After a week of mostly stories of this flavor, “Scientist smacks down filthy climate change denier, film at 11“, this article in the Telegraph by Tom Chivers is refreshing and gets it close to 100% right.
Click image for the full report.
So many stories have been written this week with my name and words in them, and only two journalists contacted me in advance to ask me to comment. The first was Oliver Morton of The Economist, the second was Andrew Revkin of the NYT. I thank them.
Another new report worth having a look at is from AAAS here. Mr. Eli Kintisch was gracious enough to correct an error he made, and very quickly. He interviews Dr. Muller after the hearing, and it is well done.
This contrasts with the Salon.com reporter Andrew Leonard who not only left an error in place (conflating Willis Eschenbach with me) but refused to do anything about it, even when it was pointed out that many bloggers downstream were repeating the error without checking. Then without permission, Leonard published my complaint emails and that of Mr. Eschenbach in a second story. I’m truly disappointed in his lack of basic journalistic etiquette. I’m also disappointed that the salon.com editors have not responded at all to our early emails. Suffice it to say I won’t be talking to anyone at salon.com ever again.
I appreciate Mr. Chivers taking the time to read, understand, and present the situation in a thoughtful way.
The only thing I dispute, and it’s a minor point, is his characterization that I was blaming Professor Muller in my comment “post normal science political theater”. I’m not, and if anyone got that impression besides Mr. Chivers, I say that is why it is always best to ask. My comment is labeling of the event and the situation, not the person(s) involved. Muller was asked to testify, he didn’t go seeking it.
In fact, it may surprise many to learn that Dr. Pielke Sr. and I have been carrying on a constructive dialog with Dr. Muller via email this week. We’ve been in touch every day. Dr. Muller has shared some additional results with me, Dr. Pielke and I have pointed out what we feel are some errors, he’s countered, we are both looking at the issue. We are also both trying to understand the situation about station siting better. While it appears simple on the surface (no pun intended) it is a much more complex problem than I thought it to be when I started out. I hope to have more in a future post. For now I have more important duties, see the upcoming announcement at 3PM PST.
For another look at station siting analysis done entirely independently, I suggest this recent article on WUWT:
An investigation of USHCN station siting issues using a cleaned dataset
Mr. Gibbas (who did that study linked above) has agreed to provide more data, and in a post upcoming soon, the cleaned data he used will be made available online.
@beng – I was aware that most of the UHI effect takes place at the initial buildup around the thermometer. But that misses the point entirely. The point is that the alarmists have told us that there has been a certain amount of climate change in the US, and we skeptics have doubted the accuracy of that claim on the basis of questions about UHI effects. Now Anthony has confirmed that what they’ve told us is consistent with the data from well sited rural stations. Therefore it doesn’t matter when or how much UHI happened because Anthony found the same climate change at the verified rural stations.
@Theo Goodwin – You quote Anthony that “According to the best-sited stations, the diurnal temperature range in the lower 48 states has no century-scale trend.” That quote can be a little misleading. It’s not saying that the best stations show no climate change trend, it’s saying that the best stations show no trend in the difference(range) between the daily (diurnal) high and low temperatures. That quote is not inconsistent with his finding that the best sited stations did show a trend of both the daily high and low increasing. I don’t recall what Muller said about the trend in the diurnal range.
Anthony did did speculate that the lack of UHI effect on average temp trends was coincidental, but that doesn’t appear to have been based on the data or calculations, but rather is just a speculative guess to explain the amazing lack of UHI effect on average temp trends. Remember that even if the city trend in the US is accurate just by coincidence, that doesn’t change the fact that the alarmist reported climate change trend for the US has apparently been accurate and not contaminated by UHI, as demonstrated by its consistency with well sited rural stations.
But what about the rest of the world? Climate change is said to be considerably less in the US than in some other parts of the world, such as the arctic. Is it possible that the error of US city stations won’t coincidentally cancel out in other places? Thinking about it more I thought (or perhaps remembered) another possibility for how UHI might have been canceled in US cities. It is said that the modern electronic thermometers have better radiation shields than the old CRS boxes, and thus give a more accurate cooler temperature at the hottest part of the day. It may be possible that cities have been converted to electronic sensors faster than rural stations, and thus the warmer lows caused by UHI could be canceled by the cooler highs of the superior electronic radiation shields. But I don’t know if the electronic sensors really are more common in the city locations or if the cooler highs would happen to just cancel out the warmer lows. I would guess that the easy availability of electricity would make electronic thermometers more common in cities. Furthermore, if the same increase in electronic sensors happens in other UHI affected locations around the world, then UHI might still be unable to explain global warming, though just by the lucky cancellation of errors.
It might be however that the lack of cancellation by the superior radiation shield can explain arctic warming somewhat, because if there is UHI(or research station heat island) increasing the lows, the superior radiation shield might not cancel much of that error due to the lower influence of the sun at high latitudes. Thus the coincidental correction of UHI effect in the US might go uncorrected in arctic regions.
I’m surprised that Anthony’s study didn’t do a comparison by shield type.
Remember that we’ve only seen the abstract. Or do you have access to a fuller version?
Mindbuilder says:
April 9, 2011 at 5:28 pm
@Theo Goodwin – You quote Anthony that “According to the best-sited stations, the diurnal temperature range in the lower 48 states has no century-scale trend.”
Excuse me, but can you find ONE Warmista who does not assert that global warming must show up most strongly in higher diurnal temperatures? No diurnal trend means no global warming, at least according to Warmista accounts.
Sorry, Mindbuilder Dude, but I am not interested in playing games. In my book, you are just another Warmista and I will not respond to you in the future.
I just realized a flaw in the theory of my previous post. As pointed out by beng, most UHI effects would have occurred a long time ago, before the introduction of electronic thermometers. So if there was any UHI trend difference between urban and rural caused by the radiation shields of electronic thermometers, Anthony’s study should have picked it up from way back.
@Theo Goodwin – The part you quote doesn’t say there wasn’t a trend in daytime(diurnal) temperatures at the best sited stations, it says there was no trend in the diurnal(daily) temperature range. In other words, the difference between the daytime high and nighttime low hasn’t changed at rural sites, but has been getting smaller in cities. We can’t tell from the abstract of his study, but he may have found that there is an upward trend in both the highs and lows in both rural and city stations. I’m not familiar with claims by alarmists that the daily range of temperatures will change, though I could have easily missed them.
@barry – I don’t have a fuller version of Anthony’s article, but he stated at the link I’ve given that he didn’t control for instrument type.
The theory goes that nighttime (minimum) temps warm faster than day time (maximum) under GHG warming (the opposite is said to be the consequence of solar warming). The ‘alarmists’ say that this is seen in the global record. I don’t know if this is assumed or measured as happening on regional scales, as in the lower 48.
Barry says: http://wattsupwiththat.com/2011/04/07/some-true-journalism-my-thanks-to-tom-chivers/#comment-639409
it is indeed alleged that due to increased green house gases heat is trapped that cannot escape from earth. So if green house gases were to blame for the warming, it should be minimum temperatures (that occur during the night) that must show the increase (of modern warming). We have reasonable accurate data from weather stations all over world for the past 35 years. On studying these data, I find eactly the opposite trend! For example, I looked at the average monthly temperatures during winter here in Pretoria, when I would expect more green house gases in the air due to local veldfires:
http://letterdash.com/HenryP/assessment-of-global-warming-and-global-warming-caused-by-greenhouse-forcings-in-pretoria-south-africa
I also looked at the data from Marion Island, which lies in the southern Indian ocean:
http://www.letterdash.com/HenryP/marion-island-assessment-of-climate-change-in-the-southern-indian-ocean-due-to-the-increase-in-greenhouse-gases
It appears from these reports that warming has stalled. There has been no measurable amount of warming for the past 35 years here in South Africa (when most of the increase in carbon dioxide occurred) whereas minimum temperatures have actually declined – this is the opposite trend that you would expect if the theory of warming due to an increase in greenhouse gases were true….
Here, we have some average temperature data since 1946,
http://img502.imageshack.us/img502/8705/navacerrada.gif
This is in Spain, which I think I also pretty much average as average goes for a place to stay on earth! The station has been very accurate in its recordings and only one day’s recordings are missing. Note that the average minimum temps. since 1980 have stayed constant.
Similarly, another weather station that showed no significant increase of minimum temperatures is that of Armagh in Northern Ireland. http://www.arm.ac.uk/preprints/445.pdf (we have accurate records here going back 200 years!)
I have also looked at the data from La Paz in Bolivia and found similar results as reported for Marion Island. These observations all support the conclusion that it simply cannot be the increase in greenhouse gases that caused modern warming.
barry wrote:
That’s interesting, I hadn’t heard that. When Theo mentioned something like that, I tried to think of what effect CO2 would have on the diurnal range, and guessed it would have the opposite effect. I guessed that if CO2 effectively trapped sunlight that it would tend to raise daytime temps more. But maybe it could have more effect blocking upward radiation at night.
Thank you, Mr. Watts. The world needs you
CO2 traps no sunlight (I read some years ago about atmospheric radiative dynamics from sunlight cutting across the atmosphere, and there may be a small amount of absorption by CO2. but the details are hazy – I do remember the effect re CO2 was either very small or non-existent). CO2 absorbs long-wave (upwelling) infrared radiation, not short-wave sunlight (or hardly any). therefore the effect is more pronounced at night without solar heating. It’s also more pronounced in winter for the same reason (less insolation).
Barry says: therefore the effect is more pronounced at night without solar heating. It’s also more pronounced in winter for the same reason (less insolation).
that is the theory by those claiming that man is responsible for changing the climate. . but I could not find any proof of that. I am saying more carbon dioxide is better and the net effect of any warming and cooling by CO2 is probably zero (AS SHOWN BY SIMPLE OBSERVATION OF THE DATA)
http://www.letterdash.com/HenryP/more-carbon-dioxide-is-ok-ok
Henry,
a handful of local records amount to anecdotal evidence. Local factors may influence the relative trends. A broad sample has the virtue of canceling out the noise and revealing a truer trend (law of large numbers).
Anthony Watts’ thoughts on small sample sizes:
I don’t know the rating for the stations you’ve mentioned, but even if they are perfectly good, it would be a mistake to conclude that such a small number is representative of global or even national trends. Much more number crunching needs to be done.
As an example, according the Australian Bureau of Meteorology, the annual minimum temperature trend is greater than that for the maximum temperature trend – this follows the mainstream uderstanding of what should happen with GHG warming. Rottnest Island off Western Australia has a positive diurnal trend (max temp trend greater than min), and the same is true for Geraldton and Halls Creek. There will be others. If we based our understanding on these stations, we would come to a false apprehension. Some stations show a positive diurnal trend, some show little or no trend, but the majority show a negative diurnal range (minimum temps are warming faster than maximum daily temps). Only by processing as much data as possible can we begin to figure out what is happening generally.
This is what makes Anthony’s upcoming paper, Fall et al 2011, so exciting. For the first time, Anthony and his colleagues have a large enough sample, and have crunched the numbers. This is a significant peak in the surfacestations project – it’s what surfacestations has been leading to. I am greatly looking forward to reading the paper.
Barry, thanks. You say: a handful of local records amount to anecdotal evidence (to the scientific world)
I do agree with you. We do need to examine more stations. I was hoping and asking all those frequenting here to help by looking at the records of their own local weather stations.
But what is the chance of me picking up on a handful (namely 5) of weatherstations, situated in various places all over the world, and all showing:
1) no change in mean temperature, in other words “no local warming” or “local cooling”
2) minima going down (at a rate of ca. 0.03 degrees C per annum)
3) maxima rising (at a rate of more than 0.05 degrees C /year)
That co-incidence is too big. I therefore predict that we can take these findings of mine from local to global. So, I think I already know what is in the paper?
I should qualify the above post, by saying that I only looked at the past 35 years.
But what is the chance of me picking up on a handful (namely 5) of weatherstations, situated in various places all over the world, and all showing:
1) no change in mean temperature, in other words “no local warming” or “local cooling”
2) minima going down (at a rate of ca. 0.03 degrees C per annum)
3) maxima rising (at a rate of more than 0.05 degrees C /year)
After your first post I mined the BOM for diurnal range trends. I couldn’t find any, so I had to check the anomaly maps to find exceptions to the IPCC rule, and came up with three after a bit more searching. What is the lesson I should learn from that experience?
Answer: nothing. Coincidence is the currency of superstition. Data is the currency of science. The more you spend, the better quality your results.
When I wrote
I meant – After your first post I mined the BOM for positive diurnal range trends. I had to look hard to find examples that, like yours, ran counter to the expectations of the IPCC.
Barry
the odds against me being not correct are enormous.
Coincidence is the currency of superstition?
Life itself was impossible if it were not for (many) coincidences!
rather use the site that I used to find historical data from your country
http://www.tutiempo.net/en/Climate/datos.php?stn=689940
Click on world climate
look for the station nearby you.
Or, as I have learned, rather look for a weather station somewhere on an island.
It will probabbly give a better average of earth’s global cliamte.
Blessings you all…
Henry,
Nearest Island to where I live on the BOM’s best site list is Moreton Island.
Mean temperature since 1900 has risen by 1.2C. Minimum temps have risen faster than maximum. If this is representative of world temperatures, then the official records are too low, and the diurnal range trends support the IPCC recommendation.
Coincidentally, this matches the first five searches I did. Incredibly, this is the opposite of what happened for you.
So how do we reconcile our different research results, both obtained by coincidence?
I looked at another island further away – <a href=http://www.bom.gov.au/cgi-bin/climate/hqsites/site_data.cgi?variable=DTR&area=aus&station=098017&dtype=anom&period=annual&ave_yr=T<King Island. The mean temperature rise matches that of the globe, but there is almost no trend for the diurnal range.
The data are variable, whether from islands or the mainland. In order to find out the national or global average, ‘coincidence’ is not going to cut it. Having checked out more of the BOM best sites, preliminary results are that warming and diurnal range trends fall in line with that anticipated by the IPCC etc. But even this larger sample isn’t enough. More data is better.
Barry
I would not trust those “made up graphs” going back to 1900 that you keep referring to me. Who made those graphs?
Can you bring me a calibrated certificate of a thermometer that is so old? How often was it read (by a person) to come to the “mean” for the day (24 hours)?
We have now electric thermometers (thermocouples) that continuously measure and record the temperature and give an average at the end of the day. So, the accuracy from before 35 years is definitely in question.
You could start here here;
http://www.tutiempo.net/en/Climate/Norfolk_Island_Airport/07-1976/949960.htm
but start looking from 1976. From then onwards I think we are reasonably sure about the accuracy and frequency of measurements. Follow the same procedure as I did for Marion Island. See what you get?
Henry, it’s this simple – you cannot derive a meaningful analysis of global temperature trends from one or a handful of weather stations. It is also the view, as I quoted upthread, of the owner of this website. It is exactly because coincidences like the one you have experienced occur that scientists and statisticians do not rely on small data samples. Small samples can easily mislead. We don’t expect uniform trends all over the globe. Some places may cool or exhibit trends different from global (or national).
I’m curious, however, about the provenance of the site you have directed me to. I am on tour and not in a position to do a long mining exercise there, but I would be intrigued to see what the data showed for Moreton Island on that website. Should you be willing to assess, let’s agree on a start year before analysis to obviate cherry-picking. How about 1976, as per Norfolk Is?
The data and graphs are compiled at the Australian Bureau of Meteorology, the national repository for Australian weather and climate data.
Did you not click on even one link? The information you asked for is right at the top of the page.
Barry says: The data and graphs are compiled at the Australian Bureau of Meteorology, the national repository for Australian weather and climate data.
yes, but who was involved in making those data from 1900?
What equipment/method/manipulations etc. were done to go back that far ?(look at the Armagh site quoted in my blog where these specifics are given.)
Moreton is not listed on my preferred site….strangely enough. Brisbane airport is listed, and I would be inclined to trust those data more than Moreton’s (because of the presence of “the climate change centre” there )
and then only from 1976, as suggested.
There are three listings for Brisbane airport. Which one to take?
Remember, this is only my hobby, so I it may take me also quite some time to do an analysis on an Australian station..
Henry, Australia’s data is relatively good, but like everywhere else, not perfect by any means. I scoured the BOM quite some time ago, as well as mined google for studies and metadata on the instrumental records. Sir Joseph Banks began a legacy of great interest in the natural phenomena of the ‘new world’. I can’t speak to the quality of particular stations, and currently am not in a good position to do much research on the metadata. We have a challenging climate here – lots of drought and flood – and accuracy in records and forecasting is vital. Many records stretch back to 1900. The ones that don’t, or where the data is patchy in the past, the BOM will only give records from the year the data is solid.
If you decide to have a look at Brisbane airport, I will stick here with you – I’d see it as very sporting of you to check it out. Norfolk Island data at the BOM will likely be hard to come by, as it’s not listed as one of the ‘best sites’. However, if there are any regulars here attending to our discussion, they’ll probably roll their eyes at the suggestion of using airport temps. 🙂
(I do believe we’re having a reasonable discussion – delighted!)
Ok. Brisbane it will be.
Historical Weather: Brisbane Airport M. O, Australia
Weather station: 945780 (YBBN)
Latitude: -27.38
Longitude: 153.1
Altitude: 4
I will check from 1976
to cover the past 35 years
How much was the bet for?