittedly it’s less clear on the 250km graph, though NA still seems to have the hotter spots.
Why is the coast off Equador always cold
MikeC (Comment#67094) January 24th, 2011 at 9:08 pm
Exactly what I said but different. LOL
Thermometers have been there since the late 1800’s…okay. What evidence do you have of this..or more importantly what evidence do you have that they were accurate down to tenths of a degree, covered an area big enough to represent the entire state; were used enough to represent a year or a decade’s worth of data; were they calibrated to each other?… and even further, what evidence do you have they could record temps LOW enough? 🙂 😉
Here is some text I found:
Makeshift Alaskan Thermometers
We had no thermometers in Circle City that would fit the case [reach low enough] until Jack McQueston invented one of his own. This consisted of a set of vials fitted to a rack, one contained quicksilver (mercury), one of the best whiskey in the country, one kerosene, and one Perry-Davis Pain Killer. These congealed in the order mentioned and a man starting on a journey started with a smile at the frozen quicksilver, still went at whiskey, hesitated at the kerosene, and dived back in his cabin when the Pain Killer lay down. -Recorded by William Bronson, The Last Grand Adventure, describing Alaska in the Gold Rush era of the 1890’s.
http://books.google.com/books?id=SV229set7RIC&pg=PA51&lpg=PA51&dq=when+was+the+thermometer+used+in+alaska?&source=bl&ots=j8eii1jHT1&sig=TzywvuMBTGHodIcq9ZPQOBTtzWM&hl=en&ei=Q78-TfjcE4jWtQOnwvSmBQ&sa=X&oi=book_result&ct=result&resnum=9&ved=0CFYQ6AEwCA#v=onepage&q&f=false
[page 50]
Wayne Arnold, The coast of Ecuador is always cold because it’s a major area of cold water upwelling.
Liza, let me answer your questions one at a time.
I’m sure the thermometers were there in the late 1800’s because I pulled them up on the station list over at GISS and looked at the start dates
They were not measuring to within a tenth of a degree… but you get tenths after averaging them all together.
(the rest of your questions require a little creativity to answer)
The area they covered is not so important because we can always use the theory that the temperature trends here on Kodiak Island the same as all the mountains and valleys for 1200 km in every direction.
The data at Holy Cross, for example, which opened up in 1893 is only missing 5 years worth of data between 1900 and 1920…ummmm but remember, Liza…. TRENDS! Were talking TRENDS!
I’m sure they were all calibrated in the factory they came from and that the thermometer factory in New York knew exactly where zero was, just like the thermometer factory in London. Remember liza, it was only in the late 1800’s when our civilization was technically advanced.
And finally, I’m sure they always went out in major sub-frigid temperatures and braved the deadly freezing wind to get the temperature every night at midnight. After all, with a name like Holy Cross, they must be telling it the way it was.
“The area they covered is not so important because we can always use the theory that the temperature trends here on Kodiak Island the same as all the mountains and valleys for 1200 km in every direction.”
Yes siiirrrreeeee bob. I think the key word there is “theory”. 😉
“I’m sure they were all calibrated in the factory they came from and that the thermometer factory in New York knew exactly where zero was, just like the thermometer factory in London. Remember liza, it was only in the late 1800′s when our civilization was technically advanced.”
Sure because everyone knows the temperatures less than zero, you know up to minus 100 or 70 degrees F don’t matter to the “trend”. LOL
Some Lunatic Denier masquerading as a Meteorologist got on the radio this morning and said this winter in Kentucky has been colder and snowier than normal.
People with a political agenda will say any outrageous thing these days.
Andrew
“And finally, I’m sure they always went out in major sub-frigid temperatures and braved the deadly freezing wind to get the temperature every night at midnight. After all, with a name like Holy Cross, they must be telling it the way it was.”
Can’t you just picture it? 🙂 How did the candle stay lit…or were they holding a lantern? Lantern in one hand; pencil and paper in the other in their long johns? 😉
Brrr you guys. I hate being cold. (Hi AndreKy!) I am shivering just reading these things. :p
There is no evidence that they were, or that anyone everclaimed they were, or that they need to be.
If there are ten thermometers, accurate to 1 degree, reading 1,1,1,1,1,1,1,1,1,1.
Then after a period of time they read.
1,1,1,1,1,2,2,2,2,2.
We could conclude quite safely that they are recording a rise in temperature.
bugs, I repeat…Sure because everyone knows the temperatures less than zero, you know up to minus 100 or 70 degrees F don’t matter to the “trendâ€. LOL!!!!!!
And bugs…hey no problem..the temp could be .8 in real life that day but you know they just rounded it up to 1 for kicks. All is good when it comes to the data… Hide the decline and all that…LOL!
Liza, That actually happens. I was talking to an observer once who was happy that the National Weather Service installed his MMTS closer to his house because he was tired of getting up early in the morning and walking through the snow to get to his CRS. I could tell by his expression that on very cold mornings he just blew it off.
bugs,
Except no thermometer readings work the way you describe.
Andrew
Except now, most of the thermometers are automated and they round it up or down to the nearest degree automatically… or the observer does it in case his MMTS isn’t connected to a computer yet.
Brrr.
I know my husband has a heck of a time when it’s raining writing down numbers on paper..which he has to do. The paper gets wet..the pencil/pen smears -all kinds of stuff happens. People sitting in front of computers all day can’t see, feel or imagine any of these real life things it seems to me.
Whilst the danger of staring a pretty pictures is that we think that we see Little and Large Bears in the sky and so on, it does look to me like a pulse or throb emanating from the north pole; an inhale/exhale of warmth/cold, if you will.
Interesting pictures Zeke, thanks.
liza (Comment#67113) January 25th, 2011 at 8:06 am
They do matter. Zeke has listed all the temperatures, the data is open, he shows exactly how he got the trend. No lols or caps needed.
liza (Comment#67120) January 25th, 2011 at 8:36 am
I know my husband has a heck of a time when it’s raining writing down numbers on paper..which he has to do. The paper gets wet..the pencil/pen smears -all kinds of stuff happens. People sitting in front of computers all day can’t see, feel or imagine any of these real life things it seems to me.
The scientists can’t win, can they? They don’t understand the problems of raw data from the field, but the raw data from the field is no good, so we can’t trust the temperature record. The problem is you don’t seem to have bothered to look at what they do when they take raw data, that has problems, and take those errors into consideration.
… actually bugs, the context of the temps, LOL’s and caps was that people will avoid going out in the cold to do their volunteer service for the National Weather Service when it’s very uncomfortable (like in Alaska where the context of the conversation was), creating a cold bias in the record…
excuse me… creating a warm bias in the record
“The problem is you don’t seem to have bothered to look at what they do when they take raw data, that has problems, and take those errors into consideration.”
Quite to the contrary, bugs. The current method is to apply a mathematical formula which was designe3d to catch large discontinuities. The published numbers indicate this formula misses discontinuities at an average of .4C. And since there are many more warm discontinuities in the record than cool, there is a warm bias simply due to the formula itself. Not to mention the statistical margins of error and etc.
I think a normal reasonable person; with absolutely no belief system set in stone about such things, or funding needs would see nothing on these graphs out of the ordinary or alarming. Especially if someone cares to “look at what they do when they take raw data, that has problems (and the large gaps in history for people to “fill in” as needed”), and take those errors into consideration.” Especially if they have a healthy idea of the history of this planet’s climate and how much we don’t know about how it operates on such small time scales!
Liza, Not to mention that these guys (especially the USHCN) have been doing this for years. A lot of them talk about comparing what they read on a particular day to what other stations are doing… so when they want to fake it because they do not want to go out into the blizzard, they know how to get the information to make it close. The daily numbers go through the quality controll checks and those numbers have to be pretty far off in order to be thrown out.
MikeC (Comment#67124) January 25th, 2011 at 8:52 am
“people will avoid going out in the cold to do their volunteer service for the National Weather Service when it’s very uncomfortable (like in Alaska where the context of the conversation was)”
When I was in the Air Force in Alaska our unit shared the barracks with the Weather people. When it got cold hourly readings became start of shift readings with the hourly readings ending up being whatever the forecast predicted. For the purposes of aviation, accurate temperature readings are required when it is hot in order to determine max takeoff weight. On the cold side once one is well below the temperature where icing is a potential issue a few degrees off really doesn’t matter, except to those people who come along 20 years later and do a statistical analysis to determine a long term trend.
Zeke: I recently had a go at the differences in the warming and cooling periods using the trends in Zonal Mean Temperatures (also available through GISS). In the following graph, the early warming period starts in 1917:
http://i51.tinypic.com/2v8j1gg.jpg
Curiously, if I started the early warming period in 1907, the Arctic trends then greatly outweighed the trends in the recent warming period:
http://i55.tinypic.com/2e67o1w.jpg
Refer to the post:
http://bobtisdale.blogspot.com/2011/01/notes-on-polar-amplification.html
Roger Andrews and I in the comments then discussed and illustrated how the rise in Arctic temperatures preceded the warming of the lower latitudes during the early warming. I split the Arctic land surface data into two subsets hoping it might be revealing (no such luck), compared them to scaled global GISS LOTI, with all of the data detrended:
http://i51.tinypic.com/rrjwva.jpg
And then using northern hemisphere GISS LOTI:
http://i52.tinypic.com/35hea9i.jpg
Curious.
harrywr2 (Comment#67135)
Sounds about right. The USHCN station observers who we have been talking about are low paid or volunteer observers. In the old days when the whole network was CRS (big wooden boxes with two thermometers in it) the observer would go out and record the high and low for the previous 24 hours. Now most of them have the MMTS which is an electronic system which is wired to a display panel which is inside their house, so it’s not as much of a problem as it used to be… except that the thermometers now have a habit of being located closer to buildings etc.
I have a question about the surface stations I don’t think anyone has asked. What is there range? For example, how large of a radius can the temperature stations measure temperature? Also, have they tested this by placing another temperature station inside an existing radius to ensure the temperature readings match each other?
I have a question about the surface stations I don’t think anyone has asked. What is there range? For example, how large of a radius can the temperature stations measure temperature? Also, have they tested this by placing another temperature station inside an existing radius to ensure the temperature readings match each other?
Are we really questioning global warming because Olaf couldn’t be bothered to go read a thermometer a few times in January 1923?
Shoosh,
People have asked that question, notably James Hansen and his collaborators at NASA when they did the analysis underlying the current GISTEMP methodology which justifies their decision to assign temperature anomalies to areas where no surface measurements are taken based on the measurements at other stations within 1200km. If the question interests you, read their papers. For instance, http://pubs.giss.nasa.gov/cgi-bin/abstract.cgi?id=ha00700d and the other papers among the references given at http://www.realclimate.org/index.php/archives/2010/01/2009-temperatures-by-jim-hansen/.
Shoosh, The thermometers sense the temperature of the air that flow across them.
No Borris, just a few thousandths of it… and had you taken the time to read the thread you would have noticed that the point was made that the problem is almost non-existent now because the thermometers have been automated. Quit trying to contrive an issue.
@MikeC
ho ho ho very funny. I meant, what is their accuracy range? 100 miles? 200 miles? I don’t think we should base anything off of the surface stations. I’m sorry but they clearly have not been tested for errors.
I don’t understand the defense of the thermometers. If Olaf couldnt’ be bothered to check the stations then we should simply throw the record of that station out, or at least the portion that is not verified. What good is the data from a station if it is wrong?
I must still be missing something. Industrial Revolution, man-made CO2 has purportedly caused an increase of 0.7C over the past 150 years or so. Meanwhile -based on this thread alone- there appears to be little contention as to the fact that during the 20th century there were three very significant cyclical variations in temperatures.
Given that the anthropgenic contribution to CO2 concentrations has been growing at an increasing rate, where does that leave CO2 as the single key forcing [ N.B: not my words but the IPCC’s]?
Until that question is answered in a convincing and verifiable fashion, the rest is a possibly interesting conjecture from an academic perspective, but an otherwise completely irrelevant and dangerous and very onerous premiss from a policy point of view.
Boris (Comment#67150):
No. What people are questioning is how accurately we can say what the average global temperature was in the past, or even at a particular location.
In order to make a meaningful comparison between the temperatures of today and past climates, we have to have fairly accurate past climate data.
If the error bars on the past climate data are to large, then there can be no comparison.
Say we know the temperature 1000 years ago to plus or minus 25 degrees C. Say it is 50 F plus or minus 25 F (global average). If the global average for 2010 is 52.05 plus or minus .1 F, what can you really say? Is it 2.05 degrees warmer than 1000 years ago? You don’t know because it could be anywhere in the range from 25 F to 75F 1000 years ago. (By the way this is an example with made up numbers).
So to the extent that anything makes any past climate data have larger error bars, it diminishes the agility to compare to today.
That is what people are saying, in my opinion.
Bob Tisdale, interesting comments.
It looks to me like the polar amplification is decreasing over time (we’d expect this actually).
Is this your impression too?
Shoosh, It depends on the density of the network in any given area. Stations can be 30 miles apart and they can be 600 miles apart. There are still some areas on the maps where there is no temperature represented (gray areas on Zeke’s maps)
The smoothing on the maps made by Zeke were 1200 km in the first set and 250 km in the second set.
Since there is overall warming from the begining to the end of the century, any baseline used will always show that the end of the century warming is “warmer” than the mid-century warming. That pretty much makes the GISS based temperature data being presented worthless for what is being discussed. What needs to be shown is the rate of change, not the difference from the GISS baseline.
That would be more work but unless I am reading the numbers wrong, would show little difference between mid and late century warming.
Vernon,
What is being shown are spatial trends. They don’t have a baseline per se.
Zeke,
How do you show them if you do not have a base line for each grid box?
Because each grid box shows the temperature change from the beginning of the period to the end of the period specified.
Carrick (Comment#67159): Sorry to reply to your question with a question, but, why would you expect a decrease in polar amplification with time?
Doug Allen (Comment#67092)
January 24th, 2011 at 8:50 pm
“I’m waiting for Zeke’s reply to SteveF about the off-setting aerosols. I agree the maps seem to caste doubt on the the role of aerosols”
Me too. The strongest cooling in the 1940-70 period appears to be in the high arctic. Plenty of studies show that sulphate aerosols produce ‘arctic haze’ which causes WARMING. Mann & Jones co-authored a paper (circa 2003) which stated that the effect of aerosols was “regionally specific” – for the very reason given in SteveF’s point 1. If aerosols were responsible the cooling should have been concentrated over the industrialised mid-latitude regions in the NH.
John Finn, If the aerosols did create changes in local temps, the GISS homoginization adjustment would most likely catch them and take them out… there’s just no hope for the global dimming fantasy
Bob Tisdale:
Because the polar amplification is mostly related to the area of the permanent snow&ice surface area, which of course correlate with the mean albedo of that region. As the area of permanent snow (and permanent sea ice for that matter) gets smaller, the effect associated with the remaining permanent snow/ice gets smaller.
(Of course this is saying a different thing than saying we would expect the latitudinal effect itself to disappear with enough warming, which I agree won’t occur.)
Carrick,
“Because the polar amplification is mostly related to the area of the permanent snow&ice surface area, which of course correlate with the mean albedo of that region. ”
.
I think (not sure) that there is supposed to be a considerable water vapor amplification component as well (warming air, more moisture, more radiative forcing) which is not supposed to depend on snow/ice cover. So warming at high latitudes is suppose to increase heat lost to space by less than the rise in surface temperature alone would suggest.
… ahhhh the two different definitions of Polar Amplification emerge
Zeke,
You have to have a base line to get anomalies. What base line are the reference boxes using in your graphics?
MikeC (Comment#67107)
Wayne Arnold, The coast of Ecuador is always cold because it’s a major area of cold water upwelling.
Thanks for the info.
Re:- Accuracy. The Aus BOM states that temp readings prior to 1910 should not be used. Before 1910 stevenson screens were not widely used, so the results cannot be trusted.
I wonder how many parts of Africa & Sth America used stevenson screens in the late 1800s early 1900s
Vernon,
The anomaly baseline is irrelevant in calculating the trends, and these maps are showing the trend, -not- the anomalies.
Re: Carrick (Comment#67168)
My understanding was that the polar amplification is, at least in great part, due to extension of convection cells poleward in a warming regime (e.g., Schneider 2010).
oliver, Another one? That makes 4 different definitions now.
I’m thinking someone needs to put together an encyclopedia of climate terms.
Arctic Amplication
1) Increased surface trends the farther north you go
2) Increased heat transport through the troposphere at the Arctic
3 Decreased winter Arctic sea ice leading to faster melt time in the melt season
4) extension of convection cells poleward in a warming regime
I’m glad sea ice aint my bag
…Amplification
MikeC,
Your definition (1) sounds like the right one. Your (2)–(4) would seem to be explanations for (1), not new definitions. (aren’t 2 & 4 quite similar though?)
oliver:
Technically that’s not “polar amplification”, since it has to do with latitudes around ±30°… I was referring to that as a “latitudinal effect”. I’m not so sure about the “great part” either: How does an effect that exists primarily in subtropical explain the fact that most of the amplification is seen at greater than 60°N?
INaw, they all come with their own little itty bittys… I hadn’t seen the one by the late Dr S til you plugged it.
SteveF:
As I stated, there are other effects besides permanent ice loss. I haven’t seen a table that estimates their relative magnitude, though I’m sure it exists. I’d have to think about why this effect you describe should be more dominant near the poles though.
I was explaining why I expected a decrease in polar amplification. I doubt it would saturate even if all of the arctic ice became non-permanent.
Re: Carrick (Comment#67180) January 25th, 2011 at 11:25 pm
GCMs runs without ice-albedo feedback have demonstrated a clear polar amplification due to increased meridional heat transport. All that is needed is a uniform (or tropical) increase in forcing.
Dr. Shooshmon, phd. (Comment#67156) January 25th, 2011 at 2:01 pm
On average, they will wrong low as much as high. Anyway, the anomalies are what is used, so if they are biased high, it makes no difference.
Hi Zeke,
Would it be possible to produce a map of the 1970-2010 250km warming zoomed in on the Canada-US border, say between 40N and 60N. From the global map, it looks like Canada is warming much faster than the US.
AJ
Re: Polar Amplification
At constant relative humidity, a Joule causes a greater increase in temperature at low temperature where the absolute humidity is low than in the tropics where the absolute humidity is high. While the ghg radiative forcing is higher in the tropics, it’s more likely that the surface temperature won’t get high enough to radiate it all away so the excess heat will be transported to higher latitude. That happens now. More solar radiation is absorbed at low latitude than is radiated and more is radiated at high latitude than is absorbed. The crossover latitude is about 40 degrees N and S and the peak heat transfer is about 5 PW. We also know that when the planet was a lot warmer on average than now with no polar ice caps and boreal forests in Antarctica, the tropics weren’t very much warmer than they are now.
oliver:
Again I’m not arguing that other mechanisms wouldn’t be present. The ice-albedo effect is certainly real, and would be expected to decrease as the globe warms (and vise versa as it cools more). So you’d expect a smaller polar amplification effect at higher temperatures, simply because one of the additive mechanisms is diminishing.
That’s the extent of my point.
Perhaps it’s the case that the ice-albedo effect is negligible, but I somehow doubt that.
Carrick,
“I’d have to think about why this effect you describe should be more dominant near the poles though.”
.
I think because radiative heat loss from the surface is very efficient in the absence of water vapor. Once the troposphere is warmer the water vapor concentration becomes higher at low altitudes, and you have to rely more on convection to transport heat away from the surface. Some rather nifty graphs DeWitt generated a few weeks back show how the delta-T per watt lost to space (the magnitude of the greenhouse effect) increases from the poles to the equator. I do not know how large the effect of this moisture amplification would be compared to changes in albedo.
DP @ 67190
Yes, the earth pumps energy poleward. I wonder if the warm, convected CO2 nearer the poles than before radiates more energy to space than absorbed from the earth. In other words, does increased CO2 concentration actually increase the rate of energy loss to space, thanks to the heat pump that is the earth? I realize that I’m poorly presenting this anti-intuitive concept.
==============
Kim,
I don’t think so. Any added infrared absorbing gas (CO2 or other) should add to the total greenhouse warming.
Can CO2 be an agent for increased poleward transport of energy? That’s a little better.
============
Kim,
Well, maybe in the sense that warmer tropics due to CO2 would increase heat transport poleward via amplified Hadley circulation (more moist tropical convection —-> more Hadley circulation).
Is the increased energy retained nearer the equator from the radiative effect of CO2 balanced by increased loss poleward? Is it less or more rather than balanced. Does anyone know?
==============
It would explain a lot if it were balanced. If not, then maybe there is a problem with CO2, the magnitude of which would depend upon the magnitude of the deviation from balanced, either positive or negative.
=============
Re: kim (Jan 26 09:28),
That’s the converse of saying that polar amplification is due to increased energy transport to high latitude. If it isn’t balanced, then, depending on the direction of the imbalance, heat is retained or lost and the temperature goes up or down in the direction needed to restore balance. So the system always strives toward balance, but may never actually be in balance.
Yes, I understand. But is it imbalanced? And whether it is balanced or not, it seems there is a mechanism against runaway.
=====================
Speaking of ice/albedo, we’re within reach of a 32 year low in global sea ice area (Arctic plus Antarctic) as reported by Cryosphere Today. Today’s result was 14.59 Mm2 compared to the low in the record since 1978 of 14.49 Mm2 in 2006. Arctic area actually dropped a little today so when today’s Antarctic area comes out tomorrow, there’s a good chance there will be a new record low.
Here’s the graph again of albedo vs. sine(latitude). The high albedo at high southern latitudes isn’t going away any time soon and the area of high albedo at high northern latitudes isn’t very large compared to the rest of the planet.
Re: kim (Jan 26 09:58),
It’s been pretty close to balance for the last few years according the the ocean heat content data. OHC is a direct measure of total radiative imbalance and it’s been nearly constant since 2005, if the ARGO data are correct.
Let me try this. The radiative effect of increased CO2 tropically is to trap energy. The radiative effect of increased CO2 poleward is to let more energy escape. And it’s because the atmosphere poleward is warmed convectively.
What is the balance between these two radiative effects of CO2.
=========================
For the purposes of energy balance, these two effects of increased CO2 are opposite in sign. Somewhere between the equator and the poles, the net effect is zero. This latitude is higher when the earth is warmer, and lower when the earth is cooler.
================
Kim, What’s really gonna flick youf bick is when you start to consider the difference in heat transport between the atmosphere and the oceans.
Oh, yes, heat is convected poleward atmospherically and predominantly oceanically. But it’s only the GHG’s in the atmosphere which have opposing effects according to latitude.
================
Re: kim (Jan 26 10:06),
That’s wrong. The radiative effect of CO2 at high latitude is still to initially reduce emission. It’s just not as strong an effect as at the equator. This figure is from GISS from a publication from Hansen and others. I’m not sure why they included a color scale for negative values, as there aren’t any.
CO2 increases emission from the stratosphere everywhere, resulting in cooling.
AJ,
Here you go: http://i81.photobucket.com/albums/j237/hausfath/TEMPTRENDinGHCN_GISS_HR2SST_250km_Trnd1212_.png
Kim, The lattitude thing is about water vapor… not that I’m trying to sell the theory…
Oliver, DeWitt, SteveF, I’ve been muddling through the peer reviewed literature. Here’s a recent analysis paper (Graversen, et al Nature, 2008):
Accompanying figure.
Does this suggest we can neglect the effect of surface albedo feedback on temperature entirely then?
I noticed they claimed it might be more important in the future:
(Emphasis mine.) Is this claim plausible?
There is also another publication which comes to a similar conclusions to by looking just at models. This one is not behind a paywall. [Winton, GRL, 2006]
Manabe’s original paper is here. He seems to be assuming that surface albedo feedback is playing a big role, based on that data analysis, this doesn’t seem to be the case.
If anybody wanted to look at this further, doing land-only temperature trends, and breaking it out by season would probably be the most interesting approach to answering the question of whether the polar amplification factor (which is near 1.8 according to Winton’s modeling) is time dependent or not.
Sounds like an interesting mini-research project.
Thanks for the feedback (;-)), I learned something new today.
I meant to point out that for polar regions that breaking out by season is interesting because one should see the dominant effect from surface albedo feedback due to ice loss to occur in summer and/or early fall.
So the prediction is that you should see a larger polar amplification factor in summer than in winter.
Dew. P. @ 1:20 PM
Thanks, I expected there to be something wrong with it. But, your link doesn’t work, and I don’t trust GISS and Hansen. Is the link to theory or data?
GHGs in general export energy at the poles because they are warmed convectively. Why do I suspect Hansen’s work doesn’t take this into account, and only takes into account the lower incident light angle.
============
I wonder if I can expand the thought a little. If the heat engine that is the earth imports energy toward the equator and exports it poleward then there is a balance point at some latitude. Poleward of that latitude the earth is losing more energy than it is gaining. Poleward of that latitude, all of the GHGs have their radiative effect losing more heat to space than retaining it here on earth. It’s not just CO2.
C’mon, there’s got to be something wrong with this idea. I just don’t see it yet. Help.
==========
Lisa:
Us backwards Canadians didn’t get thermometers until 1981.
Hudson Bay Company Forts measured and recorded temperatures before that but they were mostly Orkneymen. You might be interested in their river freeze and breakup records; continuous from c.1750. A month on the front and a month on the back.
Kim, That works for land but not nearly as much for ocean,
Re: Carrick (Comment#67212) January 26th, 2011 at 3:16 pm
I could really not speak to future changes in the importance of the ice albedo effect. If you imagine a future scenario where summer ice has all but disappeared, then the surface albedo must change quite a bit. On the other hand, then the heat exchange between atmosphere and ocean would be quite different than it is now, and I have no idea what that would do to (esp. low) cloud formation or any of the other important mechanisms.
As for the claim quoted from Graverson 2008, it does appear right after another line which… sort of makes it sound less like a strong claim and more like they don’t want to offend anyone… 🙂
MikeC @ 5:32
Yes, there are toy aspects to my model. The latitude of balance will be different for land and water, I’m pretty sure, but the general principle, that GHGs have opposing effects at the equator and at the poles stands.
============
John McManus (Comment#67216) January 26th, 2011 at 4:42 pm
Do you mean me? 🙂 Hudson Bay has interesting geology. Isn’t it something about there being less gravity there? Or something like that?
Re: kim (Jan 26 15:28),
Sorry about that. The link worked in preview. Here’s the same image that I saved. This has nothing to do with trusting Hansen. It’s basic radiative transfer calculations. You can get close to the same numbers using MODTRAN.
Re: kim (Jan 26 16:04),
As I pointed out above, the latitude where absorption equals emission is about 40 N and S. Here’s a plot of absorption and emission vs sine(latitude) so it corresponds to surface area. The crossover point is also the point of maximum energy transport. At low latitude, ocean transport dominates, at high latitude its atmospheric mainly.
Re: Carrick (Jan 26 15:22),
The maximum effect is going to be near the minimum when the percentage change in ice cover is the greatest. The minimum is about mid-September. There’s usually a lag in this sort of thing too. For the Cryosphere Today Arctic area record, the greatest difference between maximum and minimum ice cover comes in mid-October. The difference was 3.78 Mm2 and the minimum was 4.11 Mm2, so the loss in coverage over the years is about 50%. But by then , insolation has dropped a lot too.
You do start to see an increase in max-min starting in mid-May. The difference then rapidly recovers to about 1.2 Mm2 where it stays about flat until May again. If one were really gung-ho and had the software , one could probably calculate the difference in absorption over the years. Pixel level ice concentration data are available and the insolation for any given location can be calculated. But then you would need to have the cloud cover as well. Sounds like too much work to me.
Thanks, DeWitt. Those are helpful. But aren’t they theoretical calculations of the effect of the concentration of CO2? Where’s the data that shows what convectively warmed CO2, as well as the other GHGs, do at the various latitudes?
================
Alright, presuming that I understand your second graph, I think it illustrates my point. Raise the concentration of CO2 and both absorption and emission will rise, but the difference will remain the same.
==========
Polar amplification is also related to the fact there is less energy to start with at the poles so the same forcing change would impact the poles by more.
Eureka Canada has an average temperature of -20C (or an average surface energy level of 230 watts/m2). The average temperature around the equator is +20C (or a surface energy level of 420 watts/m2).
If the whole surface of the Earth warmed by 16.5 watts/m2 (the amount predicted for doubled GHGs including feedbacks although it probably isn’t equal across the whole surface, this is just an example), …
Eureka would go up by +4.4C while the equator would only go up by +2.8C.
-16.5 watts/m2 (as in the beginning of an ice age) would reduce Eureka’s temperatures by -4.7C (and stop it from rising above zero in the summer) while the equator would only fall -2.9C
kim @ 3:45 AM
Well, no. Add CO2 and absorption will relatively rise over emissions below 40degrees lat, making the difference greater but emissions will relatively rise over absorption above 40 lat, making the difference less and leaving the amount under the difference curve the same.
Well, it’s possible I don’t understand the graph. I’m not very happy with the explanatory power of my first paragraph.
I think that the earth has a natural negative feedback system that is really a defense system of the earth. I also think that negative feedbacks make more sense than positive feedbacks solely based on common sense. I believe the earth was designed for special purposes so it makes sense to me that it should have an extraordinary ability to quickly remove any excess heat.
Re: kim (Jan 27 08:35),
Absorption of solar radiation is only very slightly affected by increasing CO2 and that is mainly a change in the location of the absorption from the surface to the atmosphere. So absent a change in the albedo distribution, the absorption curve will stay the same and the emission curve will flatten (go down at low latitudes because the surface temperature will increase less than the local forcing and increase at high latitudes because emission will increase more than the local forcing) leading to increased net absorption at low latitudes and increased net emission at high latitudes from increased poleward heat transfer.
That makes sense to me, DeWitt; thanks for staying with me. Oh well, I knew it was albedo all along anyway.
==============
Re: kim (Jan 27 15:12),
You’re welcome. I had thought about the consequences of polar amplification on absorption and emission but had a less clear picture of what was happening. It now makes more sense to me as well.
The thing that strikes me is the amount of ocean data for 1900-1940 even for the 250km smoothing.
They have better ocean data than they have for Africa, South America and much of Asia.
Something, methinks, is not right there…
Re: TimTheToolMan (Jan 28 07:36),
Ships. Throw a bucket over the side. Haul it up. Stick a thermometer in it. The information goes in the ship’s log (or somewhere) and is reported later. Lots of ships compared to weather stations in the less traveled areas. The bucket method was later replaced by thermometers in the engine cooling water intakes. Some think the bump in SST temperature during the 1940’s was caused by inadequate correction during the change. If you go back in the archives of Climate Audit, there are several posts on the topic of bucket correction.
There is always the ongoing question of: What exactly ARE you measuring with either the bucket or the intake. You are not measuring the “surface” so what then? (Conversely, an airborne or satellite reading may measure the “surface” but is that what we really wanted in the first place?)
Earlier in the thread, bugs noted that polar amplification and stratosphere cooling was one of the “fingerprints” of global warming that is looked for.
I think it has been shown that polar amplification is more cyclical than anything and polar amplification is a natural result of a globe warming or cooling (the energy level math differentials prove the point in my opinion).
No 2 issue and the main point of this post, is that Volcanoes are responsible for “stratosphere cooling.” The measurements going back to Agung in 1963 and 1964 prove that point.
But have a look at the daily UAH lower stratosphere temps. Pinatubo’s aerosols first intercepted solar radiation in the stratosphere, warming it very significantly. And then Ozone destruction, resulted in it cooling off since less solar radiation was now being intercepted in the stratosphere. The temperatures have been flat or rising for 15.75 years now. (technically, this should have also warmed the surface if anyone wants to do the math on that).
http://img232.imageshack.us/img232/29/lowerstratospheretemps.png
And then I need to do a little poke on the lack of the tropical troposphere hotspot (which is apparently not a fingerprint now since increased solar radiation could also cause a tropical hotspot – somehow that is convincing that the lack of tropical troposphere hotspot doesn’t matter but then a logic step was missed there – there is, in fact, less warming in the lower tropical troposphere than just about anywhere else).
Re: Bill Illis (Jan 28 19:10),
The satellite LS temperature isn’t a good measure of stratospheric cooling due to addition of CO2. The LS weighting function peaks between 15 and 20 km. At that point CO2 cooling is minimal. Where you would really want to look is 30-50 km. See this site for example. It still looks relatively flat at 37-52 km from about 1995 on, but there’s more of a trend. It also looks like a large eruption has long term as well as short term effects on the temperature of the stratosphere.
Re: oliver (Jan 28 13:24),
You’re not measuring the surface air temperature either. The standard is 2 m above the surface. A bucket sample would be the average temperature of water near the surface. Engine cooling intake temperature would be from somewhat lower down. A satellite would read the skin temperature. It’s still useful for weather forecasting.