Watts et al paper 2nd discussion thread

The first press release announcement thread is getting big and unwieldy, and some commenters can’t finish loading the thread, so I’m providing this one with some updates.

1. Thanks to everyone who has provided widespread review of our draft paper. There have been hundreds of suggestions and corrections, and for that I am very grateful.  That’s exactly what we hoped for, and can only make the paper better.

Edits are being made based on many of those suggestions. I’ll post up a revised draft in the next day.

2. Some valid criticisms have been made related to the issue of the TOBS data. This is a preliminary set of data, with corrections added for the “Time of Observation” which can in some cases result in double max-min readings being counted if not corrected for. It makes up a significant portion of adjustments prior to homogenization adjustments as seen below in this older USHCN1 graphic. TOBS is the black dotted line.

TOBS is a controversial adjustment. Proponents of the TOBS adjustment (Created by NCDC director Tom Karl) say that it is a necessary adjustment that fixes a known problem, others suggest that it is an overkill adjustment, that solves small problems but creates an even larger one. For example, from a recent post on Lucia’s by Zeke Hausfather, you can see how much adjustments go into the final product.

The question is: are these valid adjustments? Zeke seems to think so, but others do not.  Personally I think TOBS is a sledgehammer used to pound in a tack. This looks like a good time to settle the question once and for all.

Steve McIntyre is working through the TOBS entanglement with the station siting issue, saying “There is a confounding interaction with TOBS that needs to be allowed for…”, which is what Judith Curry might describe as a “wicked problem”. Steve has an older post on it here which can be a primer for learning about it.

The TOBS issue is one that may or may not make a difference in the final outcome of the Watts et al 2012 draft paper and it’s conclusions, but we asked for input, and that was one of the issues that stood out as a valid concern. We have to work through it to find out for sure. Dr. John Christy dealt with TOBS issues in his paper covered on WUWT: Christy on irrigation and regional temperature effects

Irrigation most likely to blame for Central California warming

A two-year study of San Joaquin Valley nights found that summer nighttime low temperatures in six counties of California’s Central Valley climbed about 5.5 degrees Fahrenheit (approximately 3.0 C) between 1910 and 2003. The study’s results will be published in the “Journal of Climate.”

Most interestingly, John Christy tells me that he had quite a time with having to “de-bias” data for his study, requiring looking at original observer reports and hand keying in data.

We have some other ideas. And of course new ideas on the TOBS issue are welcome too.

In other news, Dr. John Christy will be presenting at the Senate EPW hearing tomorrow, for which we hope to provide a live feed. Word is that Dr. Richard Muller will not be presenting.

Again, my thanks to everyone for all the ideas, help, and support!

=============================================================

UPDATE: elevated from a comment I made on the thread – Anthony

Why I don’t think much of TOBS adjustments

Nick Stoke’s explanation follows the official explanation, but from my travels to COOP stations, I met a lot of volunteers who mentioned that with the advent of MMTS, which has a memory, they tended not to worry much about the reading time as being at the station at a specific time every day was often inconvenient.. With the advent of the successor display to the MMTS unit, the LCD display based Nimbus, which has memory for up to 35 days (see spec sheet here http://www.srh.noaa.gov/srh/dad/coop/nimbus-spec.pdf) they stopped worrying about daily readings and simply filled them in at the end of the month by stepping through the display.

From the manual http://www.srh.noaa.gov/srh/dad/coop/nimbusmanual.pdf

Daily maximum and minimum temperatures:

· Memory switch and [Max/Min Recall] button give daily

highs and lows and their times

The Nimbus thermometer remembers the highs and lows for

the last 35 days and also records the times they occurred. This

information is retrieved sequentially day by day. The reading

of the 35 daily max/min values and the times of occurrence (as

opposed to the “global” max/min) are initiated by moving the

[Memory] switch to the left [On].

So, people being people, rather than being tied to the device, they tend to do it at their leisure if given the opportunity. One fellow told me (who had a Winneabago parked in is driveway) when I asked if he traveled much, he said he “travels a lot more now”. He had both the CRS and MMTS/Nimbus in his back yard. He said he traveled more now thanks to the memory on the Nimbus unit. I asked what he did before that, when all he had was the CRS and he said that “I’d get the temperatures out of the newspaper for each day”.

Granted, not all COOP volunteers were like this, and some were pretty tight lipped. Many were dedicated to the job. But human nature being what it is, what would you rather do? Stay at home and wait for temperature readings or take the car/Winnebago and visit the grand-kids? Who needs the MMTS ball and chain now that it has a memory?

I also noticed many observers now with consumer grade weather stations, with indoor readouts. A few of them put the weather station sensors on the CRS or very near it. Why go out in the rain/cold/snow to read the mercury thermometer when the memory of the weather station can do it for you.

My point is that actual times of observation may very well be all over the map. There’s no incentive for the COOP observer to do it at exactly the same time every day when they can just as easily do it however they want. They aren’t paid, and often don’t get any support from the local NWS office for months or years at a time. One woman begged me to talk to the local NWS office to see about getting a new thermometer mount for her max/min thermometer, since it wouldn’t lock into position properly and often would screw up the daily readings when it spun loose and reset the little iron pegs in the capillary tube.

Some local NWS personnel I talked to called the MMTS the “Mickey Mouse Temperature System” obviously a term of derision. Wonder why?

So my point in all this is that NWS/NOAA/NCDC is getting exactly what they paid for. And my view of the network is that it is filled with such randomness.

Nick Stokes and people like him who preach to us from on high, never leaving their government office to actually get out and talk to people doing the measurements, seem to think the algorithms devised and implemented from behind a desk overcome human urges to sleep in, visit the grand-kids, go out to dinner and get the reading later, or take a trip.

Reality is far different. I didn’t record these things on my survey forms when I did many of the surveys in 2007/2008/2009 because I didn’t want to embarrass observers. We already had NOAA going behind me and closing stations that were obscenely sited that appeared on WUWT, and the NCDC had already shut down the MMS database once citing “privacy concerns” which I ripped them a new one on when I pointed out they published pictures of observers at their homes standing in front of their stations, with their names on it. For example: http://www.nws.noaa.gov/om/coop/newsletters/07may-coop.pdf

So I think the USHCN network is a mess, and TOBS adjustments are a big hammer that misses the mark based on human behavior for filling out forms and times they can’t predict. There’s no “enforcer” that will show up from NOAA/NWS if you fudge the form. None of these people at NCDC get out in the field, but prefer to create algorithms from behind the desk. My view is that you can’t model reality if you don’t experience it, and they have no hands on experience nor clue in my view.

More to come…

0 0 vote
Article Rating
378 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
July 31, 2012 5:36 pm

I have some additional thoughts up here: http://rankexploits.com/musings/2012/initial-thoughts-on-the-watts-et-al-draft/
Its intended as constructive criticism to help improve the paper, not an attack on Anthony or anything of that sort, so be civil.
[snip – Zeke we already covered this issue in email – Anthony]

July 31, 2012 5:38 pm

Climatists’ have been, “locked into a simple-minded identification of climate with greenhouse-gas level. … That climate should be the function of a single parameter (like CO2) has always seemed implausible. Yet an obsessive focus on such an obvious oversimplification has likely set back progress by decades,” (Richard Lindzen, July 2012)
Facts are facts: the surface temperatures of the Earth constantly fluctuate. A supposed rise of temperatures of 0.8 °C in 150 years is just too small to worry about. This study analyzes the surface record as reported by the NOAA and shows half of the global warming in the USA is artificial. In addition to recording temperatures at poorly-sited stations located at airports and in urbanized areas that will of course show increases in temperature trends, Well sited rural, non-airport stations show a warming nearly three times greater after NOAA adjustment is applied.

July 31, 2012 5:42 pm

For folks interested in the details of the adjustments NCDC does, this presentation by Menne at the 2011 AMS conference is a good accessible introduction: https://ams.confex.com/ams/19Applied/flvgateway.cgi/id/18288?recordingid=18288

Steve Huntwork
July 31, 2012 5:46 pm

TOBS has always been a dating issue and never a temperature offset.
Think of a Sin wave being sampled by two observers 1/3 out of phase. Both will correctly record a +1.0 maximum and a -1.0 minimum over their periods of data recording.
To “adjust” one set of valid observations with a set temperature offset, would be dumb beyond belief.
Except for very rare occasions, the maximum temperature will occur around 2:00 PM local time and the minimum around local sunrise.
For after midnight recordings, both of the maximum and minimum temperatures will be dated for the previous day.
For post-sunrise recordings, the maximum temperature occurred on the previous day and the minimum was recorded on the CURRENT day.
This has always been a timing phase offset and never an error with the actual recorded data. Those minimum and maximum temperatures actually occurred.

Christoph Dollis
July 31, 2012 5:46 pm

Thanks for this post and explanation of differing viewpoints on TOBS adjustments.
If I hadn’t made the point clearly in the other thread, I’ll reiterate it in paired down form here: I appreciate and respect both our host, Anthony Watts, and Steve McIntyre precisely for their airing out differences publicly and honestly. This is how it should be done, in my opinion.
McIntyre didn’t say definitively that he will sign on as a co-author in the final version to be presented for publication. I hope he does after properly weighing in on the paper.
REPLY: Steve contributed greatly to the first effort, creating a whole section, not naming him would have been a travesty. It will be his choice for the final though. – Anthony

July 31, 2012 5:47 pm

Anthony, when will the list of station ids used in this paper and their classifications be made available? And maybe some detail about how the classification was made?
REPLY: Hello Ron,
We want to at least get the paper into submission form first, so that we can preserve the years of work that Evan and I and the many volunteers put into it by being able to submit first. Bear in mind that gathering this data caused out of pocket expenses and much time away from my family, plus my openness has been abused in the past. Both Menne et al and Muller/BEST submitted papers before we even got ours out the door. I don’t think it is unreasonable to get my submission in first for once.
That said, I’m intending to provide it to allow for full replication as I did to Dr. Muller and when Fall et al 2011 was published.
We do have description on how the classifications were done in the PowerPoints as well as the paper. – Anthony

William Roberts
July 31, 2012 5:47 pm

Hey look at that, all suggestions for robust trend analysis are completely ignored!
REPLY: No that’s an assumption on your part. We’ve looked at everything in the thread, including yours. Suggestion: you’ll get further in this thread by being more cordial. People who also use their real name rather than lie with the first comment made by using a fake name, also tend to be taken more seriously. Thanks for your consideration. – Anthony

Jim Steele
July 31, 2012 5:51 pm

TOBS adjustments always confused me. If a thermometer measures the maximum and minimum, why would TOBS affect a multidecadal trend? I had the impression the TOBS could alter a monthly or annual average because the depending on the Time of Observation the temperature could carry over into the next time period. But why would that affect a trend?

Christoph Dollis
July 31, 2012 5:53 pm

“TOBS has always been a dating issue and never a temperature offset.
“Think of a Sin wave being sampled by two observers 1/3 out of phase. Both will correctly record a +1.0 maximum and a -1.0 minimum over their periods of data recording.”

Ummm, the difference is the temperature doesn’t fall and rise with the regularity of a Sine wave.

Geo
July 31, 2012 5:53 pm

Thanks for all the work. You are correct, the pre-posting of this has been a remarkable success, and will assist in carving the way in how science can be reviewed by ALL!! I have one question….do you feel at all “bothered” by “tiresome” distracting questions from those who “only want to find fault with your work..”? ;P
REPLY: No of course not, if I was I’d be ignoring suggestions like the fact that TOBS needs to be addressed. So we are. I don’t mind valid criticisms of the work. Small minded people who want to tear me down personally or speak in third person lagomorphic languages do become tiresome though. – Anthony

Christoph Dollis
July 31, 2012 6:00 pm

“You are correct, the pre-posting of this has been a remarkable success ….”

I think he jumped the gun by a week myself.

Steve Huntwork
July 31, 2012 6:01 pm

“Ummm, the difference is the temperature doesn’t fall and rise with the regularity of a Sine wave.”
The phase of the sampling periods is what I was trying to express…
A difference in sample timing phases will never be a linear offset.
A very simple concept if you think about it…

July 31, 2012 6:03 pm

I finally found a link that explained TOBS in a way that made sense to me. Apparently bias is only introduced if the time of day that the measurement is made changes. Am I to assume that NOAA actually knows this information and can correct for it in a meaningful way?
http://climateaudit101.wikispot.org/Time_of_Observation_Bias

Andrew
July 31, 2012 6:04 pm

Anthony,
You deserve credit for putting your money where your mouth is!
Not everyone has the cojones to open yourself (and et al.) up to the public criticism for your work. Leading with your chin hurts sometimes, but I am glad you are willing to do so!

Christoph Dollis
July 31, 2012 6:06 pm

The phase of the sampling periods is what I was trying to express…

Sure, but also the “observers” to use your analogy (or varying measuring times) aren’t going to be equidistant apart either.
The point is I just don’t know the magnitude of difference TOBS makes, but I know it isn’t likely to be a neat zero.

July 31, 2012 6:12 pm

Think of TOBs this way:
Suppose for 5 years you take one temperature measurement a day at sunset. At that point you change the observation time to morning and for the next 5 years take daily measurements at sunrise.
If you take the measurements, turn them into monthly means, and calculate the trend over 10 years you will get rather biased results.

Kelly Haughton
July 31, 2012 6:18 pm

It will be interesting to see what the “peers” do for the peer-reviewed journal. How will the peers be selected? What will they say? Will they read all the comments?

Steve Huntwork
July 31, 2012 6:24 pm

“Sure, but also the “observers” to use your analogy (or varying measuring times) area’t going to be equidistant apart either.”
…………………..
Having been one of those trained monkeys that obtained the raw weather data at each and every hour, every day of the year, so I kinda know how things were done.
For some weather stations, the observer would record the maximum and minimum temperatures over the last 24 hours with those thermometers. He is human and was not willing to do this stuff at midnight every day. So, for human observers, the standard was around 8:00 AM each morning when there was enough light to even find the darn shelter.
For those of us who did this for a living, the observations were conducted every hour and summary reports were sent out to the World every six hours. The next morning, another team would perform “quality control” and insure that what we issued wa absolutely correct. And if not, then a message correcting the errors would be transmitted.

davidmhoffer
July 31, 2012 6:26 pm

Entropic Man;
I see some papers to make one wince, too. Nevertheless, these are the ground rules that scientists and scientific journals go by.
>>>>>>>>>>>>>>>>>
It is indeed sad that otherwise intelligent people cling to the past that they are comfortable with. Odder still that someone would first promote peer review as the means to filter out bad science, and in the very next comment, admit that it doesn’t achieve that goal.
The fact of the matter is that there is a huge body of science that is published without the benefit of peer review nearly every day. The paper’s are called patents. Peer review journals do not and never did have any lock on the progress of science, the notion that they do is falsified every single day at patent offices all over the world. As for that portion of science that does progress through the journal system, I’m afraid you are stuck in the past. The world is changing, the internet permits for an exchange of information at a speed and cost that journals in their traditional format can never compete with. The only thing keeping them alive is the kind of attachment to the past the compels me to pick up an actualy newspaper from time to time to enoy along with a cup of coffee, despite my knowledge that the content was obsolete before it was printed, and current content from the same reporters is available instantly for a fraction of the cost on my desk top computer.
Will there be more crap on the internet because it is accessible to all? Of course. Will a tiny cabal of politicaly motivated scientists be able to get editors fired and legitimate science suppressed?
Not a chance. Welcome to the new world.

July 31, 2012 6:29 pm

That TOBS is a PITA.
How many sites are there like this?comment image
There was 4 interpolated monthly data points, not in a CRS but a larger well ventilated enclosure, liquid in glass with 7 a.m and 7 p.m readings since day one, did have 4 moves to 4/10 of a mile from its original location.
http://climatetrends.colostate.edu/map?station=53005

Christoph Dollis
July 31, 2012 6:30 pm

Cool Steve, thanks for the lowdown.
However, as Zeke points out, even if the recording times are tightly controlled, when they change, that’s going to skew things, so it should be accounted for in a perfect world. How much of a difference it makes to the end product I can’t say and will leave that for McIntyre and others to elucidate further.

Rob MW
July 31, 2012 6:33 pm

Anthony,
I know you don’t have time for this question but I’ll put it any way, just in case.
If the paper was to somehow use a completely unadjusted TOB, would this then shift the focus of the paper into the area of “Unintended Consequences” (i.e. – debunking other papers in this area) and away from your et al. results that are achieved using contentious yet perpetuated TOB’s ??

A. Scott
July 31, 2012 6:33 pm

Ron Broberg says:
July 31, 2012 at 5:47 pm
Anthony, when will the list of station ids used in this paper and their classifications be made available? And maybe some detail about how the classification was made?

I believe I have read a comment that this is being “addressed.” My inference is the information will be provided to select folks who can properly review it.
From my limited non-technical understanding the data is readily available publicly with the exception of Anthony’s siting results using Leroy 2010. This includes the raw and adjusted temp data along with the Leroy 2010 rating spec’s, which would allow anyone to do their own duplication of the work.
To me that seems preferable here – anyone attempting to duplicate should start from the beginning – rather then working backward from the conclusions.
The paper notes they applied the readily available specs of Leroy 2010 to the Fall 2010 USHCNv2 data set.
They identify the data they use: “We make use of the subset of USHCNv2 metadata from stations whose sites have been classified by Watts (2009)”. and “site rating metadata from Fall et al (2011)”. They further narrow: “Because some stations used in Fall et al. (2011) and Muller et al. (2012) suffered from a lack of the necessary supporting photography and/or measurement required to apply the Leroy (2010) rating system, or had undergone recent station moves, there is in a smaller set of station rating metadata (779 stations) than used in Fall et al (2011) and Muller et al. (2012), both of which used the data set containing 1007 rated stations.”. Steven Mosher and Zeke one would assume would have access to this station data as Watts used the same data as Muller 2012 in this regard.
They gave a description of the data used, their methods – how they calculated numbers – and presented their conclusions.
To me it would be much more relevant and useful it would seem, for those interested in replicating to follow the entire process – and see how their siting category counts came out. And only THEM compare to Watts conclusions.
I would also be interested in seeing how the USCRN stations, which were designed per Leroy 1999 (“which was designed for site pre-selection, rather than retroactive siting evaluation and classification”), fare under a review using Leroy 2010.
Watts 2012 notes “Many USHCNv2 stations which were previously rated with the methods employed in Leroy (1999) were subsequently rated differently when the Leroy (2010) method was applied in this study”…
Again, it would be very interesting, and potentially valuable, to see if the new USCRN sees the same siting quality results using Leroy 2010.

A. Scott
July 31, 2012 6:35 pm

One more TYPO:
299 As in Fall, et al (2011), Menne (2010), and Muller (2012), only the heat source/sink
300 proximity and area ratings from Leroy 2010 and are do consider ground-level vegetation
301 or shade.
I suspect from the paragraph following that should say we only consider heat sink/source under Leroy 2010, and not shade or vegetation ….

Brad
July 31, 2012 6:43 pm

Why not just offer a TOBS and non-TOBS graph with a discussion?

Jim Steele
July 31, 2012 6:45 pm

Zeke Hausfather says: Think of TOBs this way:
Suppose for 5 years you take one temperature measurement a day at sunset. At that point you change the observation time to morning and for the next 5 years take daily measurements at sunrise.
That is not a good analogy Zeke.
That assumes the maximum and minimum is dependent only on whatever time the observation was made. I used minimum/maximum thermometers for 30 years. It didn’t matter when you read it. If the high was 70 F, it left the marker at that point as the temperatures then retreated. The only observer error that could happen, occurred if the observation was not done daily or if it was not made before the next maximum was reached.

Brad
July 31, 2012 6:46 pm

Also, why isnt it possible to gets TOBS to be more exact? The sites should have evening temps even if morning temps are used in the database.

Steve Huntwork
July 31, 2012 6:46 pm

Anthony, I have always supported your data quality control efforts. How anyone could object to efforts to insure quality control is beyond my comprehension.
What has happened between Steve McIntyre and you is none of my business. That is something that the two of you must work out.
But please, keep honest!

L.
July 31, 2012 6:51 pm

Sorry if this has been discussed earlier, but has the Watt’s paper been dicussed on the ‘warmist’ sites yet..?
I had a look on desmog and realclimate, but couldn’t find anything.
I would be interetsed in a critique by people more qualified than myself.

Bill Illis
July 31, 2012 6:52 pm

TOBS is a relatively newly discovered problem.
It was only discovered about 200 years ago.
The former Weather Bureau of the United States was already adjusting for it 1909.
http://babel.hathitrust.org/cgi/pt?id=uva.x030228867
110 years later, we are still making adjustments to temperatures of just 5 years ago apparently. Obviously, the dedicated weather-people of the United States did not know what they doing (in 1873 or in 2007).

Phil
July 31, 2012 6:56 pm

Jason Samenow has rounded up early criticism/concern. Links below:
Josh Halpern/Eli Rabett
http://rabett.blogspot.com/2012/07/bunny-bait.html
Steve McIntyre
http://climateaudit.org/2012/07/31/surface-stations/

Gerry Parker
July 31, 2012 6:58 pm

Anthony,
I agree with what you’ve done so far with regards to the stations, but there are additional points to consider. The entire premise by the team that an unknown signal mixed with an unknown noise can be descrambled by averaging, gridding, applying (seemingly) arbitrary offsets is beyond me. I have worked with some pretty sharp DSP algorithm folks descrambling noisy signals, and from my experience, there are too many unknowns for their approach to work. In fact, from my experience it is snake oil of the first degree.
Also, the mechanics of measuring the UHI of a city is something that is relatively trivial which I believe you have done yourself (recognizing it requires a lot of labor and some technology), so that is something that could factually be determined for many urban areas. The sectioning measurement could better determine the UHI affect on some of the stations by sectioning near the stations (although the ability to capture the dynamic nature would be limited since it would be one or more snapshots in time). This could be used to show that for X sections in a year, the average UHI near station Y was measured to be (some value). That seems much less arbitrary and based on measured data.
Gerry Parker

July 31, 2012 6:59 pm

Zeke Hausfather says:

Suppose for 5 years you take one temperature measurement a day at sunset. At that point you change the observation time to morning and for the next 5 years take daily measurements at sunrise.
If you take the measurements, turn them into monthly means, and calculate the trend over 10 years you will get rather biased results.

You have got to be kidding!
Thanks to a little algebra, a monthly mean can be viewed as
MonthlyMean(j) = ((Sum of Max(i))-(Sum of Min(i)))/2, i = days(i) of month j
If I change the time of Observation from evening to morning, The most that happens is that I replace the High (from the last day of one month) with the High from the last day of the next month. And over the year, that will even out. For a 10 year trend you ought to see no effect at all.
And that assumes you have good meta data on the station and that the highs and lows were not already corrected in the record.
Ok. Maybe if you need the daily Max and Min to get a daily mean, to do something by day, you might need to sync the high and low values. But you don’t need to do this for Monthly means and multiyear trends.

July 31, 2012 6:59 pm

A. Scott: To me it would be much more relevant and useful it would seem, for those interested in replicating to follow the entire process – and see how their siting category counts came out. And only THEM compare to Watts conclusions.
A., replication (aka “auditing”) requires the original data set. If Anthony feels it is appropriate to hold the data in private, away from public view, at least until publication, then the next step is to attempt to reproduce it by following his methods. Anthony explains above that he believes there is sufficient information in to reproduce their Leroy 2010 classification for USHCN stations. “We can, however, quite accurately determine heat sink coverage by use of satellite and aerial imagery and in particular, that of Google Earth aerial photography and its distance measurement tool.” That certainly tells us where the data came from, but only hints at how it was processed.
So neither Watt’s data nor his classification methods are currently available for public review. It’s important to understand the limits of this blog review. Nevertheless, I believe it could be possible to develop an independent classification per Leroy 2010 of USHCN stations.
REPLY: if you want to do that, since you are in Europe, why not start with GHCN instead of USHCN and contribute something useful and new as I did? There’s a whole mess of stations in GHCN just waiting to be rated right in your own back yard. – Anthony

scp
July 31, 2012 7:01 pm

A thought, and maybe it’s wrong, but I’ll toss it out there – Instead of adjusting for TOBS, is there no way to slide the window so the day starts and stops at a different time? i.e. Offset the calendar by a number of hours to some different time zone in order to avoid the need for an adjustment?

Steve Huntwork
July 31, 2012 7:15 pm

@scp says:
Now you are starting to understand what I was trying to say about the TOBS adjustments.
Lable the Y axis as temperature and the X axis as time.
The TOBS adjustments are along the X axis.
Amazingly simple if you think about it.
Slide the time window and the temperatures will match. It is that simple!

July 31, 2012 7:18 pm

Europe? I guess even your penchant for revealing personal information of commentators is fallible! 😆
REPLY: Ron, Someone told me you were based in Germany, I took them at their word. So where are you based?
The suggestion still stands, there’s a need to get GHCN metadata/ratings. Apparently there was a group that was going to get around to doing it, but they have postponed to 2015. BTW my name is Anthony since you have trouble with that.
– Anthony

Eric Barnes
July 31, 2012 7:18 pm

Reading the Karl TOBS paper here ftp://ftp.ncdc.noaa.gov/pub/data/ushcn/v2/monthly/karl-etal1986.pdf it would seem to me that you could experimentally determine TOBS bias for any site(s) that also had hourly data (the airports im assuming) based on an hour that you choose to reset the thermometer. IMO this would be a fairly easy way to bound the magnitude of the problem. Hasn’t anyone done that? Has anyone seen the tobs fortran code? I haven’t been able to find it.
Algorithm.
Use Hourly data to get min/max (reset at midnight).
Calculate max, min and mean values +trend.
Use hourly data but reset at 5pm say.
Calculate max, min and mean values +trend.
compare steps 2 and 4.

wayne
July 31, 2012 7:19 pm

Christoph Dollis
Ummm, the difference is the temperature doesn’t fall and rise with the regularity of a Sine wave.

Christopher, a minimum is a minimum and a maximum is a maximum, please tell us all why you keep drawing time into the picture concerning this paper. Now if you are speaking of ~WWII changeover from single readings in the afternoon to single COOP readings at specific times in the morning I would see your point. This paper however is the 1979-2008 span. I would be surprised if any of the older single time readings even plays into this time period of records. See?

July 31, 2012 7:29 pm

If you were to get maximum TOBS bias, you would read at the maximum temperature at about 3 in the afternoon. Then you would get a double high reading on a cooling trend. Unless the time of reading was changed often, you would just have one double reading and the rest a day off. Generally, people are a little curious about what the actual high temperature of the day might be. If I were reading the temperature, I might be incline to make sure I didn’t have the wrong temperature for the wrong day. The neighbors might start calling me an idiot or skeptic or something. I am sure that climate scientists considered the human element. Didn’t TOBS adjustments start when the geniuses realized the MMTS and the other automated system gave false readings? Oh, wait, they believed the high tech stuff first didn’t they?

MrX
July 31, 2012 7:30 pm

Do any current thermometers keep monitoring temperature all day long and can I download this data anywhere? What sampling frequency is available? If anyone has a link to a handful of station data like this (preferably in Canada or the US), could you please post it? I’d like to do some simple analysis and also apply a possibly new algorithm for data that is cyclical (basically, it can detect shifts between the patterns, and/or give you some error analysis on slightly noisy data compared to a baseline or aggregate set of cyclical data, or even compare two sets of data)

Steve Huntwork
July 31, 2012 7:31 pm

TOBS adjustments are kinda like dating an unknown sample of wood by matching the tree rings with known historical data to identify it’s date.
Archaologists do this all the time…

Eric Barnes
July 31, 2012 7:36 pm

MrX says:
July 31, 2012 at 7:30 pm
Hourly data is here …
ftp://ftp.ncdc.noaa.gov/pub/data/noaa/

James Humbolt
July 31, 2012 7:37 pm

[SNIP: James, I’m not going to get into this. I’ve had it with rude, hostile, superior, snarky and sneaky commenters. Sometimes you need to hit them with a two-by-four to get their attention. -REP]

Steve Huntwork
July 31, 2012 7:39 pm

MrX:
You asked: “Do any current thermometers keep monitoring temperature all day long..”
There are so many sources of max / min thermometers available on the market today, that I could not hope to list all of them.
However, see if this link will help to answer your question:
http://www.bing.com/images/search?q=max+min+thermometers&qpvt=max+min+thermometers&FORM=IGRE

July 31, 2012 7:39 pm

From the article:
In other news, Dr. John Christy will be presenting at the Senate EPW hearing tomorrow, for which we hope to provide a live feed. Word is that Dr. Richard Muller will not be presenting.
Muller is making the rounds. This afternoon he was a live guest on a.m. radio 910 in the San Francisco Bay area. The blurb I caught said: “Richard Muller, who was a skeptic funded by the Koch Brothers, has reversed his views…” [or something very close].
Muller the chameleon is still pretending that he used to be an AGW skeptic. I would call him a self-serving opportunist.

Rob MW
July 31, 2012 7:46 pm

It seems to me that Anthony et al. will be damned by the climate science community if he uses unadjusted TOB’s, or damned by the skeptic community if he doesn’t.
Toss a coin Anthony and the rules are; heads you lose and tails you lose !!!

Paul Westhaver
July 31, 2012 7:47 pm

much better-er… I like the graph. My brain responds to this form of representation.

Gunga Din
July 31, 2012 7:48 pm

TOBs, Time of Observation biases. That was one of those terms I wasn’t familiar with and didn’t know what difference it might make until I followed the link above, “Steve has an older post on it HERE which can be a primer for learning about it.”
We report our precipitation to the NWS. One day a few years ago I took a call from them questioning our monthly report. We were taking our reading at about 6 AM and put it down under the previous days total. (Midnight was a very busy time for us, especially before we had any automation.) That was usually not a problem. The call was because there had been a heavy downpour between midnight and 6 AM on the first day of the month but nothing after 6 AM. Our report showed “0.0” precipitation while everyone else had a couple of inches at least. The time we took our reading had no effect on total for those two months but it did effect which of the two months it fell in.
I hope that might help someone understand what everyone is talking about when “TOBs” comes up.
(I’m still working on why it might matter that much to the paper.)

July 31, 2012 7:51 pm

MrX says:
July 31, 2012 at 7:30 pm
“Do any current thermometers keep monitoring temperature all day long and can I download this data anywhere?”
Mr. X, click the icon on the right sidebar that says “Monitor your own weather”. Anthony doesn’t really push it, but his company sells numerous weather related instruments, including inexpensive usb dataloggers that can be set to record temps from very short to very long intervals.
I worked in a facility where almost every room had chart recorders for temp & humidity. The recorders and circular paper charts could be set for 1 day, 1 week, 1 month, etc. I am retired now, but I am sure there are models available. Ours were made by Honeywell. But Anthony’s are less expensive, and the versatility, specs & accuracy look to be even better than the paper chart recorders.

James Humbolt
July 31, 2012 7:55 pm

[SNIP: Once again, this is not up for discussion. A first time-commenter pursuing this tack looks and sounds much like a “concern troll”. -REP]

Steve Huntwork
July 31, 2012 7:55 pm

Sorry Anthony, but every military weather observer / forcaster attended the same school and we all knew each other. Army, Air Force, Navy and Marine – we all trained together.
We do get a little upset when “scientists” years later try to tell us that our records were wrong and must be “adjusted.”
We were there and took those measurements every hour and fully understood how strict the quality control was. In the military, if you make an error, someone may die because of it.

A. Scott
July 31, 2012 7:55 pm

Ron Broberg says:
July 31, 2012 at 6:59 pm

A. Scott: To me it would be much more relevant and useful it would seem, for those interested in replicating to follow the entire process – and see how their siting category counts came out. And only THEM compare to Watts conclusions.

A., replication (aka “auditing”) requires the original data set. If Anthony feels it is appropriate to hold the data in private, away from public view, at least until publication, then the next step is to attempt to reproduce it by following his methods. Anthony explains above that he believes there is sufficient information in to reproduce their Leroy 2010 classification for USHCN stations. “We can, however, quite accurately determine heat sink coverage by use of satellite and aerial imagery and in particular, that of Google Earth aerial photography and its distance measurement tool.” That certainly tells us where the data came from, but only hints at how it was processed.
So neither Watt’s data nor his classification methods are currently available for public review. It’s important to understand the limits of this blog review. Nevertheless, I believe it could be possible to develop an independent classification per Leroy 2010 of USHCN stations.

Semantics. Replicate/reproduce … ‘replicate – reproduce or make an exact copy of …’
The source data is well identified, and all available already as far as I can see. The key issue in question would be the application of the Leroy 2010 standard to that data and the results thereof.
The Leroy 2010 standard seems well identified. An attempt to reproduce the Watts results would have far more value than trying to replicate his exact results. Knowing the answer ahead of time adds a strong bias in my non-technical opinion.
This does not seem to be hard work. Grab the readily available identified data set – read the Watts report and follow their process – apply Leroy 2010 to the data and see what you come up with for siting bins. Run the calcs from those and see if they match Watts 2012.
If they do commend him. If they don’t THEN ask him to work with you to identify the differences. This will only strengthen the Leroy 2010 standard and its application.
Add a TOB’s review on top and see where that leads as well.
The difficult work is reviewing the sites. If I recall there were 779 to start with – they are your beginning station data. Watts et al spent over a year completing that work. I can well understand his reluctance to put that in public domain before the paper is ready for submission.
It is that work – how the Leroy 2010 is applied – that is important here. It is that work that should be replicated or reproduced. And having the answers beforehand would not seem to make a lot of sense here – at least to me.

July 31, 2012 7:59 pm

If the tennor of the comments on slashdot is any indication, this paper has the AGW croud totally freaked out!
Secondly I can see tobs being adjusted to 2 PM for the expected high temperature, but how do you adjust for the Tmin expected at just before sunrise? It shouldn’t be to much of an issue in the tropics, but in high latitudes, sunrise changes quite a bit, high enough it can take weeks.

Bill Yarber
July 31, 2012 7:59 pm

And observation from today and related questions for Zeke:
As I left my club today about 1:45pm, the temperature reading in my car was 95F. I check this reading regularly against the time/temp displays all around town and I’m normally +/- 1F. My drive home of 7 miles (only 4 mile south of starting point) takes 15 minutes this time of year. The temperature at my condo complex was 86F, or 9F less. My condo is only 1,500 feet from the ocean while my club is about a mile from the ocean. Also, it was a cloudless sky at my club and the cloud cover was nearly 50% (approx) by the time I got home. But those two locations were only 4 miles apart on a direct line.
How would your TOBS algorithm adjust those two readings taken only 15 minutes apart?
How can you use area spacings of up to 1,200 square meters to approximate the temperature of the Earth (or even the USA) when the temperature can vary by 9F over a four mile distance?
I have the feeling that you know your algorithms and their functionality extremely well, but you’ve lost site of the real world variables.

Paul Westhaver
July 31, 2012 8:03 pm

serious question…
Does the tree ring growth data, that Mann excluded, in favor of hiding the decline agree better with the corrected temp data set?

A. Scott
July 31, 2012 8:05 pm

I would also note the Leroy 1999 siting results data is known from prior studies … it would seem that station data would be an excellent place to start. Add the Watts comments on classification category changes and appears you could well narrow the station list initially pretty quickly – at least create a list of most likely culprits to start with …

ttfn
July 31, 2012 8:05 pm

TOBS is the temperature at the time of observation. The observer records the max/min temps, records the current temp (TOBS) and resets the device. The reported TMAX and TMIN are valid for the time period between resets (the time of the previous ob to the current ob), If the observer takes the reading after sunrise, TMIN will be the low that occurred on the day the observation was taken UNLESS the previous day was colder. For example, the observer recorded TMIN 25 degrees and TOBS 26 degrees on the 10th at 8am and reset the device. On the 11th, the observer recorded TMIN 26 degrees and TOBS 42 degrees. Obviously, the TMIN recorded on the 11th actually occurred on the 10th seconds after the device was reset. The closest we have to TMIN on the 11th is TOBS unless the 12th happens to be warmer than the 11th and so on. It would be nice if the MMTS timestamped TMAX and TMIN. Just another example of the mess that is climate. Trees are probably more accurate.

Werner Brozek
July 31, 2012 8:07 pm

Time of observation can get extremely tricky. In Edmonton, Alberta, Canada, for example, the daylight hours can vary from about 17 hours in June to 7.5 hours in December. So if readings are taken at 7:00 AM and 7:00 PM every day, for half the year, this would be after sunrise and before sunset. Then for the other half of the year, it would be before sunrise and after sunset. Then there is daylight saving to throw in a further complication. Rather than having a specific time of observation, perhaps times connected to local sunrise and local sunset would be more useful. But who has a computer that can sort that out on a global basis? And who is going to get the readings at just the right times which can change by 3 minutes every day in a place like Edmonton?
REPLY: The max and min thermometers have memory posts on the meniscus. Captures the high and low of the day. – Anthony

davidmhoffer
July 31, 2012 8:08 pm

While I understand the TOBS issue, what I don’t understand us why TOBS must be applied to this particular analysis. Is the point not to determine if station siting issues influence the trend? Would it not be as valid to compare both trends with TOBS out as it would be to compare both trends with TOBS in?

Christoph Dollis
July 31, 2012 8:11 pm

Christopher, a minimum is a minimum and a maximum is a maximum, please tell us all why you keep drawing time into the picture concerning this paper.

Because minimax aren’t the only temperatures recorded, hence TOB adjustments.
Will you tell Steve McIntyre he’s a moron trying to figure out something completely unnecessary or shall I?
REPLY: Retract your statement about Mr. McIntyre or get out permanently. I have zero tolerance for that – Anthony

Steve Huntwork
July 31, 2012 8:17 pm

“If the observer takes the reading after sunrise, TMIN will be the low that occurred on the day the observation was taken UNLESS the previous day was colder.”
Good assumption unless other raw data (such as atmospheric pressure) is available that actually showed a frontal passage moving though the area on that specific date. Even a frontal passage will require a sudden change in temperatures, pressure and wind directions.
Keep it simple and use the 99% standards of normal TMIN and TMAX date and times, unless other information becomes available that can identify a frontal passage.

Christoph Dollis
July 31, 2012 8:17 pm

Anthony, I was being completely sarcastic. I don’t for a second think that about McIntyre: my point is the opposite. That he’s obviously trying to calculate TOBs because he feels they are significant.
I should have added a /sarc tag. It was a back-handed compliment. In my posts today, I’ve done nothing but laud Steve McIntyre.
REPLY: OK point of order – ALWAYS use the sarc tag when using sarcasm And, that’s probably my cue to step away for awhile, I’m getting really tired. – Anthony

Jim Steele
July 31, 2012 8:19 pm

Several studies show that tree ring data agrees well with the maximum temperature.Unadjusted maximum temperatures throughout the USA on average do not exceed the 1930’s and 40’s and thus reveal a slight decline.
The latest Esper paper agrees well tree rings but their mean Scandinavian temperatures likewise does not exceed the 40’s.
I think Maximum temperatures are more representative of the heat stored in the atmosphere because of the tremendous vertical mixing of mid day convection. Minimums are more often biased by land use, GHGs, wind patterns and what not that disrupt the boundary layer. Any warming not radiated back to space by dawn can be carried to the stratosphere by the next day’s convection. Minimums likely do more to obscure any climate trend.

July 31, 2012 8:22 pm

Bob Johnston says: July 31, 2012 at 6:03 pm TOBS example
http://climateaudit101.wikispot.org/Time_of_Observation_Bias
This example is WRONG. It is wrong because a change to 3 pm observations changes not only the 2 days discussued, but all days in a time series read at 3 pm. Therefore, the average will not be the same as that given.
TOBS is tricky. You have to work the examples quite carefully. Others are also wrong.
In Australia, the vast majority of observations in relevant years were made at 0900 hrs. It is rare for there to be a need for any TOBS correction. If NOAA or GISS whomever add a TOBS correction to Australian data, they will create an error. Probably, the same applies for many other countries outside USA.

Rud Istvan
July 31, 2012 8:23 pm

AW, keep up the good work. Remain open to comment. TOBS is going to be a distraction rather than a problem, as pointed out on this thread already by others.
I would however suggest that since 71% of Earth is ocean, your wonderful analysis of US siting bias be extended to oceanic measurement error. That way, becomes much more compelling. There are some possible ways to begin such an effort. Satellites versus thermometers, various thermometer depths. Stuff that is already in the literature, and provably ‘over-corrected’ by databases such as GISS.
There are also critiques of AR4 beyond data quality. The combination should become quite powerful for AR5. I will get my own book chapter on this out before the election, since was not originally targeted at AR5.

NeedleFactory
July 31, 2012 8:24 pm

At http://rankexploits.com/musings/2012/initial-thoughts-on-the-watts-et-al-draft/, Zeke Hausfather says [Leroy 2010] results in a less strict criteria than that of Leroy 1999, and considerably more stations are rated as class 1 or 2 (160) than in the prior classification scheme (~80). This should by itself raise a yellow flag: if using a more strict classification criteria found no difference in trend, why would a more lax classification criteria result in very significant differences in trends (at least in the raw data)?
I think Zeke uses strict in two different senses here: (1) how many sites get “high/low” grades; (2) how many sites are “good/bad”. It seems to me that the newer classification scheme will improve the rating of some sites and lower the rating of others.
Consider: if the older scheme jumbles all sites around, then all five classifications less difference one from another than if they are “more correctly” assorted, so that the classifications better reflect real differences in stations.
Have I missed something in Zeke’s reasoning?

Christoph Dollis
July 31, 2012 8:28 pm

NeedlesFactory, great observation. I noticed that too, but you took the time put it very cogently.
It seems to me Leroy 2010 is simply a better standard — of course the surface area of nearby heat sinks matter.

Steve Huntwork
July 31, 2012 8:35 pm

Many times today I have seen statements of 02:00 PM or even 03:00 PM times of observations.
This may be good on a theoretical basis, but NOBODY has ever based their 24 hour weather records upon those times. If I am wrong, then please show me a single station anywhere in the World that used those times.

Konrad.
July 31, 2012 8:37 pm

I posted this comment on another thread in response to vvenema’s unfounded joy at having found a TOB “out” from the pain of the paper. I feel it is also relevant to this thread.
——————————————————–
vvenema says:
July 31, 2012 at 1:06 pm
——————————————————–
Hand flapping about TOB adjustment in the hope that it will remove this thorn from the side of the consensus is not going to work. In fact it may have the opposite effect than what you intend. You are just attracting attention to Tom Karl’s pet rat TOBy.
There are two types of TOB adjustment that may be valid. The first is time zone adjustment, to account for the true sidereal position of the sun over a station. The second is a one time only step adjustment to individual station data for a change between evening or morning reading of max/min mercury thermometers. The second type of adjustment cannot be validly made from a desk in a distant city. It can only be valid if it is made on an individual station basis with direct reference to individual station paper records. Which method do you think Toms rat TOBy has been using? Yes, that’s right, from a desk in a distant city, not on an individual station basis and with no supporting metadata.
To make a valid TOB adjustment you would need to know whether an individual station was making evening or morning readings of a mercury thermometer, if and when the reading time for that station changed and when the station changed to an MMTS sensor. To achieve this for USHCN stations would require a project similar in scale to Anthony’s surface station project.
Anthony has shown that you cannot adjust for station site issues from behind a desk with any amount data smearing. NOAA thought you could. NOAA also thinks you can adjust for TOB from behind a desk with no supporting individual station metadata. So by all means make a fuss about TOB adjustment invalidating Anthony’s work. Lets drag TOBy squeaking into the disinfecting sunlight.

Christoph Dollis
July 31, 2012 8:38 pm

No worries, Anthony.
Great work, on balance. I’ve critiqued you today for rushing to press so it may not seem like it, but your site is and has been for years one of my favourites and I greatly appreciate your work. I think it’s valuable not just to science, but to humanity … considering the terrible ramifications of bad environmental and economic policy.
You’ve earned a break tonight, that’s for sure! It’s a valuable discussion and once the paper’s tightened up (linguistically too), i hope to see it published. If you’re right, it’s huge.
We’ll need and want more “SurfaceStation”-type projects in other countries. Like Watts 2009, this should improve how science is down and increase, however marginally (!), the numbers of people looking at popular climate science with a skeptical mindset.

John Trigge (in Oz)
July 31, 2012 8:52 pm

Anthony,
I seem to recall but cannot find it now, that your original interest was to investigate the effect of different paint surfaces on Stevenson Screens. From that study you determined that the surface coating did have some effect on the temperature readings.
Where in the ‘adjustments’ is this taken into account and, if not, why not?

johanna
July 31, 2012 8:55 pm

I notice that Steve McIntyre said on his blog that, unfortunately, he has to get into TOBS, which he regards as a time sink (presumably with non-commensurate results).
I’m guessing that one of the reasons for his frustration is that the significance of TOBS varies a great deal by location. In the tropics, the temperature often does not vary much throughout the 24 hour cycle. In the desert, it can vary enormously, as it can in places (like Melbourne, Australia) where you can have ‘4 seasons in one day’. Trying to find appropriate corrections for these sorts of variables would be challenging, to say the least.
From personal experience, I know that any decent piece of work one does raises further questions – otherwise it is not a decent piece of work. But, at some point you just have to draw a line under it and say ‘it is what it is, it’s not the Grand Unifying Theory, just another small step forward. We acknowledge that X,Y and Z are not yet resolved.’
My best wishes to you and your family, Anthony, and good luck with resisting the taunts to follow every rabbit (or rabbet) down every hole.

markx
July 31, 2012 9:06 pm

Christoph Dollis: July 31, 2012 at 5:53 pm
Said:” …. Ummm, the difference is the temperature doesn’t fall and rise with the regularity of a Sine wave. …”
On average, yes it does.
And we are not talking about an “on the spot observation”, it is the recording of an earlier maximum or minimum.
So IMHO Steve Huntwork: July 31, 2012 at 5:46 pm is right on the money.

Christoph Dollis
July 31, 2012 9:13 pm

markx, the average amplitude of a Sine wave is always 0. This isn’t the same thing as temperature values despite that, yes, they both rise and fall.

Christoph Dollis
July 31, 2012 9:17 pm

Johanna, Steve McIntyre said that temperature records generally are a time sink, and that basically he accepts the official records, if not in fact then as a reasonable basis to start working from. .

July 31, 2012 9:20 pm

Well done. I’m happy to see this study being discussed in the mainstream media. Doing great science is one thing. Getting the press to cover it is another. Well done.

JR
July 31, 2012 9:25 pm

No, Steve Huntwork 5:46pm is *NOT* right on the money. Anyone that correctly understands TOB does not claim that the actual recorded temperatures are in error. Changing the time that the max and min thermometers are read will cause a change in the resulting computed monthly mean values vs. what they would be if the time had not changed. This will show up as a step change in the temperature time series at the point where the time of observation is changed, and if you add a step change to a time series, you change the trend. If the step change is artificial – e.g. because the methodology changed and not because the climate changed, then you just artificially biased the trend. The question is not whether TOB needs to be adjusted for, it is by how much.

July 31, 2012 9:41 pm

Christoph Dollis says:
July 31, 2012 at 9:13 pm
“…the average amplitude of a Sine wave is always 0.”
Only if it is not clamped.☺

AndyG55
July 31, 2012 9:41 pm

What doesn’t make any sense is that the UNHCN TOBS adjustments since 1970 have ALWAYS increase the warming trend.
MUST be just coincidence , I guess.

Manfred
July 31, 2012 9:41 pm

Eric Barnes says:
July 31, 2012 at 7:18 pm
Reading the Karl TOBS paper here
ftp://ftp.ncdc.noaa.gov/pub/data/ushcn/v2/monthly/karl-etal1986.pdf
———————————————
I haven’t found a reference for the data in his table 1 (Percent of am, pm and md cooperative weather stations in the US). That data is essential for the amout of correction and in the paragraph referring to table 1, there is only a reference to a “personal communication” with Shoun in 1985.
And is TOBS applied as well outside the US and on what data base ?

Gunga Din
July 31, 2012 9:44 pm

Christoph Dollis says:
July 31, 2012 at 8:38 pm
No worries, Anthony.
Great work, on balance. I’ve critiqued you today for rushing to press so it may not seem like it, but your site is and has been for years one of my favourites and I greatly appreciate your work. I think it’s valuable not just to science, but to humanity … considering the terrible ramifications of bad environmental and economic policy.
=============================================================================
A little off topic, but I think this wasn’t so much a rush to press as a use of a new medium. I’ll liken it to the time when religious and legal papers were all written in Latin. The people didn’t know latin so they didn’t really know what the clergy and the lawyers were talking about. (WARNING! I’m about to twist a figure of speech to the breaking point.) This paper hasn’t yet passed through the Latin process of peer-review but it (a draft ot it) was put out there in English for the people to read and try to understand and critique even though much of it is still Greek to us. (Whew!) “People-Review” before “Peer-Review” is a good idea.
Someone once told me, “It’s what you learn after you think you know it all that matters.” I don’t know if they were quoting someone else or not, but maybe it’s time the peer-review process learn something that matters.

dalyplanet
July 31, 2012 9:46 pm

I am still not ‘getting’ how TOB affects more than the day of adjustment.There is a recorded min and max and if it starts high and goes low or the other way or starts in the middle after one day of shift will be balanced again.

markx
July 31, 2012 9:52 pm

JR : July 31, 2012 at 9:25 pm
said:…. ” Changing the time that the max and min thermometers are read will cause a change in the resulting computed monthly mean values vs. what they would be if the time had not changed. This will show up as a step change in the temperature ….”
OK, granted.
But then, how many times have they changed the TOBS? How many steps are we talking about?
… and is it always in a direction which adds an error, and so needs to be adjusted down?

Christoph Dollis
July 31, 2012 9:56 pm

“Anyone that correctly understands TOB does not claim that the actual recorded temperatures are in error.”

Two separate issues here. To be sure, a lot of people here don’t yet understand TOB (and it’s a new concept for me also). However, recorded temperatures may still be in error due to poor site design and maintenance (you didn’t dispute that; just throwing it out there, since that’s a major part of Watts’ paper).

Christoph Dollis
July 31, 2012 10:02 pm

A little off topic, but I think this wasn’t so much a rush to press as a use of a new medium.”

Not calculating the TOBs, which Anthony feels are minor anyway, isn’t my major concern. My major concern is more interpersonal: that McIntyre felt a bit rushed and not yet comfortable with having his name on the paper. I feel (and I think it’s little in doubt) that it would have been better to give Steve a bit more time. Anyway, water under the bridge and all that.
The discussion has started and it’s an important discussion. I believe Anthony has definitely identified differences in the rate of temperature increases in various stations depending on their Leroy 2010 classification, and also shown that airports are nearly wholly unreliable for fine climatic measurements. Worse, the data homogenization-adjustments use the worst sites as a starting point rather than the best sites, revealing a bias toward warming on the part of most climate scientists.

Gunga Din
July 31, 2012 10:03 pm

“People-Review” before “Peer-Review” is a good idea.
I should add that I left out “Press-Review”n purpose.

Christoph Dollis
July 31, 2012 10:09 pm

“is it always in a direction which adds an error, and so needs to be adjusted down”

Changing the measuring time could add a mean error up or down. On balance, it may be very minor as Anthony feels. However, to be thorough and as certain as possible, it’s right to address it as Steve is doing.

Manfred
July 31, 2012 10:11 pm

dalyplanet says:
July 31, 2012 at 9:46 pm
I am still not ‘getting’ how TOB affects more than the day of adjustment
—————————————–
If you are reading data around minimum temperature time of a day and then reset, the next measured temperature will be the same minimum again and may remain the minimum for the next day as well if it doesn’t get colder. You then have a 2 day minimum value measured on one day for both days and that is a cold bias. The other way round if you read around the maximum time.

Jim Steele
July 31, 2012 10:13 pm

JR says:
Changing the time that the max and min thermometers are read will cause a change in the resulting computed monthly mean values vs. what they would be if the time had not changed. This will show up as a step change in the temperature time series at the point where the time of observation is changed, and if you add a step change to a time series, you change the trend. If the step change is artificial – e.g. be…. The question is not whether TOB needs to be adjusted for, it is by how much.
JR, I think your notion of an artificial step change my be the key issue. However a one day mistake should not impact the annual average so any such adjustment seems totally unnecessary… unless you are concerned about how any step changes affect the homogenization process. Often their algorithm interprets a step change as the point where adjustments needs to me made.

Christoph Dollis
July 31, 2012 10:15 pm

By the way, I think it’s possible TOB adjustments will be significant in and of themselves. However, I have no reason to think they’ll trend any differently for airports vs. non-airports, rural vs. suburban vs. urban, Class 1 sites vs. Class 4, etc.
I’m curious, though, now that Steve McIntyre is looking into it. Might as well see what he comes up with.

Nick Stokes
July 31, 2012 10:19 pm

” Steve Huntwork says: July 31, 2012 at 5:46 pm
“TOBS has always been a dating issue and never a temperature offset.

Steve, you frequently refer to your experience at stations taking hourly readings. That’s not what TOBS is mainly about. It’s dealing with the history of stations that used min-max thermometers, which were reset once a day, and which recorded mechanically the min and max between the resets.
The intention is to represent the day’s temperature as the mean of the min and max for that day. Not ideal, but it’s the record that we have, and OK if it is consistent.
But suppose that the reset is at 5pm. Then when a cool day follows a warm one, it can happen that the highest temp over the 24hr period was at 5.01pm on that previous warm day. So in effect, the warm day is counted twice. If a warm day follows a cool, however, it is very unlikely that 5.01 pm will be the coldest point of the warmday cycle. So warm days are double counted, cold not. That’s a warm bias. Morning readings have a similar cool bias.
As long as it is consistent, that doesn’t matter much. The problem comes when you shift from evening reading to morning. Then there is a corresponding cool bias, and the change gives a cooling step.
Now the NWS relied a lot on COOP volunteers. It asked them to reset in the evening, but read rain gauges in the morning. They tended to shift after a while to reading both in the morning. That’s a cooling trend.
There’s no supposition here. The reset times are recorded. There is a lot of information on the diurnal cycle, and day-to-day variation. There’s a clear need to adjust when reset times changed, and the numbers are there to do it.
davidmhoffer says: July 31, 2012 at 8:08 pm
“While I understand the TOBS issue, what I don’t understand us why TOBS must be applied to this particular analysis.”

As McSteve says, it’s an important confounding effect. The reason is that TOBS applies where a change was made, and this occurrence is very likely to be related to CRN status. And it is a big effect.

DavidA
July 31, 2012 10:23 pm

There’s lots of stations and TOBs should be randomly spread across the board – you’d think. So can’t see how accounting for it will make a whole lot of difference to the main conclusion.

Maus
July 31, 2012 10:28 pm

JR @ 9:25: ” Changing the time that the max and min thermometers are read will cause a change in the resulting computed monthly mean values vs. what they would be if the time had not changed. ”
Angels dancing on the head of a pin. The midpoint of two extremums of a range is not the same as the average of 24 individual samples over some given time or even just 2 samples spaced 12 hours apart. Such a midpoint is neither a mean, median, nor mode; and it makes not a lick of difference if you call it ‘Tmean’, ‘Tgreen’, or ‘greenT’. It does not and cannot represent any notion of area under a curve.
The entire fret about TOBS is simply a canard based on a mathemagical sophistry. Every Tmin/Tmax pair represents the extremums over, notionally, the prior 24 hours. The only manner in which this can be effected by time of observation is if the *specific time* at which a Tmin event and a Tmax event occur is not fixed day to day. But that only states that what we are doing *is not taking an samples at specific time intervals.* But we already knew that, didn’t we?
Changing the TOB for a given site can only reduce or increase the interval over which we collect extremums *one time* per change. That is, it can only produce a single spurious outlier. If you have enough such outliers then it is certain that your dataset is garbage for that site and no amount of massaging will repair it. But if a single spurious outlier is enough to destroy your trend over 365 24 hour sampling periods?
Then you were trying to speak more accurately then your instrumentation allows. Nothing but GIGO.

Christoph Dollis
July 31, 2012 10:30 pm

Thanks for that great explanation, Nick — it’s the best I’ve seen so far. From what you say it’s a bigger effect than I thought, too.

Christoph Dollis
July 31, 2012 10:33 pm

“Angels dancing on the head of a pin. The midpoint of two extremums of a range is not the same as the average of 24 individual samples over some given time or even just 2 samples spaced 12 hours apart. Such a midpoint is neither a mean, median, nor mode; and it makes not a lick of difference if you call it ‘Tmean’, ‘Tgreen’, or ‘greenT’. It does not and cannot represent any notion of area under a curve.
“The entire fret about TOBS is simply a canard based on a mathemagical sophistry.”

Now that I understand it better, I don’t agree — not at all:

“But suppose that the reset is at 5pm. Then when a cool day follows a warm one, it can happen that the highest temp over the 24hr period was at 5.01pm on that previous warm day. So in effect, the warm day is counted twice.”

Now I will say there’s no reason to assume the effect is more or less at the various types and classifications of sites. So maybe it’ll all average out. However, it appears to be a significant introduced bias and should be accounted for if the data is available to do so.
Why not remove unnecessary doubts and avenues for attack?

July 31, 2012 10:33 pm

At the end of the day, what I take away from this is that so many data quality issues surround just the last century or so of data that trust necessarily becomes a matter of faith.
While fascinated by this discussion, I spent some quality time this evening perusing the papers here: http://www.clim-past.net/8/issue4.html
The abstract, discussion and conclusions sections of which will really broaden your perspective as you further ponder when we live, the half-precession old Holocene, the efficacy of TOB, station siting etc.
Do the same for http://www.deas.harvard.edu/climate/seminars/pdfs/Tzedakis_etal_2012.pdf and see what sense you make out of all this.
Makes for an interesting evening, just sayin.

July 31, 2012 10:38 pm

Anthony struck a nerve with media matters,
http://mediamatters.org/research/2012/07/31/conservatives-still-trying-to-dispute-global-te/189034
REPLY: Yeah they’ve pretty well flipped out. Can’t have any questions you know.
And what a bunch of hypocrites, they quote everyone else, including 3rd person disturbed Bunny Boi who they outed:

Watts’ failure to make certain adjustments to the raw data, as NOAA has done, is a serious flaw knowledgeable bloggers say. Specifically, Watts did not apply a time of observation bias correction according to Howard University chemistry professor Josh Halpern, who blogs under the pseudonym Eli Rabett.

…but they don’t dare link to the Watts et al paper for anyone to read for themselves. -Anthony

markx
July 31, 2012 10:56 pm

As someone in here pointed out a long time ago…
If you find an added historical uncertainty to the parameter you are measuring, rather than “adjust the historical data”, should you not simply widen the error bars?

johanna
July 31, 2012 10:58 pm

Can I please reiterate that Watts et al is not proposing a Grand Unifying Theory of Climate, or even of climate measurement. It is clearing some of the underbrush so that we can move forward. People who complain that it doesn’t address their particular concern are missing the point.
At a minimum, this thread tells us that TOB is a highly contested area. Blaming Watts et al for not resolving it to everyone’s satisfaction is like complaining that antibiotics do not cure the common cold.

Christoph Dollis
July 31, 2012 11:05 pm

“If you find an added historical uncertainty to the parameter you are measuring, rather than “adjust the historical data”, should you not simply widen the error bars?”

Sure, but if the data (records of times temperatures were taken) exist and introduced biases can be removed mathematically following recognized, sound procedures, why not do so?

Maus
July 31, 2012 11:29 pm

Christopher Dollis: “Now that I understand it better, I don’t agree — not at all:”
Fair enough. But in the quoted example you gave that would be true of every cold day that followed every warm day. And every warm day that followed a cold day. And so on. No matter what single specific time we choose for the reset of the Tmin/Tmax we are always taking a single hard mark over the previous 24 hour frame we’re dragging along. And then erasing the entire frame of data. We cannot do otherwise as we only have a single recording device.
This is similar to, but not synonymous with, Nyquist frequencies. Such that if we wish to avoid any issues arising from a specific time at which to reset the device then we need a pair of devices. Each recording the extremums over a 24 hour period but in which their resets are staggered on 12 hour periods. Any and every other solution is going to fall prey to the specific case you quoted. (Though this only allows us some continuity over daily periods, it does not allow us to speak intraday. For which we’d need 4 devices on staggered 6 hours resets. Which only allows us to … ad nauseum.)
For all else in regards human issues, which are certainly part of the instrumentation method, see Mr. Watts’ reply.

Kev-in-UK
July 31, 2012 11:33 pm

As with others, I find the TOBS issue somewhat perplexing and, to be honest, I do think it is looking for and making unnecessary complications..
My impression is that TOBS is the result of having the modern continuous Temp instruments and the realisation that the (Tmax-Tmin)/2 values do not actually represent anything like an average ‘temperature’ of any given day? Even manual recorded TOBS temps (the actual temp at time of obs) is essentially useless except for comparison to the Tmax and Tmin recorded at the same time to note whether the TOBS was ‘between’ Tmax and Tmin? I can’t imagine there are many stations where manual TOBS temp coincided with either Tmin or Tmax on a regular basis?
Hence, there seems little point in trying to adjust old manual data – it is like putting lipstick on a pig, and it is much better to separate the old and new datasets and use them independently.
The new datalogging type temp records can always be made to look like the old records (by a simple algorithm to ‘extract’ standard Tmax and Tmin?) if we need to make a general comparison to older data – but old data cannot be made up to modern standard without introducing some bias via various assumptions, etc, etc.
My general impression is that the older RAW data should be able to show obvious step changes from TOBS changes (if they are there), when plotted – kind of in a similar way to station relocations? But overall, I would expect them to be very very small effects? – I dunno, perhaps changing TOBS could cause a local dip or rise in the ‘average’ for a few days – but after that it would settle down again? Does anyone seriously think this would have a massive effect on long term trends? Has anyone ever done such an analysis? I presume it would require the manual records to be available, with details of the TOBS changes too? I would certainly prefer this kind of approach rather than some estimated/assumed guesswork on the actual effect of TOBS being ‘added’ en masse to millions of data points!
Kev

Nick Stokes
July 31, 2012 11:36 pm

Anthony,
“TOBS adjustments are a big hammer that misses the mark based on human behavior for filling out forms and times they can’t predict.”
There’s nothing that can be done to improve data where observers were writing incorrect information on the forms. But if you believe what they wrote on the forms is worth taking account of, then it is there on the record, and can be analyzed by people sitting in offices as well as anyone.
The 2009 BAMS paper of Menne et al has a Fig 3 which shows the trend of observation times that observers actually reported. And Fig 4 shows the resulting effect of a TOBS adjustment on trends, based on the Fig 3 data and the known diurnal and day-to-day variability.
REPLY: Noooo…Fig. 3. Changes in the documented time of observation
in the U.S. HCN. is about the times they assigned the observers. There’s no proof the observers adhere to it. – Anthony

July 31, 2012 11:41 pm

While MMTS TOBs is more complicated, there are also a lot of CRS stations that experienced TOBs changes in the 1979-2010 period, and that definitely introduces bias. It should also be reasonably easy to test the MMTS data (coupled with the TOBs recorded in that station metadata) and see if there are any detectable correlations between TOBs changes and temp changes. I might play around with the data once I get bandwidth; anyone know if the TOBs metadata is available in an easily digestible form?
REPLY: What Zeke illustrates is what I see as the biased nature of the whole network. There’s so many problems that all sorts of issues have to be invented to deal with it, and all the corrections they invent go UP. And like NCDC, Zeke’s trying to find the answers in the data, rather than looking at the reality of the measurement environment. Nobody wants to deal with that side of it. Color me unimpressed with the whole adjustment crusade. – Anthony

Latimer Alder
August 1, 2012 12:18 am

Anthony’s update certainly hits the mark with my experience of human behaviour. You get exactly what you pay for and ‘police’. And if that is ‘nothing’ and ‘not at all’ respectively then some folks will react accordingly. You will get poor quality work. We all do it to a greater or lesser extent. To pretend otherwise is naive and simplistic.
The more I read and understand of how this basic ‘climate’ data has been collected and processed and massaged and adjusted interpreted, the more certain I am that it is just not fit for the purposes to which it is put. 4
And when your base data is so fundamentally flawed, any theories or predictions that rely on it are pretty much useless also. Which seems to cover a huge swathe of climatology.

J.Hansford
August 1, 2012 12:19 am

John Trigge (in Oz) says:
July 31, 2012 at 8:52 pm
Anthony,
I seem to recall but cannot find it now, that your original interest was to investigate the effect of different paint surfaces on Stevenson Screens. From that study you determined that the surface coating did have some effect on the temperature readings.
Where in the ‘adjustments’ is this taken into account and, if not, why not?
================================================================
Seems to be an element of the smart Alec in you post….. But you do hit on a valid point, the point being of course, that you can’t do “adjustments” for all the confounding elements intrinsic within the surface temp data…. The real fact is, these temperature measurements were never meant to be used to the resolution and exactitude that they are now being subjected to. I think this is Anthony’s opinion also… That’s why he said that even the condition of the Stevenson screen paint will have an effect on the temperature recorded inside, which then morphed into the fact the neglect, condition and siting was even worse at these sites then could have ever been imagined.
So basically, the surface record should be scrapped and the Satellite record adopted. The surface observation record is not really scientific data beyond the purpose that it was originally intended, which was simply Meteorological observations for the purpose of weather forecasting and general information…. not data collected for determining anthropogenic CO2’s effects on temperature and global climate down to a resolution of a 1000th of a degree F.
…. The “adjustments” end up creating bias and compounding errors…. which is the subject of this paper that Anthony is doing.
I hope Anthony agrees with my summation of his views and motives…Pretentious poppet that I be.. 😉

August 1, 2012 12:20 am

John Trigge (in Oz):
” … effect of different paint surfaces on Stevenson Screens … ”
Sure does. I have an RH/Temp logger set up in a semi-compliant Stevenson Screen. Siting probably not as bad as a “5” under Leroy 2010, but maybe not far off 🙁
Here at 19°S NQ late July I was getting ‘spikes’ from the low angle morning sun, one of which was higher than the maximum at 3pm. Additional (3rd) coat of paint made a difference. I might try cenospheric paint if I can find a sample pot, or the spheres themselves – preferably a metric handful rather than a 500kg bag 🙂
johanna says:
” … I’m guessing that one of the reasons for [StevM’s] frustration is that the significance of TOBS varies a great deal by location. ”
At least there is now scope for further study of TOBS, Leroy Classification etc. Data loggers starting around $41 with an accuracy of +/- 0.5 C, 12 hour sampling. My rig set up for less than $200: original purpose was to learn about data logging for assessing thermal performance of building materials.

Ian
August 1, 2012 12:25 am

I have a horrid feeling that if my comment is read I shall reveal to those readers my lack of knowledge. With regard to TOBs with the max/min thermometers I’ve used the maximum and minimum temperatures are clearly identifiable. I can see that if the thermometer isn’t read until very late the next day then “today’s” temperatures may well obscure “yesterday’s” temperatures. Possibly advances in technology have overcome that but for earlier readings, human nature being what it is, how can the TOB be accurately known? On another point, which I suspect has been done to death, knowing the SD or SEM of the decadal trends would seem very necessary to see if differences from the different groups of surface station are or are not statistically significant

kadaka (KD Knoebel)
August 1, 2012 12:28 am

From Nick Stokes on July 31, 2012 at 11:36 pm:

The 2009 BAMS paper of Menne et al has a Fig 3 which shows the trend of observation times that observers actually reported.

“Not Found”
File can be found here:
ftp://ftp.ncdc.noaa.gov/pub/data/ushcn/v2/monthly/
And why are you messing with Goddard’s old “hacked” site anyway, that Steve said had been taken over by someone he thought he could trust but was proven wrong?
Gee Nick, you’re not verifying your links. On the predecessor thread you couldn’t calculate temperature trends correctly. Is something bothering you to distraction, buddy?

Manfred
August 1, 2012 12:37 am

Nick Stokes says:
July 31, 2012 at 11:36 pm
There’s nothing that can be done to improve data where observers were writing incorrect information on the forms.
—————————————
I would have 2 suggestions to deal with that:
1. Compare COOP stations with other stations with “really” known observation time.
2. Delete ALL entries, where either tmin or tmax is double on consecutive days and fill the gaps with gap filling algorithm which are already available.

A. Scott
August 1, 2012 12:37 am

Nick Stokes
Why wouldn’t the TOB adjustment simply be a single, identifiable, “step” change, easily discerned in the data?
And as Anthony notes considering the why would there be ANY need for a TOB adjustment for the Max-Min recording stations that have been in use for a long time now?
If I recall MMTS systems are well over 50% of the stations?

Arnost
August 1, 2012 12:39 am

Konrad: “To make a valid TOB adjustment you would need to know whether an individual station was making evening or morning readings of a mercury thermometer, and if and when the reading time for that station changed …”
Excellent point.
Given that Anthony is only referencing the period from 1978-2008, then something like 50% of the stations would have ALREADY changed TObs (Time of Observation) from evening to morning. So these cannot be an issue. See DeGaetano 2000
http://journals.ametsoc.org/doi/pdf/10.1175/1520-0477(2000)081%3C0049%3AASCSOT%3E2.3.CO%3B2
Note: the above paper links to a dataset for all the HCN stations that identifies the TObs changepoint. So if push comes to shove, it should not be a great amount of work to use this or even to go through each site’s COOP B-91 forms (as they have the TObs on it) and confirm EXACTLY when and IF a change in TObs occurred on each of Anthon’y sites. And if necessary make an appropriate adjustment to the site record.
As to the size of the TOB likely overall adjustment… Menne et al 2009 say: “The net effect of the TOB adjustments is to increase the overall trend in maximum temperatures by about 0.015°C decade−1 (±0.002) and in minimum temperatures by about 0.022°C decade−1 (±0.002) during the period 1895–2007.” And in a pre-print here:
https://ams.confex.com/ams/pdfpapers/141108.pdf
Say: 0.012°C dec-1 and in minimum temperatures by about 0.018°C dec-1 over the period 1985-2006
This is a tenth of the trend Anthony has identified [i.e. “The Raw Tmean trend for well sited stations is 0.14°C per decade lower than adjusted Tmean trend for poorly sited stations”] – and given that the TOB issue affects at most some 30-40% of the stations used – the impact should be even less!
Time of Observation Bias may not really be a major issue and in no way invalidates Anthony’s paper’s main thrust that: good stations are adjusted by bad.

A. Scott
August 1, 2012 12:41 am

Anthony … can I suggest making the new paper a small linked “badge” and place it at top of right hand column for easy reference?

son of mulder
August 1, 2012 12:45 am

So if I understand correctly the TOBS issue means that if I take a high reading at say 3PM each day and tomorrow the high doesn’t reach as high as the temp at 3.01PM today then tomorrow’s high reading will be as at 3.01 PM today so there would tend to be a warm bias in the data for daily highs. Similar and opposite if I took low readings at 3AM and the next night was warmer I’d get an artificial low for tomorrow.
So I’d suggest that if most readings were taken during the day there would be a tendency to higher than real Tmax on average and if most readings were taken during the night there would be a tendency to be lower than real Tmin.
But that wouldn’t affect the trend of Tmax readings or the trend of Tmin readings as the bias is built in all along the time series unless practice changes at some point.
And wouldn’t most readings be done at start of day or end of day so lower risk of an issue.
So I need to understand why TOBS is an considered a significant issue for the trend results of Anthony’s study.

Brian D
August 1, 2012 12:46 am

Damn good point, Anthony.
But for those who would like to see the difference in reading time monthly means, I took the hourly data from the airport near me via wunderground and computed the midnight, 7am, and 5pm monthly averages. All am readings adjusted to 7am criteria and all pm readings to 5pm criteria. That’s what I remember from the Karl paper. Also, they do some slight adjust for the end of month discrepancy.
March 2012
Midnight: Max = 47.3F(8.5C), Min = 29.0F(-1.7C), Avg = 38.2F(3.4C)
7am: Max = 46.8F(8.2C), Min = 27.3F(-2.6C), Avg = 37.1F(2.8C)
5pm: Max = 49.4F(9.7C), Min = 29.3F(-1.5C), Avg = 39.4F(4.1C)
Guess NOAA should do a survey on those issues you brought up, Anthony. They might be surprised at what is happening out there in the network.

sean71
August 1, 2012 12:48 am

The TOBS question is very simple. If your paper is discussing station quality bias, you could compare two different methods of combining TOBS adjusted data. That would answer the question which the paper superficially seems to address.
If you wish to address ALL of the adjustments, you need to include them all in your analysis. This is clearly a much more complex problem, so performing the first (quality based) analysis is the obvious first step. I guess this will be as inconclusive as the Fall 2011.
In changing 2 variables at once, you loose the ability to identify which variable is responsible for the result you observe. Post TOBS has a higher trend than pre-TOBS; well, yes… and?
Having commented on this paper elsewhere already, I feel slightly cheated.

A. Scott
August 1, 2012 12:52 am

The USHCN Version 2 Serial Monthly Dataset page:
http://www.ncdc.noaa.gov/oa/climate/research/ushcn/

David Schofield
August 1, 2012 1:07 am

Simple question but it jumps out at me.
Does the stepwise graph above show that observers were a lot less diligent from 1970 than previously? If so why?
I don’t believe they were but isn’t that an implication of the adjustments?

August 1, 2012 1:19 am

The TOBS adjustments are not there to correct any mathematical or observational failures, they are there primarily to all yet another round of temperature nudging toward the great holy goal of Global Warming. The observations in the field, shown in the article, reveal just how human nature really works and how temperature recording will be done in practice. Such is life!

David Ross
August 1, 2012 1:20 am

Anthony,
From reading this thread and from gleaning comments on the warmist blogosphere, I think TOB is the main issue that this paper’s critics will use to attack it, and has got to be comprehensively addressed.
Regardless of how and whether you finally use TOB adjustments I think the paper needs a lengthy section describing the issue, like your …
“Why I don’t think much of TOBS adjustments”
…obviously with a more scholarly tone.
However, anecdotal evidence doesn’t look good in a scientific paper, no matter how compelling.
Weather station technology has changed over time, and I think there is no-one more informed about the details and implications of that for the United States dataset than you.
You mention the “advent of MMTS, which has a memory.” Are these “advents” sudden enough to demonstrate a step change in recordings? Is it known when the equipment changed at each station? Failing that, even the dates of introduction of new MMTS devices and subsequent sales figures might be instructive.
I think a clear before and after (introduction of MMTS and similar) comparison of data quality (both at individual stations and across the network) would be useful.

Kev-in-Uk
August 1, 2012 1:25 am

”REPLY: Noooo…Fig. 3. Changes in the documented time of observation
in the U.S. HCN. is about the times they assigned the observers. There’s no proof the observers adhere to it. – Anthony”
Precisely! – but this demonstrates a couple of important points:
1) the (old) data we have has to be taken at FACE value only – massaging it for supposed/expected errors could and would likely only compound any real errors – especially if the supposed errors aren’t actually there at all, or are perhaps intermittent!
2) apart from for clearly obvious and well recorded ‘introduced’ errors, such as station relocations, instrument/thermometer changes, etc – actual data ‘alteration’ is rather silly IMO. In any data series, it would be better to always produce and maintain the data as separate sections (as required) of the series, such that a detailed graph of the alterations alongside (or on top of) a graph of the raw data could always be referred to – in a suitable scale so that the alterations can be clearly seen and visually judged to see if they appear appropriate. And there’s the other rub – climate variation isn’t a smooth graph – so how can you make an assumption that a sudden temp change did not actually occur in the historical record? – you have to be very careful in making those assumptions, presumably aided by comparing with nearby stations, etc.
3) Maintaining the data as individual sections, rather than trying to meld them together as a complete series must be considered paramount – that way, all and any subsequent adjustments can be reviewed and assessed as potentially valid or not, especially if they are applied as a ‘blanket’ type adjustment to a melded series. Constantly updating and joining data and calling it ‘good’ – is not the way to work.
4) Homogenising data between stations must surely be evident as potentially highly flawed if the base data is itself flawed? Gridding data and averaging, etc – even moreso!
As an old school scientist and engineer it seems to me that the original data is always to be preserved, even if just to hold it up as a pile of cr*p! Manipulation and reproduction ALWAYS introduces further errors – even if just from simple typos?….
If one produces a piece of work saying I’ve adjusted this to get that – it should always present the before and after case and detailed reasoning. It strikes me, that in the historical climate data sense, this is not adhered to – and we have different versions of the data, and presumably adjustments upon adjustments, etc, etc. The question now is – what and where the flip is the ‘real’ data?

Martin Lack
August 1, 2012 1:28 am

In my recent post on my blog, Anthony, I have conceded that there may be some validity to your criticisms (but so that you cannot accuse me of blatant self-promotion, I will not even attempt to include a link to it). However, the fact remains that the data you have examined relates to 2% of the Earth’s surface (within which you accept 50% of warming is real). Despite this, you seem to want the World to believe that you and your colleagues have uncovered the real story; and that the vast majority of climate scientists are insidious, incompetent, or simply imbeciles. But, be honest, how likely is that? Indeed, is it more or less likely that the WTC was brought down by a team of controlled demolition experts?
Also, if global warming stopped in 1998, perhaps you can also explain to me why May 2012 in the USA was… “the 327th consecutive month in which the temperature of the entire globe exceeded the 20th-century average, the odds of which occurring by simple chance were 3.7 x 10-99, a number considerably larger than the number of stars in the universe.” — Bill McKibbin (Rolling Stone magazine)

M Courtney
August 1, 2012 1:45 am

So what is the audit regime of the data collection?
How do the organisers check that the data is accurately reciorded at the right times.
I’m asking about the quality system that lies behind the data collection. There must be a written system. Does anyone know if it’s any good?

August 1, 2012 1:56 am

Christoph Dollis says:
July 31, 2012 at 6:00 pm: “You are correct, the pre-posting of this has been a remarkable success ….”
I think he jumped the gun by a week myself.

Thereby accruing seven days of outside review for error-spotting in data, methodology, descriptions and graphs; correction/explanation of ambiguous sentences, spelling, and grammatical usage; and addressing areas which may warrant additional scrutiny.
I think that was a pretty good idea, myself.

A. Scott
August 1, 2012 1:56 am

I re-read the report to identify the data referenced … here is the list I came up with.
I even included the NCDC station history metadata link in case you don’t want to do the extensive visual and/or onsite inspection Anthony and his help spent well over a year doing.
Seems everything is there to reproduce the work for the people who question the results. I did it on my own – would saved me half an hour if I’d simply read the References 😉
The USHCN Version 2 Serial Monthly Dataset page – scroll down to find the 4 data sets for each set of station ratings, along with the MMTS and Cotton Region Shelter (Stevenson) site information used for Menne (2010):
http://www.ncdc.noaa.gov/oa/climate/research/ushcn/
The NCDC Station Histories appear to be here:
http://www.ncdc.noaa.gov/oa/climate/stationlocator.html
WMO-CIMO endorsement of Leroy(2010) standard is here:
http://www.wmo.int/pages/prog/www/CIMO/CIMO15-WMO1064/1064_en.pdf
Leroy(2010) is here:
http://www.jma.go.jp/jma/en/Activities/qmws_2010/CountryReport/CS202_Leroy.pdf
Watts(2009) is here:
http://wattsupwiththat.files.wordpress.com/2009/05/surfacestationsreport_spring09.pdf
And Muller’s Station data is here:
http://berkeleyearth.org/pdf/berkeley-earth-station-quality.pdf
Fall (2010):
http://pielkeclimatesci.files.wordpress.com/2011/07/r-367.pdf
http://www.surfacestations.org/fall_etal_2011.htm
NOAA’s Climate Reference Network Site Handbook (see Sec. 2.21)
http://www1.ncdc.noaa.gov/pub/data/uscrn/documentation/program/X030FullDocumentD0.pdf
Watts Surface Stations Project site master list:
http://www.surfacestations.org/USHCN_stationlist.htm
(the brief notes should provide an initial screen of suspect stations)

John Wilbye
August 1, 2012 2:08 am

Pardon my ignorance here but is TOBS really a factor? If there is an error introduced by TOBS I suggest the same error will occur in those badly sited stations as in the better sited stations so the effect is nullified and therefore it does not need to calculated.

Terry
August 1, 2012 2:22 am

A Scott says “Why wouldn’t the TOB adjustment simply be a single, identifiable, “step” change, easily discerned in the data?”
As I understand it TOBs adjustment for a single change should not be a step change, it should be a spike. The subsequent Tmin Tmax range should then return to normal since you are still looking at the range for a single 24 period. The real problem is when the TOBs (which is recorded, faithfully you would hope) changes on a daily basis.

Eric (skeptic)
August 1, 2012 2:29 am

Smokey, do a google news (past week) search on “Richard Muller”. Then do the same search on “Anthony Watts”. The result is what you would expect, but you will probably be amazed at the orders of magnitude difference.

Alexej Buergin
August 1, 2012 2:41 am

Do I get that right: In the age of computers and automatic measuring the quality does not get better, but worse? And the time of observation was a problem during the last 30 years?
When I had a close look at US-Meteorology 20 years ago (aviation-related, at ERAU in Daytona Beach), I had the impression they knew what they were doing.

Slabadang
August 1, 2012 2:50 am

Introduction of “sommertime”!
It varies between coutries. In Sweden it was introduced in 1980. It was made uniform in EU 1996 and I havent found out how the other EU countrys and countrys outsida the EU have applied sommertime and I havent a clue how this change have influensed readings. It ought to be cinsidered by TOBS? Or is it irrelevant?

Arnost
August 1, 2012 3:27 am

If as Anthony says: “they stopped worrying about daily readings and simply filled them in at the end of the month by stepping through the display.”. Then this suggests that the Max/Min temps would be recorded on a strictly daily basis (i.e. midnight to midnight). This being the case when the old base station was replaced by the new Nimbus, the practice of relying on stored temp data would have resulted in effectively a change in time of observation from the morning to midnight.
And that would have introduced a warming bias.
I haven’t heard of anyone discussing this let alone adjusting for it. Is this an issue? Can it explain why Anthony is finding the lower trend?

Entropic man
August 1, 2012 3:37 am

Kelly Haughton says:
July 31, 2012 at 6:18 pm
“It will be interesting to see what the “peers” do for the peer-reviewed journal. How will the peers be selected? What will they say? Will they read all the comments?”
This should help you understand the peer review process.
http://www.agu.org/pubs/authors/pdf/PeerReview_Guide.pdf
.

Ian W
August 1, 2012 3:39 am

Using atmospheric temperature to assess the amount of heat in the atmosphere shows ignorance of physics.
Averaging atmospheric temperatures compounds the nonsense.
Arguing over the time of observation of the temperatures is akin to arguing over the color of the angels wings that are to be counted on the head of the pin.

August 1, 2012 4:02 am

I still don’t get why so many are hammering the TOBS thing. I have understood for years that reading the units at different times can make a minuscule difference.
But this paper is about the way the “adjustments” are made. The paper is about the huge amount of poorly sited units which read warmer due to location and then have even more upward adjustments made to them — and to beat all, the good sites are adjusted upwards to match the bad ones!
TOBS is a smokescreen in the context of this paper.

Peter Ellis
August 1, 2012 4:05 am

@Steve Huntwork For a pure daily cycle, then the time of observation is irrelevant.
The problem arises when warm/cold spells of a few days’ duration (i.e. weather) are combined with a daily wave form (i.e. diurnal cycle).
Consider what happens if you have a diurnal cycle varying from 10 degrees at dawn (say 5am) to 20 degrees in early afternoon (say 2pm), and you’re doing the readings at 4pm – i.e. in the afternoon shortly after the daily max. Each day’s min/max observations will come up as 10/20 degrees at 5am/2pm, and everything is fine.
But now think what happens when you have a warm spell lasting 7 days, which raises the overall temperature by 5 degrees so the cycle now goes from 15 degrees at 5am to 25 degrees at 2pm. For the first 7 days, you’re fine: it records the daily min/max as 15/25 degrees at 5am/2pm, exactly as it should. On the eighth day, you’ll record the min correctly as 15 degrees at 5am, since the warm spell is over and you’ve dropped back to the previous normal. However, the max will get recorded as something like 24 degrees at 4:01pm on the previous day; i.e. the last day of the warm spell gets double counted. A seven-day warm spell has contributed 7 times to the daily min, and 8 times to the daily max, and the average has therefore been subtly boosted slightly higher than it should be.
There’s no deliberate cheating or lies associated with this. The same issue can works in reverse: for stations taking observations near dawn, then the last day of any given cold spells will get double counted. The overall effect is in principle symmetric: stations taking readings at 4pm will tend to read higher than the “true” value since they double-count warm days, while stations taking readings at 4am will tend to read lower than the “true” value since they double-count cold nights. However, humans being humans, you don’t *get* large numbers of weather stations all taking their readings at 4am, whereas in the first half of the century 4pm was a common time to take daily readings. This applies particularly to rural stations with reduced accessibility. So, rural stations tended to slightly overstate temperatures in the first half of the record, and thus the raw values from these stations will underestimate the true warming trend – exactly as Anthony sees in this work.
The fact that the new station rating shows no difference in trends in the homogenised data is strong evidence that the homogenisation procedures correctly remove both UHI and time-of-observation bias.

Tucci78
August 1, 2012 4:10 am

So my point in all this is that NWS/NOAA/NCDC is getting exactly what they paid for. And my view of the network is that it is filled with such randomness.

Dear God.
And bear in mind that I’m an atheist.
How the hell can anyone come remotely close to quantifying such “randomness” in order to set some sort of value on the degrees of uncertainty in the whole “NWS/NOAA/NCDC” melange?
Billions of dollars have been spent on “research” to support the ginormous “We’re All Gonna Die!” AGW fraud and Mr. Watts reports that:

One woman begged me to talk to the local NWS office to see about getting a new thermometer mount for her max/min thermometer, since it wouldn’t lock into position properly and often would screw up the daily readings when it spun loose and reset the little iron pegs in the capillary tube.

Like most SF magazines, Analog actually reaches its subscribers as much as a couple of months before its stated publication date. The November 2009 edition was therefore in print and in the hands of readers long before FOIA2009.zip hit the ‘Net. In that edition – written long before the issue went to print – physicist and SF writer Jeffery D. Kooistra published his “The Alternate View” column (titled “Lessons From the Lab“) about Mr. Watts’ early SurfaceStations project report, Is the U.S. Surface Temperature Record Reliable? (May 2009).
Kooistra started by recalling his experience as an undergraduate, in which he learned something that all former science majors can readily appreciate:

When you work in experimental physics, you have it drilled into you that without proper calibration, at the end of the experiment you will have, as my professor one time screamed at me, no data.

He then went on briefly to discuss Mr. Watts’ project and recently published monograph, concluding:

I have long wondered why most of my fellow physicists haven’t been as skeptical of global warming alarmism as I have been. I think one reason, perhaps even more important than their politics affecting their judgment, is that they naturally assume other scientists are as careful in how they obtain data as physicists are. I’ve been a global warming skeptic for some time now, and it didn’t even occur to me that most of the time the thermometers would be “sited next to a lamp.” What’s really ironic is that, if someone claims to see a flying saucer, which hurts no one and costs nothing, debunkers come out in force. But let a former vice-president claim environmental apocalypse is upon us, and suddenly we’re appropriating billions and changing our lifestyles.
Cripes.

This particular “popular science” article in a pulp magazine helps explain why fans of “hard” science fiction – Analog‘s staple product – have tended overwhelmingly to be skeptical about the great AGW boojum since its inception, and why few of us were surprised (though most of us were delighted) by Climategate 1.0 on 17 November 2009.
We’ve been tracking this hoax-based hysteria among the mundanes for decades.
This understood, to learn that things are actually as bad in the “NWS/NOAA/NCDC” surface stations system as Mr. Watts describes…

Some local NWS personnel I talked to called the MMTS the “Mickey Mouse Temperature System,” obviously a term of derision. Wonder why?

…is to me the equivalent of hearing that the Mayo Clinic has been given over wholly to the practices of homeopathy.
Not that I don’t believe Mr. Watts, but it’s kinda difficult to get my hands around just how rotten this “global” pretense of precise climate knowledge actually is. It’s like something out of a sick and twisted horror novel.
And, remember, I’m a fan of “hard” science fiction.

Entropic man
August 1, 2012 4:31 am

davidmhoffer says:
July 31, 2012 at 6:26 pm
Entropic Man;
I see some papers to make one wince, too. Nevertheless, these are the ground rules that scientists and scientific journals go by.
>>>>>>>>>>>>>>>>>
Will there be more crap on the internet because it is accessible to all? Of course. Will a tiny cabal of politicaly motivated scientists be able to get editors fired and legitimate science suppressed?
Not a chance. Welcome to the new world.
————————————————————————————————————————–
If your baseball team wants to play in a league, you play by that league’s rules. If you want to publish, you go through peer review. Most scientists use it as a filter to keep their reading list manageable by winnowing out the worst of the rubbish.
To get some perspective on this problem, look at http://arxiv.org/
This archives e-prints of unpublished and unreviewed papers. Some are good science, put there to establish priority in advance of publication. Others are there because they bad science that will never reach publishable standard.
Without peer review there is no way for someone reading the papers online to distinguish between good and bad. Like all systems human peer review is imperfect, but better than nothing at all.
There is discussion of ways to update quality control of scientific papers in a digital age, though any solution would still have to winnow the wheat from the chaff.

Glenn Tamblyn
August 1, 2012 4:37 am

Several comments about TOBs.
An early paper by Schaal et al 1977 identified a 1.0 – 1.5 C cooling impact on the US temperature record due to TOBs changes. This was for the period 1910 to 1975 where a large percentage of the COOP observer network progressively changed. For each individual station, a single change in TOBs represents a single step change in bias.
Given the way the older system worked, with thermometers recording max & min temperatures, with the thermometers being reset when the observer takes the reading, having a reading schedule close to the time of day when either the maximum or minimum temperature occurs increases the likelihood that the value at that time may end up being counted for 2 days, whether max or min. The earliest schedule used by COOP, as reported by Schall et al, was to measure at midnight. This was an ideal time, a substantial number of hours away from the normal time of max & min. The schedules then progressively shifted to 7 AM & 7 PM. 7 AM, a particularly problematic time. In winter, in darkness, the minimum can easily occur around then. So you can easily record a minimum twice because the thermometers don’t tell you when the minimum occurred.
Schall et al reported that by 1975, 55% of the COOP network had changed their schedules, up from 10% that were on that schedule in 1910.
1975 is near the starting point of this study. And the MMTS units weren’t being rolled out till decades later. What proportion of the 45% of COOP sites that hadn’t switched their TOBs already did so over the study period.
Then the introduction of MMTS automatically constitues a TOBs change. Now it isn’t getting max/min based on observation time, but max/min for a particular day. So intrinsically it is cancelling out and previous TOBs bias, which is itself a TOBs bias.
Another question or issue. Is there any correlation between station siting ratings and TOBs bias. Is there any pattern in which stations experienced TOBs changes, and does this correlate with station ratings. Could stations with poor siting due to being in less pristine environments have been more or less likely to experience a TOBs change. For example, in an environment that is getting built up, could the observer for that site be someone who now works nearby and takes the reading during office hours, whereas in the past they where someone who lived nearby and sampled before going to work.
A final comment on TOBs biases and how they can be important. John Christy is listed as a co-author. John has years of experience with the Satellite Temperature record at UAH and how it is analysed. One significant issue that was addressed nearly a decade ago was the problem of Diurnal Drift. The satellites are in orbits that are meant to be Sun Synchronous. As they fly over the Earth they are meant to pass over each point below at either solar Noon or solar Midnight. To do this their orbits are given a planned precesion so that over the course of a year thier orbit slowly changes to remain Sun Synchronous. However the satellites don’t have thrusters to adjust their orbits so over time due to atmospheric drag and the fluctuation of gravity as they pass over different parts of the Earth, their orbits deviate and they are no longer quite Sun Synchronous. So they now pass over the Earth at times different from Noon & Midnight. This is Diurnal Drift. UAH (and RSS) developed methods to adjust for this. And the results made substantial differences to the trends reported by both groups, adding a significant amount to the trends. Diurnal Drift mattered!
Diurnal Drift is a TOBs bias!
So I wonder what input John had as a co-author in advising on just how important it was to consider TOBs biases.
The simples approach that the authors could take in the short term, until perhaps a more detailed cross-referencing of TOBs vs other factors can be done for each individual station is to cofine comparisons to looking at different station categories within only individual data sets. Compare only Raw against Raw, TOBs adjusted against TOBs adjusted, Fully adjusted against Fully adjusted. When comparing across the data sets you simply can’t have any confidence in what you are reporting. Ascibing differences to one factor when multiple factors are involved and haven’t been separately evaluated isn’t worth very much.
For example, the paper comments that good quality stations have been given warm afjustments to bring them into line with the poor quality stations. What about the possibility that the good quality stations had a TOBs change before the start of the survey period, the 55% reported by Schaal. Perhaps the good quality stations were more likely to be given MMTS upgrades compared to the poor quality ones.
I don’t know the answer to these questions. But nothing written in the paper says the authors have any idea either. And without being able to evaluate this, the conclusions of the paper simply cannot be backed up from what they have presented.The methodology is currently flawed.
If the authors wish to report some of the findings from applying part of the required methodology, fine. But this is extremely preliminary data that could change radically when all the relevent factors are considered.
So press releases and such claims to the effect that “Our findings show…..” are simply, totally unjustified. The findings don’t show anything yet; they are too preliminary.

Stephen Richards
August 1, 2012 4:43 am

REPLY: What Zeke illustrates is what I see as the biased nature of the whole network. There’s so many problems that all sorts of issues have to be invented to deal with it, and all the corrections they invent go UP. And like NCDC, Zeke’s trying to find the answers in the data, rather than looking at the reality of the measurement environment. Nobody wants to deal with that side of it. Color me unimpressed with the whole adjustment crusade. – Anthony
Anthony
Colour me disgusted.

Steve S
August 1, 2012 5:02 am

Terry,
My concern exactly. If a change in observation time did introduce a step change in daily TMAX/TMIN observations, wouldn’t that be obvious in the raw data? Wouldn’t the accumulated TOBS biases over time tend to go back and forth, rather than requiring an additional bias be applied to the data that always seems to drive the temperature up?

tonyb
August 1, 2012 5:07 am

Last year I wrote this article- carried here- which seems highly relevant to this subject; The following comments are taken from a book I referenced for the article, written by a famous climatologist over 100 years ago
Extracts;
“If the mean is derived from frequent observations made during the daytime only, as is still often the case, the resulting mean is too high…a station whose mean is obtained in this way seems much warmer with reference to other stations than it really is and erroneous conclusions are therefore drawn on its climate, thus (for example) the mean annual temperature of Rome was given as 16.4c by a seemingly trustworthy Italian authority, while it is really 15.5c.”
That readings should be routinely taken in this manner as late as the 1900′s, even in major European centers, is somewhat surprising.
There are numerous veiled criticisms in this vein;
“…the means derived from the daily extremes (max and min readings) also give values which are somewhat too high, the difference being about 0.4c in the majority of climates throughout the year.”
Other complaints made by Doctor von Hann include this comment, concerning the manner in which temperatures are observed;
“…the combination of (readings at) 8am, 2pm, and 8pm, which has unfortunately become quite generally adopted, is not satisfactory because the mean of 8+2+ 8 divided by 3 is much too high in summer.”
And; “…observation hours which do not vary are always much to be preferred.”
That the British- and presumably those countries influenced by them- had habits of which he did not approve, demonstrate the inconsistency of methodology between countries, cultures and amateurs/professionals.”
http://wattsupwiththat.com/2011/05/23/little-ice-age-thermometers-%E2%80%93-history-and-reliability-2/
That there have been problems with the instruments AND the manner in which they are read can be traced throughout the instrumental record. The BEST record to 1750 seems to be based on the data of around ten heavily adjusted stations to 1750 and is neither remotely ‘global’ nor, with the huge error bars, especially scientific-although it adds a little to our general knowledge which is to be welcomed and those involved in the number crunching to be congratulated on their diligence.
I can not really comment on the current Watts et al paper in detail as it seems to be (quite rightly) a work in progress, but I am sure Anthony-with his background-knows of the problems as much as anyone, whether you are dealing with modern or historic records.
Add in all the problems described in my article and the complications of siting, uhi, station move etc etc and i have come to believe over the last five yearsthat it is very foolish for politicians to bet the house on a CAGW scenario that-looking at history-is very difficult to see as based in reality
tonyb

rogerknights
August 1, 2012 5:16 am

Peter Ellis says:
However, humans being humans, you don’t *get* large numbers of weather stations all taking their readings at 4am, whereas in the first half of the century 4pm was a common time to take daily readings. This applies particularly to rural stations with reduced accessibility. So, rural stations tended to slightly overstate temperatures in the first half of the record, and thus the raw values from these stations will underestimate the true warming trend – exactly as Anthony sees in this work.
The fact that the new station rating shows no difference in trends in the homogenised data is strong evidence that the homogenisation procedures correctly remove both UHI and time-of-observation bias.

But Anthony is measuring only from 1979, not 1900.

Sjoerd
August 1, 2012 5:23 am

I understand why tobs leads to a bias; I just don’t see how one can correct for it using the available data.
Consider this scenario.
During the winter a shifting wind or passing front will lead to a bigger shift in temperature than the diurnal cycle. This is the case for the Netherlands; I bet there will be locations in the USA with a simular situation.
As a result, when the front passes, the warmest moment during the day can be at any point during the day. It’s not uncommon that when a cold front moves in around dawn, that the warmest moment was during the night! And on the other hand, when a warm front moves in during the night, it might be warmer at night than it was during the day.
The problem is that one can’t judge from the min/max values at which time the front moved in. Did the cold front move in at dawn, so the maximum is the maximum of the previous night (which could be the maximum at 8:01 pm on the previous day)? Or did the cold front move in at 1 pm, which would result in a ‘proper’ Tmax at 0:59 pm? In the former case, Tmax should be adjusted, in the latter it shouldn’t!
It seems impossible to me to adjust for this situation, without knowing the actual timing of the arrival of the front (which is not recorded).

Rob J
August 1, 2012 5:30 am

Someone please explain something to me. When looking at Christy’s satellite data for the lower 48 I see about 0.3 degrees per decade warming for the same time period of Anthony’s study which is pretty much in line with NOAA/NASA/BEST numbers and about twice as large as Anthony’s data for “good” sites. I assume that the satellite data is not fudged. So why are the satellite numbers in much better agreement with the supposedly faulty NOAA/NASA/BEST numbers?

rogerknights
August 1, 2012 5:32 am

Here are some fixes needed for the Powerpoint presentation. Probably most of them have already been suggested:
P.2, there’s a dot missing in the leader before “Page 2” and there’s an extra space in the next line before “Page 3”
The following words are red-underline-flagged as misspellings by MSFT. They should be added to the spell-check dictionary: gridded, microsite, mesosite, vs., inhomogeneity, vanishingly
PP. 23, 33, 42, 52: Delete “actually” from “What the compliant thermometers actually say:”. It’s argumentative innuendo–or could be seen as such. (And it’s not in sync with the usage on p. 14.)
PP. 13, 22, 32, 41, 53: “poor station” should be “poorly sited station” for parallelism with “well sited station”. Also, “well-sited” should be hyphenated. (But not “poorly sited”!) Also, a closing “)” is missing.
PP. 10, 19, 29, 38, 48: Delete “over” (redundant) from: “cover over 50% or more”
In your main paper, you should request NOAA to make available pictures of the sites you weren’t able to access. You might twit them for not doing so earlier.

AndyG55
August 1, 2012 5:35 am

So many ” we don’t know for sure” bits and pieces to the puzzle..
Why not just admit it and put error bars +/- 5C on the whole shebang !!!!

Maus
August 1, 2012 5:35 am

Peter Ellis: “The overall effect is in principle symmetric: stations taking readings at 4pm will tend to read higher than the “true” value since they double-count warm days, while stations taking readings at 4am will tend to read lower than the “true” value since they double-count cold nights.”
You were doing fine right up until this point. This is not the case for a single site. It can be the case if someone is so studiously ham-fisted as to attempt to compare two different sites, at two different locations, with two different local conditions, measured at two different times, directly as if they were synonymous with one another under the notion that the Tmin/Tmax pair recorded are somehow valid time pertinent data points rather than extremums of a range collected over the previous 24 hours. And yes, some of the professional eggheads are just this studiously ham-fisted.
The problem is not TOBS. The problem is people trying to force the data to say things that it cannot. Thus far every presentation of the ‘problem’ and the ‘need for a cure’ has relied on trying to turn a Tmin/Tmax pair that occur at some unknown time during the prior 24 hours into instantaneous measurements taken at specific times. The data is simply not there for such shenanigans. Taking hourly measurements will not give a valid Tmin/Tmax, and Tmin/Tmax measurements will not give valid hourly values.

August 1, 2012 5:37 am

It has been asserted that NOAA adjusted the data of rural stations upwards to match the urban stations. However, it is countered here that the adjustments were made because a change in thermometer type meant that readings fell:
“”In the press release it is also emphasised that the temperature trend after homogenization is stronger than in the raw data. Maybe Mr Watts thinks this is new, but, e.g., Menne et al. (2009) already stated that the introduction of automatic weather stations (the transition from Liquid in Glass thermometers to the maximum–minimum temperature system) caused a temperature decrease in the raw data of 0.3 to 0.4 °C. This temperature jump has to be and was removed by homogenization.”
The increase in the temperature trend is thus not due to adjustment of stations with a low trend to the ones with a strong trend, but due to the change in the way the temperature is measured, the transition from LiG to MMTS and also probably due to a change in the time of observation. Homogenization removes these artificial jumps and because they caused artificial cooler temperatures, the homogenized data shows a stronger trend. There is no evidence in Watts et al. that the good stations are adjusted to the bad ones. Watts et al. does not even study how homogenization algorithms function.”
Source: http://variable-variability.blogspot.co.uk/2012/07/blog-review-of-watts-et-al-2012.html
I hope this helps.

Espen
August 1, 2012 5:56 am

John Wilbye says:
August 1, 2012 at 2:08 am
Pardon my ignorance here but is TOBS really a factor? If there is an error introduced by TOBS I suggest the same error will occur in those badly sited stations as in the better sited stations so the effect is nullified and therefore it does not need to calculated.
Seems reasonable – but it needs to be investigated whether there is some correlation here (what if, for instance, badly sited stations have more TOBS issues because they in general have been more sloppily maintained?). Anyway, as long as the TOBS issue is as full of pitfalls as it is, I think Anthony did just the right thing by leaving it out of this report.

Dr. Deanster
August 1, 2012 6:34 am

Given my limited knowledge on TOBS .. it would seem to me that the ajustment would be highly station dependent, and thus, no blanket adjustment could be credibly done.
Where I live, it would seem missing the “low” would be more likely than missing the “high”. Things warm up, max around 3 pm, but cool off very slowly. So a temp reading of 4 pm would not be much different, if at all. In contrast, the min, happens just before sunrise, and warms up rapidly. Taking the min an hour late would result in a false “warm” reading.
Further, Fronts happen. Sometimes the Max is at Midnight, and the Min is at 5 pm.
Since I notice that all TOBS adjustments seem to increase the warming trend, it seem pretty obvious to me that this whole adjustment has an “end” as its goal .. ie., to create a greater warming trend that exists.

Keitho
Editor
August 1, 2012 6:40 am

I suppose with something as big as the climate sea there are bound to be a number of red herrings but really. The effect that Anthony has shown is not altered by TOB’s , by error bars, by new and improved snazzy mathematics. It simply shows how the different classifications of reporting stations have differing rates of temperature increase. He shows that this is an artifact of the methodology used by NOAA. This is important because a lot of research has been done on finding out why the NOAA data shows what it does and the data is wrong which means the research was often looking in the wrong place for the wrong answers. Much wasted time and effort.
By all means let’s look at the other niggles just so we can be certain they don’t, collectively, compensate for the error found by Watts et al. and thus make NOAA correct once more. Now that really would be a 5 sigma result worth considering. Ah, CO2 you amazing little molecule is there nothing you can’t do? /sarc

wsbriggs
August 1, 2012 6:48 am

I have an issue with the idea that we will record the min/max within a single 24 hour period every day, and then complain and insist that if the next day’s max occurs within one second after the end of the previous 24 hour period, that it is invalid. When you sample, you sample. Yes, the max for the previous 24 hour period was one second prior, but we are asking for the max temp within a 24 hour period. They fall where they fall. Humans would experience the 24 hour period exactly as the min/max thermometer does. It got cooler after X AM/PM. Arguing that the max was double counted seems to be arguing that we really don’t want to know the 24 hour min/max. This doesn’t change with continuous sampling – the max was still 1 sample after the 24 hour reset.
To exaggerate, during a 24 hour period, a micro downburst occurs in the vicinity of the instrument chilling it substantially below the temperature before and after the event. The thermometer records the min temp faithfully. Should we nuke the temp? A human standing there would have felt it. NOTE: this is NOT the same as a jet exhaust, it is a meteorological event.

August 1, 2012 7:01 am

A. Scott: Seems everything is there to reproduce the work for the people who question the results. I did it on my own
You were able to recreate the list of USHCN station ids and their associated Leroy 2010 classification? Well done! Perhaps you can sure the list with the slow half of the class.

dearieme
August 1, 2012 7:03 am

On the first large industrial experiment I organised, I strolled into the control room at 12:15 and looked at the record sheet: the results for 13:00 had already been written in. Such is mankind.

Rattus Norvegicus
August 1, 2012 7:08 am

The existence of a TOB has been known since 1848! Also look at the link in the comment this jumps to, it has a substantial list of papers on this bias, with the earliest one being 1890!
http://variable-variability.blogspot.com/2012/07/blog-review-of-watts-et-al-2012.html?showComment=1343822752641#c3498478472825631256

Entropic man
August 1, 2012 7:20 am

I know a rural station, not in the US, notorious for underreading because it is in a frost hollow. Winter Tmin values are consistently lower than others in the region, giving distorted winter daily temperatures and the lowest annual average temperature for 200 miles.
On the basis of an estimated error of about -2C relative to its neigbours it would be classified Class 3, but otherwise meets most of the requirements of Class 1. Reading through the class descriptions, they all discuss warming effects such as parking lots or air conditioners, but even shading is barely mentioned. I site such as this is not catered for at all.
Is it me, or are both the USHCN and surfacestation.org exposing themselves to sampling bias assuming in their classification that all low class stations overread? The TOBS, MMTS and SHAP corrections all assume that the temperatures need to be adjusted upwards, which would make the error in this station even worse!
In a 20 minute sample search of USHCN and surfacestations.org I found a number of stations classified as low quality due to overreading, but nothing to match the station described above.

Bill Yarber
August 1, 2012 7:32 am

Anthony
I remember seeing a graph a couple of years ago showing the number of reporting stations vs the temperature trend. It indicated a significant drop in reporting stations (most of them rural if I remember correctly) around 1990 after the collapse of the USSR, and a corresponding (though possibly not correlated with) increase in the temperature values. I realize that most of the decrease in station count was not in the US, but did your team consider the impact of fewer rural stations in your paper? A quick scan did not indicate that this change was mentioned in your article. Since you show that the rural temperature increase has been much less than the other areas and the “adjusted” final tempearture trends, I have two questions:
1) Is the impact of the reduction in rural stations in the US statistically significant?
2) If the rural trend is so much less than the trend from the other stations, why is the UHI adjustment only 0.05C? Seems like it should be on the order of .5C or more.
Bill

Bill Illis
August 1, 2012 7:36 am

I challenge those who have been able to download the Raw and Adjusted USHCN V2 data, to publish the (annual or monthly) numbers.
Steven Goodard has here.
http://stevengoddard.wordpress.com/2012/07/16/how-ushcn-hides-the-decline-in-us-temperatures/
So that results then in total Adjustments of 0.84C to the trend in USHCN V2.
– TOBs only increases the trend by +0.399C (from 1934, funny how that year has the biggest change); but,
– Other adjustments are +0.500C.
I’ve pinned all the data to 2011 so that we can actually see the total change (while others have produced charts that obscure the changes by using a baseline averaged in the middle).
Total Adjustments first.
http://s15.postimage.org/wzzrg9k0b/Total_Adjust_USHCNV2.png
Raw, TOBs and Final annual numbers next.
http://s16.postimage.org/ux7avs6qt/USHCNV2_Raw_Final.png
Now perhaps Steven’s numbers are off some. I used the most recent Final numbers from the NOAA to do the charts and they were slightly different.
But all those who have the Raw, and Final data should make the data available so that we can see just how much these adjustments add to the trend.

Bill Illis
August 1, 2012 7:37 am

Sorry TOBs adjustment above should say +0.339C (not 0.399C).

geo
August 1, 2012 7:37 am

Yeah, I ran into one of those consumer weather station in the CRS types in the Dakotas, at a sugar beet processing plant. Temps were taken by the guards at the guard station. One told me the issued thermometer developed a bubble, and they asked NWS for a new one, but it wasn’t forthcoming, so they just went out and bought their own. Tho, coincidentally, I’m sure, this happened in a Dakota’s February.

geo
August 1, 2012 7:38 am

Wireless, of course, and the guard station itself was indoor/heated.

donald penman
August 1, 2012 7:47 am

TOBS is i have been reading not the time of observation but the 24 hour period in which the Tmax and Tmin occurs and every 24 hour period has a Tmax and Tmin so the data can be adjusted to refer to the same 24 hour period without I think introducing a trend in the data.

beng
August 1, 2012 7:58 am

Seems to me TOBS is complicated. A tropical or lower-latitude site would have less bias w/the same observation-time changes as a more poleward site affected by fronts because low-latitudes would have more consistent min/max time-occurrences compared to front-affected sites. How does one quantify those different situations? One would have to spot each & every occurrence of double-reporting for each site to get it right. Estimating algorithms would be prone to the usual inadequacies, tinkering and abuse.
Yeah, I’ve briefly looked at the methods used for TOBS, and that’s why I make the remark about their efficacy.

beng
August 1, 2012 8:13 am

****
Entropic man says:
August 1, 2012 at 7:20 am
I know a rural station, not in the US, notorious for underreading because it is in a frost hollow. Winter Tmin values are consistently lower than others in the region, giving distorted winter daily temperatures and the lowest annual average temperature for 200 miles.
****
If the station is otherwise consistent/unchanging, being in a frost hollow has no effect on the trend.

Christoph Dollis
August 1, 2012 8:34 am

“How the hell can anyone come remotely close to quantifying such “randomness”

There’s always going to be some measurement uncertainty. It’s just a fact of life in science. Quantifying it is part of science.

Maus
August 1, 2012 9:06 am

donald penman: “TOBS is i have been reading not the time of observation but the 24 hour period in which the Tmax and Tmin occurs and every 24 hour period has a Tmax and Tmin so the data can be adjusted to refer to the same 24 hour period without I think introducing a trend in the data.”
Yes, that is one of the ham-fisted notions for it. But the notion for a trend is what now for climate? Last I remember was 13 or 17 years. But let’s say 13 years and that we’re dickering over a temporal jitter of +/- 12 hours with respect to solar noon. That’s a +/- 12 hours jitter over 133,880 hours covering 4,745 samples. Not counting leap years because I’m lazy.
If that’s enough to scrobble up your trend? Buddy, Jeane Dixon and Uri Geller don’t need that of much handicap to make predictions.

Entropic man
August 1, 2012 9:17 am

Bill Illis says:
August 1, 2021, 7.36am
“(from 1934, funny how that year has the biggest change)(while others have produced charts that obscure the changes by using a baseline averaged in the middle).”
You see a lot of that from both extremes in the climate change war. I have seen cherrypicked data used to support arguments for extreme warming , and for none. My favourite is the people who tell me there’s been no global warming for 15 years, starting from just before the exceptional El Nino year of 1998/sarc.
http://www.drroyspencer.com/wp-content/uploads/UAH_LT_current.gif

Entropic man
August 1, 2012 9:29 am

beng says:
August 1, 2012 at 8:13 am
“If the station is otherwise consistent/unchanging, being in a frost hollow has no effect on the trend.”
Fair enough, though it does make that region, with only two dozen stations, appear cooler than it is.
My main concern is that the apparant lack of frost hollow stations recognised in USHCN or surfacestations.org is distorting both datasets.

JJ
August 1, 2012 9:31 am

Martin Lack says:
Also, if global warming stopped in 1998, perhaps you can also explain to me why May 2012 in the USA was… “the 327th consecutive month in which the temperature of the entire globe exceeded the 20th-century average, the odds of which occurring by simple chance were 3.7 x 10-99, a number considerably larger than the number of stars in the universe.” — Bill McKibbin (Rolling Stone magazine)

May 2012 was warm relative to the 20th century average because the earth is warm relative to the 20th century average. “Warm” is not the same as “warming”. If the temperature of the earth stays the same as it is today for the next 100 years (i.e. warming stops) the temperatures recorded in 2112 will still be warm relative to the 20th century average. The odds presented by Bill McKibbin are wrong, and the conclusion that he draws from them is a non-sequitur which would not be correct even if the odds he presented were right.
Bill McKibbin is an idiot, and one should look to Rolling Stone for commentary on drugs and music, not science and math.

August 1, 2012 9:32 am

beng says:
August 1, 2012 at 8:13 am
“If the station is otherwise consistent/unchanging, being in a frost hollow has no effect on the trend.”
I agree and if the station is atop a 6288 ft mountain it has no effect on the trend:
http://weather-warehouse.com/WeatherHistory/PastWeatherData_MtWashington_MountWashington_NH_August.html

Christoph Dollis
August 1, 2012 9:42 am

“That’s a +/- 12 hours jitter over 133,880 hours covering 4,745 samples.”

If that’s all it was — that a given station might be half a day off — then no one would care. But it isn’t just that. It’s that day-in, day-out, the time the daily minimum/maximum-recording thermometers are reset can bias the results.
Anthony makes a good point about how, particularly in light of newer thermometers that need to be read less often, human nature causes people to have fudged the forms instead of paying attention to strict timing, but that’s a separate issue.

Ged
August 1, 2012 9:44 am

What I want to know is why TOBS adjustments always go up. TOBS should balance itself out. If you can go one way with the bias, you can go the other way with the bias. And over a large sample size, the stochastic effects will just nullify. But you never see a downward adjustment.
It seems completely arbitrary. Where is the empirical, experimental evidence?

Brian H
August 1, 2012 9:44 am

Places with variable weather can throw min/max thermometer assumptions out the window. If a cold day is followed by an overnight (say, 2 a.m.) warm front moving through, the MIN will still show the cold day’s measurement, even though it might be much warmer by the following sunrise. IMO only continuous plots, with frequent data point recording, can get around this. And then the whole “MIN/MAX issue is almost moot, because now you have real integrated temperature plots and actual averages, not stupid medians, are available.

John W
August 1, 2012 9:57 am

On NPR……
Global Warming Debate Heats Up, Again
Of course, when a team member defects, his/her ex-teammates shoot back quickly. A day after Muller’s op-ed piece, Anthony Watts from the privately-owned IntelliWeather, posted a piece in his blog Watts Up With That, where he claims that the data of U.S. temperature trends shows a spurious doubling due to station measurement issues.
http://www.npr.org/blogs/13.7/2012/08/01/157659554/global-warming-debate-heats-up-again

Stephen Pruett
August 1, 2012 10:05 am

” There is no evidence in Watts et al. that the good stations are adjusted to the bad ones”
There is no evidence that there was intent on the part of NOAA to adjust good stations to agree with bad ones, but the methods they chose had that effect. I would hope this is something they would want to examine and understand.

James Humbolt
August 1, 2012 10:10 am

Has a table been made available with the drafts with a list of 1+2 stations so Joe Blow can download the raw time series for each station?

Maus
August 1, 2012 10:11 am

Christopher Dollis: “If that’s all it was — that a given station might be half a day off — then no one would care.”
That one’s on me. I didn’t mention in that post that it was is in respect to aligning times between multiples stations. For a single site the notion would be meaningless, of course.
“But it isn’t just that. It’s that day-in, day-out, the time the daily minimum/maximum-recording thermometers are reset can bias the results.”
Sure, iff the temporal variance is enough to cause problems. But there are two time tested solutions for these problems. One does not alter the data recorded, nor the trend present in the data at all: Error bars. Completely safe and cheap to model. But terrible for the Chartporn Powerpoint of Doom: “On the side of the Titanic, that big freakin’ ship right there, there is a porthole that will cause a catastrophe. Which one? Can’t tell exactly, but there’s a lot of those tiny little buggers bolted right up. Smug bastards.”
The other will, like any adjustment to the data, bias the trend. Note: The data itself does not bias the trend, it *is* the data and we have absolutely no other knowledge. It’s a simple one-line in any statistical package and it’s called a moving average. If we’re concerned about yesterday in today and a part of today into tomorrow? 2 item moving average. Well known, needs no defense in general, alters the trend just as will including unicorns farts in the daily temp for stations in Wichita.
This isn’t pointed at you, mind. But if anyone wants to justify why we need unicorns farts in cornfields to homogenize the cold prairies up to the steam pipes under NYC then they need to first justify why completely harmless and common methods used even by the likes of Mann, Briffa and other notables (And unnotables, if you wish to eyeball the mere Weathermen running around here.) are wholly unsatisfactory corruptions of the empirically measured data.
And no, just no. A childhood allergy to error bars isn’t an excuse. Not even if you’re a Climate Scientist. Not even if you have a Nobel Prize. Not even if you have a Doctor’s note.
Without such justifications as to why every statisticians has gotten wrong for a century then it’s not data, not a theory, not science, but a product. But then so are used cars.

August 1, 2012 10:24 am

Christoph Dollis says:
August 1, 2012 at 9:42 am
Anthony makes a good point about how, particularly in light of newer thermometers that need to be read less often, human nature causes people to have fudged the forms instead of paying attention to strict timing, but that’s a separate issue.
================================
“Fudging”
Way back in the day when I was the opening person for the college radio station, one of my jobs was to go outside and get the weather data. (if my memory serves me correctly, it may have been a Stevenson screen). And since I really didn’t care how accurate the information was on many very,very cold days, I can tell you that there was much fudging.

Anonymous Coward
August 1, 2012 10:25 am

I find this discussion about TOB and other data adjustments utterly confounding. There are something like a dozen peer reviewed papers on TOB adjustments alone – you can find a bunch of them here:
http://agwobserver.wordpress.com/2012/08/01/papers-on-time-of-observation-bias/
It seems to me that if commenters on this site really wanted to understand data adjustments made for TOB, they would start by looking at the actual scientific literature on it, and actually try to understand how and why the adjustments are made.

Brian D
August 1, 2012 10:28 am

Well, Anthony, it was a late for ya last night. But I have an idea. Your comment concerning the human factor and how folks actually do the readings could turn out to be another mission. Create a post specifically for folks who are now doing or have in the past been observers. Kind of a online survey. Have pertinent questions along with the freedom to comment. Do a press release calling out to all to come and post on how they did their readings.(Facebook is another good avenue) It could be a very worthwhile endeavor. Just a suggestion. I really think we’d all be very surprised at the results. People are people, not robots. I know if I knew I could read an MMTS just once in awhile because of its memory capacity, I would. You’d be getting the temps you need, and it wouldn’t be a burden on me in the midst of a sometimes chaotic life.

Rud Istvan
August 1, 2012 10:39 am

Anthony, to the extent the US instrument siting problem is global, then there should be a growing divergence between satellite measured global temperatures, global surface temperatures, and global land temperatures (for example in GISS). I cannot paste in the chart here, but comparing UAH to HadCrut to GISS land only makes the case very nicely from 1980 (assuming MMTS started in the mid 1980’s). Did not run the exact statistics yet, but from 1995 to 2005 the divergence between UAH and GISS land is on the order of 0.3C, and on the order of 0.15C to HadCrut. Basically supports your conclusions about compromised data incorrectly homogenized in the wrong direction.

Frank K.
August 1, 2012 10:42 am

Brian H says:
August 1, 2012 at 9:44 am
“Places with variable weather can throw min/max thermometer assumptions out the window. If a cold day is followed by an overnight (say, 2 a.m.) warm front moving through, the MIN will still show the cold day’s measurement, even though it might be much warmer by the following sunrise. IMO only continuous plots, with frequent data point recording, can get around this. And then the whole “MIN/MAX issue is almost moot, because now you have real integrated temperature plots and actual averages, not stupid medians, are available.”
Brian H. – I’m with you. I’ve thought about this TOBS issue, and my impression is that one can’t really know very much about a continuous function (ambient temperature at a point in space versus time) given two values measured within a 24 hour window at unknown times (i.e. we don’t know exactly when the max and mix temperatures occurred). Why not simply plot the max and min temperature series as two separate discrete functions (no TOBS needed) and estimate the trends for each. I’m sure someone has done this.
Also, if one is considering the modern era of say 1960 to the present, we should have plenty of hourly data with which one can compare with nearby climate monitoring stations, again allowing the TOBS adjustment (if any) to be more precisely determined for the climate min/max readings. Maybe someone has already done this as well (and I confess I haven’t read all of the TOBs papers save for ones by Karl et al.).

Entropic man
August 1, 2012 10:54 am

J. Philip Peterson says:
August 1, 2012 at 9:32 am
beng says:
August 1, 2012
“If the station is otherwise consistent/unchanging, being in a frost hollow has no effect on the trend.””I agree and if the station is atop a 6288 ft mountain it has no effect on the trend:
http://weather-warehouse.com/WeatherHistory/PastWeatherData_MtWashington_MountWashington_NH_August.html
___________________________________________________________________
This is definately a rural site, unless someone has built a restaurant up there!
Just for amusement I calculated the ten year average, 1948 to 1957, and compared it with the ten year average for the most recent 10 years. The older mean is 47.1C and the newer 48.86C.
The trend is +0.27C per decade.

Entropic man
August 1, 2012 11:08 am

My God , they did build a restaurant up there!http://www.mountwashington.org/about/visitor/summer_visits/expect.php

August 1, 2012 11:11 am

Entropic man says:
August 1, 2012 at 10:54 am
J. Philip Peterson says:
August 1, 2012 at 9:32 am
beng says:
August 1, 2012
“If the station is otherwise consistent/unchanging, being in a frost hollow has no effect on the trend.””I agree and if the station is atop a 6288 ft mountain it has no effect on the trend:
http://weather-warehouse.com/WeatherHistory/PastWeatherData_MtWashington_MountWashington_NH_August.html
___________________________________________________________________
This is definately a rural site, unless someone has built a restaurant up there!
Just for amusement I calculated the ten year average, 1948 to 1957, and compared it with the ten year average for the most recent 10 years. The older mean is 47.1C and the newer 48.86C.
The trend is +0.27C per decade.
____________
Which month did you calculate for? I made several graphs in excel for ‘Mean Temperature’ for the months of Jan., April, June and Oct. and only April showed an increase from 1948 to 2012 of approx. 3 degrees F max. The other 3 months were even or even trended down.

Frank K.
August 1, 2012 11:21 am

Entropic man says:
August 1, 2012 at 10:54 am
“This is definately a rural site, unless someone has built a restaurant up there!”
Actually…there is a cog railway to the top, an auto road to the top, and a visitor’s center at the top, complete with gift shop. 🙂 Definitely worth a visit if you’re visiting New Hampshire (except in the winter, of course). Bring your jacket…

August 1, 2012 11:29 am

A critique of the article from Skeptical Science:
http://www.skepticalscience.com/watts_new_paper_critique.html

Hu McCulloch
August 1, 2012 11:32 am

The Nimbus thermometer remembers the highs and lows for
the last 35 days and also records the times they occurred.

This seems to imply a midnight effective TOBS unless the recorder goes to the trouble to convert it to a different effective TOBS.
Midnight will tend to double count cold nights, and therefore be similar (but not identical) to 7AM.

beng
August 1, 2012 12:00 pm

There’s a good comment at CA w/links to PielkeSr studies on the TOBS. As you would expect, the TOBS adjustments made by NCDC don’t make alot of sense…
http://climateaudit.org/2007/02/16/adjusting-ushcn-history/#comment-79771

Maus
August 1, 2012 12:46 pm

Anonymous Coward: “I find this discussion about TOB and other data adjustments utterly confounding.”
What’s confounding is that you didn’t bother to skim the abstracts of the papers at your link. A quick run through of the worst offenders:
Vincent (2009): “Not all days are affected by this change in observing time, and the annual percentage of affected days ranges from 15% for locations in the west to 38% for locations in the east.”
Playing with the reset time does not cure epistemic ignorance of when then highs and lows occur, or solve the problem that they occur at difference times on different days. Makes the same error of state that the midpoint of a range is a ‘mean’ temperature for a day. This is wholly illiterate with respect to basic mathematics.
Janis (2012): ” Surprisingly, non-calendar-day observations are similar to calendar-day observations on a majority of days. When differences do occur, however, they can be large and of either sign.”
Time of Tmax and time of Tmin vary, but are unknown. We already know that, but we cannot conjure up a stopwatch that does not exist. Not a problem of data collection but PEBKAC. Trying to ‘move’ temperatures out of a 24 hour window into some other time frame in absence of knowledge.
DeGaetano & Knapp (1993) : “Simulations using hourly temperature data indicate that the period required to reach a given GDD threshold value during a growing season often varies by 2 weeks or more solely because of observation time differences. Previous work on methods to adjust such biases has concentrated on seasonal totals and long-term averages. ”
Past work is based on the long term average of the midpoint of a range. Invalid for the purpose D&K are after, good enough. But the GDD is based on the ‘mean’ temperature of the day as purported to be the midpoint of the range. Again, mathematical illiteracy.
Karl et al (1986) : ” Using seven years of hourly data the standard errors of estimate using the model were only moderately higher than the standard errors of estimate of the true time of observation bias.”
Modeled adjustment is, if the paper is read, outside the confidence interval for estimating the TOB on the basis of hourly observations. Same issues of mathematical illiteracy as before. The paper mentions that issues of skew were found in two mentioned months, but no analysis is undertaken with respect to skew or any arbitrary bias introduced from a model out of statistical bound of an estimate of a midpoint of a range. Useless.
Byrd (1985): “The modified method involves adjusting data to a “true” mean obtained by averaging all hourly temperature values for the 24-hour period ending at midnight, rather than adjusting to the midnight standard observational mean obtained by averaging the maximum and minimum values over the same period. The adjustments are applied to mean temperatures from stations with different observation times in the region around Oneonta, resulting in spatial analysis fields which are believed to be more representative than those using the published data.”
Attempts to apply an *actual* mean — for once, thankfully — to the midpoint of the Tmin/Tmax. Mathematical illiteracy. Didn’t bother to read this one for sd, or skew issues.
Blackburn (1983): ” A scheme for adjusting these reports to eliminate the biases and make them conform to the midnight-to-midnight reports of first-order weather stations is described.”
Same mathematical illiteracy as previous. Bonus points for homogenizing weather. GIGO.
Schall & Dale (1977): ““Historical changes in time of once daily maximum and minimum temperature observations at cooperative climatological stations from 1905 to 1975 have introduced a systematic bias in mean temperatures. Unless corrected, this bias may be interpreted incorrectly as climatic “cooling” and may also affect the assessment of agricultural production potential and fossil fuel needs.”
Funny that the next ice age was the big thing back then. Once again, the midpoint of a range is not a mean temperature for any 24 hour period. Mathematical illiteracy.
Baker et al (1975): “Comparisons of the annual and monthly mean temperatures showed deviations can be of such magnitude as to discourage comparison of station temperatures and temperature-derived quantities such as HDD and GDD unless observation times are the same or corrections are applied”
Baker here properly distinguishes between a true ‘mean’ and ‘Tmean’. Kudos to baker. As noted here he concludes the homogenizing station data is a farce unless you fudge. Too bad Baker, “weather is not climate” and homogenizing is nonsense. Though, to be fair, they were still stocking up on coal for the coming ice age in 1975.
All save one of those not mentioned, that had abstracts, were about detecting TOB via various means. Which is fine for what it’s worth. But let’s reiterate Baker (1975) once more: “These data were used to calculate 1) a true daily mean, 2) a mean of the maximum and minimum between successive midnights as observed at first order stations, and 3) a mean of the maximum and minimum observed at all other hours of the day to simulate cooperative station means.”
None of these things is like the others. And the uninformed yahoo’s here knew more about the contents of your link dump then you yourself did. Good job on your efforts.

A. Scott
August 1, 2012 1:19 pm

Ron Broberg says:
August 1, 2012 at 7:01 am
A. Scott: Seems everything is there to reproduce the work for the people who question the results. I did it on my own
You were able to recreate the list of USHCN station ids and their associated Leroy 2010 classification? Well done! Perhaps you can sure the list with the slow half of the class.

Nope. And I never indicated nothing or the sort. Stop putting words in my mouth I did not say.
What I did say was the links collected appear to include all the data- both temp and station info, along with the Leroy(2010) siting standards, so you or anyone else could do the same thing Anthony has done …..go thru the temp and station data, apply the Leroy(2010) standards yourself, and generate the data the Watts report did.
I included a link to the USHCN metadata files and a link to Anthony’s Surface Stations info as well – both which give a reviewer data on the station history and siting data.

Hu McCulloch
August 1, 2012 1:31 pm

The graph of NOAA adjustments above and at http://wattsupwiththat.files.wordpress.com/2012/04/ushcn-adjustments.jpg is labeled degrees Fahrenheit, but then the footnotes refer to degrees Celsius. Use with caution!

Brian D
August 1, 2012 1:36 pm

This is a good place to go to get comma delimited files for the USHCN network. Has raw, TOBS, and full adjust files. You can also generate graphs as well.
http://cdiac.ornl.gov/epubs/ndp/ushcn/ushcn_map_interface.html

A. Scott
August 1, 2012 1:38 pm

beng says:
August 1, 2012 at 8:13 am
****
Entropic man says:
August 1, 2012 at 7:20 am
I know a rural station, not in the US, notorious for underreading because it is in a frost hollow. Winter Tmin values are consistently lower than others in the region, giving distorted winter daily temperatures and the lowest annual average temperature for 200 miles.
****
If the station is otherwise consistent/unchanging, being in a frost hollow has no effect on the trend.

This would seem to be true regarding that single station … as with UHI effect, if the conditions remain the same the station trend should be unaffected. But when it comes to the larger record, a station that consistently reads low will, in my understanding, clearly affect the larger record.
As far as I know we compare temp data to generate trends – we do not compare trend data itself. Add in a low reporting station and your larger scale trend data is now changed.

u.k.(us)
August 1, 2012 1:43 pm

Thanks for all the TOBS info everyone.
My next sleepless night will be dedicated to its contemplation, and hopefully a better understanding of the apparent issues.
It is not like global warming is gonna jump-up of the sudden and create catastrophe, so the time-sink of including the TOBS seems a reasonable course of inquiry.
If it keeps the CAGW theory alive.

A. Scott
August 1, 2012 1:49 pm

claimsguy says:
August 1, 2012 at 11:29 am
A critique of the article from Skeptical Science:
http://www.skepticalscience.com/watts_new_paper_critique.html

A lengthy display of, to me, largely gibberish – intended to refute the Watts paper through obfuscation.
While completely ignoring and failing to address it seems the whole point of the Watts paper – to apply the new WMO endorsed Leroy(2010) siting standards and see what effect it has.
Which ironically the posters in the comments acknowledge is a valuable contribution.

Christoph Dollis
August 1, 2012 2:20 pm

“A lengthy display of, to me, largely gibberish – intended to refute the Watts paper through obfuscation.”

Maybe it was wrong in areas, but it wasn’t “obfuscation” — its positions were clearly stated.

Nick Stokes
August 1, 2012 2:20 pm

wsbriggs says: August 1, 2012 at 6:48 am
“I have an issue with the idea that we will record the min/max within a single 24 hour period every day, and then complain and insist that if the next day’s max occurs within one second after the end of the previous 24 hour period, that it is invalid.”

Who is saying that it is invalid? It’s a measure. The problem is that the result depends on the time of day when you took those readings. It’s changing that time for a station that creates the need for a TOBS adjustment. Otherwise there is a spurious step change.

Susan Sample
August 1, 2012 2:23 pm

I am writing this after looking at your PDF on the figures that go with your discussion paper. Here are some suggestions, on some changes.
Figure 1; in the pie chart you used I found the class 3 and class 4 colours much too close to being the same shade. Lighten the colours up please. The class 5 colour was too strong a shade in my opinion.
Figures 4 through 8; I would widen those graph bars to show an emphasis between the three sets of data. Darken the colours a bit more.
Figures 10-18; I was a little confused about those graphs because there was no vertical lines in between that data points of each class of stations.
By the way I didn’t see figure 19, just to pass the information on to you.
I do appreciate all the time and effort you put into this, and I thank you.
Susan

Anonymous Coward
August 1, 2012 2:24 pm

Maus, you are either unable or unwilling to see the forest from the trees. Those are about a dozen peer reviewed journal articles–by people that spend a whole lot more time studying this issue than you do–that 1) document the existence of TOB in temperature measurement, and 2) set forth a methodology for correcting for TOB in temperature data. It’s really that simple.

Anonymous Coward
August 1, 2012 2:42 pm

Here is an analogy: Say we want to know the exact average height of group of people. The problem is, when we take measurements, half of them are wearing shoes and half of them are barefoot. Shouldn’t we correct for the ones wearing shoes? There really shouldn’t be any controversy about that. You can argue about the methodology one uses to remove a bias from your raw data, but not the need to remove that bias.
That is why the surface temperature data is adjusted.

Annie
August 1, 2012 2:46 pm

markstoval @ 4:02 am
You’ve put very simply what I was thinking.
I am still trying to get my head around all of this.
Amazing work Anthony et al.

TomRude
August 1, 2012 2:52 pm

Tamino weights in:
http://tamino.wordpress.com/2012/08/01/much-ado-about-nothing/
Self labelling title indeed…

A. Scott
August 1, 2012 3:19 pm

From Steve Goddard:
http://stevengoddard.wordpress.com/data-tampering-at-ushcngiss/
“Data Tampering At USHCN/GISS
The measured USHCN daily temperature data shows a decline in US temperatures since the 1930s. But before they release it to the public, they put it thorough a series of adjustments which change it from a cooling trend to a warming trend.”
Great animated graphics – here is one:
http://stevengoddard.files.wordpress.com/2012/07/ushcn26.gif?w=640

A. Scott
August 1, 2012 3:24 pm

Christoph Dollis says:
August 1, 2012 at 2:20 pm
“A lengthy display of, to me, largely gibberish – intended to refute the Watts paper through obfuscation.”
Maybe it was wrong in areas, but it wasn’t “obfuscation” — its positions were clearly stated.

Chris – they completely failed to address the most important point of the Watts paper – the difference in siting quality when applying the Leroy(2010) standards. In my opinion and observation, as is their usual M.O. – they threw a whole lotta garbage against the wall, made a lengthy dissertation strictly intended to refute – all the while ignoring the primary and most important premise. There is only one reason for that IMO – and that is to obfuscate the real story.
The old adage comes to mind:
“… if you can’t dazzle ’em with your brilliance – baffle ’em with your BS …”

August 1, 2012 3:34 pm

A. Scott, you have completely missed the point that you can’t say anything about the Leroy (2010) standards until you first homogenize the data by removing the other biases. This point was very clearly stated in the SkS post several times, so I’m not sure how you missed it.

Jan P Perlwitz
August 1, 2012 3:37 pm

Anthony Watts wrote:

The TOBS issue is one that may or may not make a difference in the final outcome of the Watts et al 2012 draft paper and it’s conclusions,…

But have the conclusions not already been foregone? At least according to the previous thread and press release, where Watts said:
“New study shows half of the global warming in the USA is artificial
PRESS RELEASE – U.S. Temperature trends show a spurious doubling due to NOAA station siting problems and post measurement adjustments.”
So what again were the previous assertion and the press release about?
Watts’s statement about “global warming in the USA” doesn’t make much sense, anyway. What is “global warming in the USA” supposed to mean? The term “global warming” refers to the globally averaged temperature anomaly trend, but not to the temperature anomaly trend in an area that comprises only ca. 1.6% of the area of the whole globe.
[REPLY: Jan, you really are better than this. “Global Warming In The USA” really was sort of a mal-mot. We’re talking about warming trends in the US. And if you think the issue is restricted only to USA temperature records, then you are not doing any service at all to your current employer. If the Watts et al. analysis proves to be valid, then the results will be, as I’ve said else where, “tectonic”. I think you recognize that. -REP]

Susan Sample
August 1, 2012 3:38 pm

After looking at my comments, I went to research on how colour graphs work in reports. I found this site,
http://www.perceptualedge.com/articles/visual_business_intelligence/rules_for_using_color.pdf
I hope it gives you some help with the presentation of the illustrated figures. So ignore my earlier comments and go for a professional look. Stephen Few has more of an understanding than myself in this matter.

August 1, 2012 3:44 pm

A. Scott: Seems everything is there to reproduce the work for the people who question the results. I did it on my own.
Ron Broberg: You were able to recreate the list of USHCN station ids and their associated Leroy 2010 classification? Well done!
A. Scott: Nope. And I never indicated nothing or the sort. Stop putting words in my mouth I did not say.
If you did not reproduce the work in question, you should not have said that you did so on your own. The confusion is directly tied to your own words. The fact is that we can only guess at which stations that Watts’ has classified, what those classifications are, what methods were used to derive that classification, and if potential biases in his classification could be affecting outcomes. As long Anthony choses to withhold that information, the public review will be incomplete.

Entropic man
August 1, 2012 3:48 pm

J. Philip Peterson says:
August 1, 2012 at 9:32 am
beng says:
August 1, 2012
“If the station is otherwise consistent/unchanging, being in a frost hollow has no effect on the trend.””I agree and if the station is atop a 6288 ft mountain it has no effect on the trend:
http://weather-warehouse.com/WeatherHistory/PastWeatherData_MtWashington_MountWashington_NH_August.html
______________________________________________________________
Just for amusement I calculated the ten year average, 1948 to 1957, and compared it with the ten year average for the most recent 10 years. The older mean is 47.1C and the newer 48.86C.
The trend is +0.27C per decade.
____________
Which month did you calculate for? I made several graphs in excel for ‘Mean Temperature’ for the months of Jan., April, June and Oct. and only April showed an increase from 1948 to 2012 of approx. 3 degrees F max. The other 3 months were even or even trended down.
——————————
I used the August figures given in the link. I could not persuade the site to disgorge annual figures and its too near my bedtime for a marathon calculation. If you can find the annual mean temperatures a calculation based on them would be interesting.

A. Scott
August 1, 2012 3:52 pm

From the silly but fun category of data …
amino Alexa rank 329,436 … Reputation: 602 (sites linking in)
SkepticalScience Alexa rank 144,526 … Reputation: 2,489
RealClimate Alexa rank 166,512 … Reputation: 4,546
MediaMatters rank 19,783 … Reputation: 9,833
WUWT Alexa rank 20,001 … Reputation: 5,112

Maus
August 1, 2012 3:56 pm

Anonymous Coward: ” 1) document the existence of TOB in temperature measurement, and 2) set forth a methodology for correcting for TOB in temperature data. It’s really that simple.”
The midpoint of a range is not an arithmetical mean no matter what you may like to do to it. It’s mathematically ignorant and, by your evidentiary references, a demonstration that no one in the last 37 years has been able to suss out the difference. But beyond the simple an anti-arithmetical notions that the midpoint of a range can represent a ‘mean’ temperature is that there are indeed timing issues. By the construction of the equipment, as a historical issue, Tmin and Tmax have no temporal reference outside a 24 hour window. Due this it is likewise improper, if you are interested in trends, to consider the midpoint at all. This is all basic lab tech junk.
But because they wish to and have been using Tmean, the midpoint, as a foolhardy and nonsensical reference to the average temperature per day then they are somewhat stuck. Numerous short-term weather forecasting models and agricultural models (planting times) are based on this lackwit use of data. Weather forecasting couldn’t care less, as a matter of necessity, as to when the measurements and resets are taken.
But as a significant issue to agricultural issues — and this was mentioned in one of the papers in your link dump — TOB issues can cause erroneous signals as to when planting should be undertaken. But this is not a flaw of the measurement nor data. It is a flaw of the use of a midpoint as representing a mean. Feel free to insert any joke here you like about aggies, sharecroppers, and other fly-over hayseeds.
And for that specific use, and given the available instrumentation, there is absolutely no other option than to fudge. Completely make things up. And it is completely meaningless if it is off by any significant degree. For example, and again from your link dump, the paper on TOB adjusts in Canada found differences of 0.5 to 12.5 degrees of both signs when attempting to ‘move’ one of Tmax or Tmin out of one 24 hour period and into one calendar day. Another paper, the one with that mentioned but did not calculate skews, find TOB adjusts binned by time-of-day whose standard deviations ranged from 0.05 (IIRC) to 0.32. This is per day. Not per month, per year, per decade, per sextadecadalwhatchamacycle. But per day.
But if we accept, arbitrarily chosen, 0.5 Celius per decade then the per calendar day warming comes in it 0.00014 (Remarkable precision, eh?) degrees. No leap years or anything fancy. But if we know there is skew and the SD of the *best* deviation from a model that only *estimates* Tmin/Tmax is 357 odd times greater than the daily warming? Then the whole idea is a nonsense.
But more to the point these TOBS adjusts are only necessary for getting a correctly wrong answer to the Tmin/Tmax for a given calendar day. But climate models don’t give a fig about such notions. They care only about trend over long term scales. To the degree that there are good arguments for TOBS, and there are for practical matters, they have no relevancy to climate science.
What is relevant to climate science, and also again you will find in your link dump should you happen to read it, is that we can detect differences or alterations in the time-of-day in which the measurement was taken. As this allows for keeping things nice and tidy for our 0.00014 per day trend line.
But there is no excuse in defending an idiotic metric chose for pragmatic reasons, due instrumentation capabilities, a century ago as any valid notion of current measurement. Let alone arbitrarily fudging it for uses that have no bearing nor relation to climate issues. Feel free to stake your claim in the AGW camp as you like, but have care not to hoist yourself by your own petard.

Brian D
August 1, 2012 3:57 pm

I looked at a station near me, and the TOBS adjustment they did to it. The metadata available through MMS and the monthly obs reports I downloaded for the history of the station (218419) which is a part USHCN and GHCN, and they don’t match the changes made via TOBS.
http://cdiac.ornl.gov/epubs/ndp/ushcn/ushcn_map_interface.html
https://mi3.ncdc.noaa.gov/mi3qry/login.cfm
Available metadata from MMS and the forms show unknown time until the start of 1948. From 1/1948 thru 8/2007 reading time is listed as 1500 LST. 9/2007 until present show 1400 LST. Now the TOBS adjustments change in the record more often than that. The dates of change they have via the adjustments are 1895-1902, 1903-1952, 1953-1963, 1964-1967, 1968-2007, and 2008-present. The switch in 2007 is the only match. They must have come across different stuff, and didn’t care to show it or they are using from somewhere else (there is a tendency to smear stuff around). And all adjustments are negative because of pm readings. Max is hit hard, especially after 1968.

davidmhoffer
August 1, 2012 3:58 pm

Anonymous Coward says:
August 1, 2012 at 2:42 pm
Here is an analogy: Say we want to know the exact average height of group of people. The problem is, when we take measurements, half of them are wearing shoes and half of them are barefoot. Shouldn’t we correct for the ones wearing shoes?
>>>>>>>>>>>>>>>>
Yes, and if we’ve been recording their heights from childhood to adulthood, how do we go back after the fact, twenty or thirty years down the road, so adjust for shoes on or shoes off in each year if that information was never recorded? How do we adjust for some periods of time when the rule was “shoes off” but some people recording the info didn’t get the memo? How do we account for periods of time when the rule was “shoes on” but some people said “I’ve always done it shoes off so I’m not changing”?
The more I read about TOBS the more I agree with Anthony. It is one thing to sit in an office and scrutinize data in the hopes of identifying these things strictly from statistical analysis, quite another to get out in the field, talk to the people who do the recording, and soon discover that all the assumptions you thought made your statistical analysis accurate were bad ones. TOBS is a red herring.
Further, even if it is legit, I think it is a distraction from the core issues, of which there are two:
1. Bad sites were in general adjusted upward. There are prescious few siting issues that can cause a reading to be lower than it should be, they almost all result in readings being higher. The adjustments are bogus from that perspective, and one needs no statistical analysis to reach that conclusion.
2. Pristine sites were adjusted upward even more, the justification being that they were outliers of some sort, in disagreement with the majority of the sites adjacent to them. This too is bogus. Since the sites were pristine, they should have been considered the more accurate depsite there being fewer of them. Again, one needs no statistical analysis to arrive at that conclusion.
The only think one needs a statistical analysis for is to compare the disparity in trends. This is where the misdirection with TOBS begins. TOBS and other adjustments only matter if one is focused on calculating an accurate trend. Given all the variability in the data and how it was collected over time, I consider that a fool’s errand. But to conclude that the methodologies being used artificially increase the trend substantially by applying adjustments in the wrong direction, making any conclusions drawn from completely unreliable, requires no more information than what I have quoted in points 1 and 2 above.

Entropic man
August 1, 2012 4:11 pm

The Skepticalscience critique raised a number of points, many of them already mentioned here. Some relate to the thesis of the paper, some to its technical quality.
What encourages me most is the last sentence.
” With said caveats carefully addressed and the conclusions amended if and where necessary, the paper has the potential to be a useful contribution to the climate science literature.”
It suggests that Mr Watts can better make his point by joining the game, rather than blogging from the stands.
[More to the point, it suggests that you should start your own blog rather than commenting critically at another person’s home on the internet. And Skeptical Science is the sole blog listed here as “Unreliable”. Search the archives to understand why. ~dbs, mod.]

Christoph Dollis
August 1, 2012 4:22 pm

@ Jan

Watts’s statement about “global warming in the USA” doesn’t make much sense

They tried to cram too much, awkwardly, into too short a headline so I’d concede that, but it’s such a minor point and you know it. You’re trying to make a mountain out of a molehill. Ally E. in the original Press Release thread came up with what I thought was a better headline, for what it’s worth.
If you have substantive criticisms to make, by all means air those, but to claim the compressed wording of a blog headline was a central claim of the study seems a stretch.
(Note: I originally placed this comment under the wrong thread due to a Google Chrome “pagination” plugin that does not play nice with this website at all! My apologies to the moderator.)

michaelozanneike Ozanne
August 1, 2012 4:24 pm

“Tamino weights in:
http://tamino.wordpress.com/2012/08/01/much-ado-about-nothing/
Self labelling title indeed…”
On a speed read seemed to be some ad-hominem bollocks followed by a proof by assertion, am I mistaken?

August 1, 2012 4:30 pm

Ron Broberg says:
August 1, 2012 at 3:44 pm
The fact is that we can only guess at which stations that Watts’ has classified, what those classifications are, what methods were used to derive that classification, and if potential biases in his classification could be affecting outcomes. As long Anthony choses to withhold that information, the public review will be incomplete.

That’s right I could not agree more and as I said elsewhere:
I agree that Anthony Watts and crew should be held to the same incredibly tough standards that are required to be met by everyone else in the field of climate science.
So what does that give him before he has to cough up the code the data and all the details — twenty, thirty years? …and a half dozen FOIAs defended to the teeth? Just askin…. 😉
Or we could hang on a few days or weeks and let them deal with the other important issues that have been raised… Relax….

August 1, 2012 4:31 pm

From reading most of the comments on TOBS, I don’t think the numbers in the Watts et al paper will be affected by adjusting for TOBS. I just don’t see it. I think it’s a wash – up or down, plus or minus. {Hey I may be a lay person, but I did have one full year of engineering courses at Penn State U. (JoePa was an assistant coach that year in 1961)}.

A. Scott
August 1, 2012 4:45 pm

Ron Broberg says:
August 1, 2012 at 3:44 pm
A. Scott: Seems everything is there to reproduce the work for the people who question the results. I did it on my own.
Ron Broberg: You were able to recreate the list of USHCN station ids and their associated Leroy 2010 classification? Well done!
A. Scott: Nope. And I never indicated nothing or the sort. Stop putting words in my mouth I did not say.
If you did not reproduce the work in question, you should not have said that you did so on your own. The confusion is directly tied to your own words. The fact is that we can only guess at which stations that Watts’ has classified, what those classifications are, what methods were used to derive that classification, and if potential biases in his classification could be affecting outcomes. As long Anthony choses to withhold that information, the public review will be incomplete.

Bull. When you read my comments in context – instead of cherry picking as you did – it is quite clear I was talking about collecting the data links I presented.
All of the data is available. There is a complete discussion in the paper and in a separate “Methods” Powerpoint on the process, and data sets used are clearly identified in the paper. There is nothing to prevent you or anyone else from doing exactly the same study as done by Anthony and his group.
The Watts paper identifies the stations used – under Sec 2.2 Station Site Classification:
“We make use of the subset of USHCNv2 metadata from stations whose sites have been classified by Watts (2009), gathered by the volunteers of the surfacestations.org project using the USCRN site-selection classification scheme for temperature and humidity measurements (NOAA/NESDIS 2002), and originally developed by Leroy (1999)”
These stations are listed at http://www.surfacestations.org
The same station data was used in Falls(2011) – the data files for that are all avail – along with the code – here:
http://www.surfacestations.org/Fall_etal_2011/si/fall_etal_2011_SI.zip
The Watts paper also notes the original “site rating data” is included in the Falls(2011) data file.
They further identify they used the Fall(2011) stations, and then applied the Leroy(2010) standards to them:
“For the purpose of this study, the original site rating metadata from Fall et al (2011), also used in Muller (2012), was supplemented with further refinements and additional station surveys inclusive from June 15th, 2011 to July 1st, 2012, followed by application of the Leroy (2010) site survey rating system to both old and new surveys (Table 1)”
No need to guess … there is your answer as to the station data used – they used the same stations as Fall (2011) , which was derived from Watts(2009). The data for both are readily available.
The station classifications under Watts are a “result” – having the “results” befoire hand is irrelevant IMO – and introduces a potential bias – to any attempt at reproducing his results.
And no need to guess at “what methods were used to derive that classification” either … the Leroy(2010) standards are readily avail – I posted them and they are included in the papers References section. Leroy(2010) is supposed to be a “standard” – it should be uniformly applied by all.
Last – any “potential biases in his classification [that] could be affecting outcomes” are also irrelevant to an attempt to reproduce his results. If you use his data and follow his well described methods, applying the Leroy(2010) siting standards to the Fall(2011)/Watts(2009) stations you will obtain your own results.
Everything is available to reproduce Watts work. The results of any proper attempt to reproduce will either confirm Watts work or they will not – on both station classifications under Leroy(2010) and on the results he obtained. If they do not, then an analysis can be made of where the differences are.
Providing the answers in advance, in this case at least, offers nothing to the process

TimTheToolMan
August 1, 2012 4:48 pm

So it seems the current TOBs correction might be an upper limit on the correction but the real correction is somewhat less if it turns out that people really are people and not the automaton temperature reading takers that had been previously assumed. That puts it in perspective…

Ripper
August 1, 2012 4:50 pm

Brian H says:
August 1, 2012 at 9:44 am
Places with variable weather can throw min/max thermometer assumptions out the window. If a cold day is followed by an overnight (say, 2 a.m.) warm front moving through, the MIN will still show the cold day’s measurement, even though it might be much warmer by the following sunrise. IMO only continuous plots, with frequent data point recording, can get around this. And then the whole “MIN/MAX issue is almost moot, because now you have real integrated temperature plots and actual averages, not stupid medians, are available
==============================================================
Yes , I have seen plenty of times where in the arid zone where I live, that the temperature before 9am was 3+ degrees warmer than the previous day’s (particularly if it rained) Tmax .
In these cases the maximum for the day before that would have been recorded would have been too warm under the old Max/Min method being read and reset at 9AM.
So I am sceptical that morning observations always result in a “cool” bias and evening observations a “warm” bias. It really depends on the season and the locality.

Christoph Dollis
August 1, 2012 4:56 pm

From reading most of the comments on TOBS, I don’t think the numbers in the Watts et al paper will be affected by adjusting for TOBS. I just don’t see it. I think it’s a wash – up or down, plus or minus.

Apparently Steve McIntyre doesn’t share your certainty.
From what I’ve recently gathered, American stations shifted from afternoon to morning measuring at some point. This is in general and by and large.
The thing is, shifting resetting the minimax thermometers to a cooler part of the day can result in more cold days being double-counted than prior to the shift. That would be a cooling bias.
By not removing the cooling bias with a TOBs adjustment, you present the homogenized measured results as cooler than the real, physical temperatures. And Watts et al. 2012’s primary conclusion is that real, physical temperatures have been cooler than we’ve been told.
So the TOBs adjustments are directly relevant, or so the argument goes.

Christoph Dollis
August 1, 2012 5:06 pm

So I am sceptical that morning observations always result in a “cool” bias and evening observations a “warm” bias. It really depends on the season and the locality.

Sure. There will be local variations. Latitude is important. Etc.
But the difficulty of the math doesn’t mean it should be ignored. If Anthony’s right and the data collection was often so bad (fudged records as to what time the measurements were taken), then maybe it’s appropriate to ignore TOBs for that reason.
In that case, it would be appropriate for him to make a strong argument to that effect in the paper. Peer reviewers and scientists and the lay public reading the paper for themselves could assess the validity of the argument.
But just, “I think it will all average out,” isn’t as good as doing the calculations. If McIntyre shows they do, in fact, nearly average out, then bully.

Konrad.
August 1, 2012 5:12 pm

There are two separate issues here.
1. Can TOB effect the accuracy of recorded temperatures, and is it valid to adjust for it.
2. Is the method of adjustment used in USHCN valid.
For (1) I believe Yes is a reasonable answer, but it may not be necessary for climate observation.
For (2) I believe No is the only reasonable answer. The TOB adjustment represents around half of the warming adjustment to USHCN data. The adjustment is not made on an individual station basis with direct reference to individual station records.
The approach of making adjustments to groups of stations can only be valid for time zone TOB. To make adjustments to station data for min/max thermometer recording times on anything other than an individual station basis using the actual station records of recording times and equipment changes is a totally invalid and unjustified procedure. “But, but we validated it against a cherry picked subset, blah, blah, blah”. It’s wrong and no amount of hand waving will make it right.
It is interesting to note that Anthony has shown that you can not adjust for station site issues from behind a desk without obtaining metadata about individual station site conditions. Ridiculously, those trying to challenge this work are claiming that an adjustment from behind a desk for TOB that does not adjust on an individual station basis using individual station records should have been applied.

August 1, 2012 5:23 pm

interestingly, at today’s EPW Committee, chaired by Sen. Boxer, when Dr. Christie quoted from this latest Study about the USA temperature record, he was accused by Boxer of submitting a “paper” which had never been published in ANY Journal, and nor was it “Peer Reviewed”.
Is it not the case that this blog IS A JOURNAL, and a multi-award winning Journal at that.
It has hundreds of times as many readers as the dead tree press journals.
It does not charge people to view the information beyond synopses.
And as pointed out in the preamble above, there IS A PEER REVIEW PROCESS
going on here in the comments. Not only that but ANYONE is free to make any
comment (within reason and propriety), whether helpful or admonitory, whereas
in the dead tree journals, only the invited few ever get to review such early work.
As in the case of Charles Monnet whose work was “peer reviewed” by a colleague
in the same office, and his own wife, before being published in a “journal”.
According to Sen. Boxer’s rules, Monnet’s paper would be acceptable as
evidence, yet Anthony Watts work is not. Charles Monnet was discredited
when it was discovered that his Polar Bear data seemed a bit haphazard,
and he was interviewed for hours by Special Agent Eric May, from the
US Government who was sent to investigate the Bowhead Whale project
that he was supposed to be researching. On the other hand The FBI have
yet to visit Watts to ask him about the surface stations record.
There is an audio transcript of that Monnet “grilling”, which when heard,
makes it abudantly clear just how haphazard Monnet’s methodology really
is, when compared to the meticulous detail and thoroughness of Watts.
That transcript, read by an actor may be heard on Video Wall #12,
at the Fraudulent Climate website.
at time of posting FC website is down, please try again later !

GlynnMhor
August 1, 2012 5:39 pm

I could see an issue possibly being created if the Time of OBservation were the time at which the temperature was recorded.
But the time at which the daily min/max temperatures are taken off the machine should generate as many upward as downward biases, and small ones at that, since those min and max temperatures occur at their own times, independent of the time they are recorded and the instrument reset.
There should not be any net shift of min/max temperatures, much less the strongly positive one shown in the adjustments made to the record.

Christoph Dollis
August 1, 2012 5:39 pm

Is it not the case that this blog IS A JOURNAL, and a multi-award winning Journal at that. [sic]

No. Not at all.

August 1, 2012 5:46 pm

“Is it not the case that this blog IS A JOURNAL, and a multi-award winning Journal at that.”
The difference is that a climate journal has an agenda, and rarely allows a paper that doesn’t fit its alarmist narrative. OTOH, WUWT allows all points of view, it reaches a much, much larger audience, and it has a much greater impact than either Science or Nature.
And WUWT doesn’t charge $200 a year for a basic subscription.

Entropic man
August 1, 2012 5:53 pm

[More to the point, it suggests that you should start your own blog rather than commenting critically at another person’s home on the internet. And Skeptical Science is the sole blog listed here as “Unreliable”. Search the archives to understand why. ~dbs, mod.]
I am here to debate the science of climate change with sceptics, on a site where they are abundant. I see no point in blogging from a pro-cAGW viewpoint when I know most sceptics are unlikely to read it.
“It suggests that Mr Watts can better make his point by joining the game, rather than blogging from the stands.”
[More to the point, it suggests that you should start your own blog rather than commenting critically at another person’s home on the internet. And Skeptical Science is the sole blog listed here as “Unreliable”. Search the archives to understand why. ~dbs, mod.]
If this site is intended as a forum for debate, then opponents of the site’s stance should be welcome as grit in your oyster. As another of your moderators said, there are people here who are quite capable of refuting any dubious statements I might make.
On the other hand, this site may be intended as a private reserve where sceptics can agree with each other in peaceful isolation, being told only what they want to hear. If so, then I can understand that the presence of outside views would be unwelcome.
As for “Unreliable”, which sites are reliable or unreliable is very much a matter of viewpoint.
Human nature being what it is, any site which promotes a different view from one’s own is automatically suspect.
Regarding Mr Watts’ paper, I agree with Skepticalscience that it has useful things to say, but will
not be properly considered outside the sceptic community until it has passed peer review, which still needs work. I have attempted to provide constructive criticism to that end.

Jan P Perlwitz
August 1, 2012 5:54 pm

Moderator REP wrote:

Jan, you really are better than this. “Global Warming In The USA” really was sort of a mal-mot. We’re talking about warming trends in the US.

So why not clearly stating so, instead of creating a context that gives the impression it was about “global warming”. Now, the meme that Anthony Watts had published a study, which alledgedly contradicts other analyses of global warming trends is already being spread, e.g. here
http://www.foxnews.com/scitech/2012/07/30/weather-station-temp-claims-are-overheated-report-claims/
In the context of his article, Watts is also quoted in this piece of Fox News desinformation:
““I believe global warming is real. No doubt about it. Not a bit of doubt,” Watts told FoxNews.com. “However, I don’t think it’s catastrophic, or as bad as it’s been portrayed.”

And if you think the issue is restricted only to USA temperature records, then you are not doing any service at all to your current employer…

This looks like a variation of the logical fallacy of appeal to authority. Or is this a subtle try to intimidate me by bringing my employer into the argument? BTW: Who do you think is my employer anyway?

If the Watts et al. analysis proves to be valid, then the results will be, as I’ve said else where, “tectonic”. I think you recognize that.

For the assessment of the magnitude of the global warming trend? No, I don’t recognize that. For the sake of the argument, even if there was an artificial trend in the US temperatures, using the adjusted USHCNv2 data set, as large in magnitude as initially claimed by Watts et al. in their manuscript, I rather expect that your expectation you formulate here will be disappointed. The adjustment procedure by Menne et al. (2009) is the issue here, which is applied to the USHCN data. These data are a subset of all data from an area that equals only ca. 1.6% of the area of the whole globe. So the effect on the global temperature anomaly trend would be rather small. Different groups use the unadjusted GHCN data and other data sets for the global analysis, and they apply their own adjustment procedures to those data to account for artificial trends. You would have to assume that the different adjustment procedures used by the different groups to eliminate artificial trends from the GHCN data are all equally flawed, e.g., the one presented in Hansen et al. (2010), doi:10.1029/2010RG000345 *, where the authors use satellite observed night lights to separate between urban and rural stations. However, the global temperature trend is not significantly different, even if the analysis is restricted to pitch black areas on the globe. Not likely that results from this analysis would suffer from the same flaws, regarding the global temperature anomaly trend. Why would it? The paper by Hansen et al. (2010) is not referenced in the manuscript by Watts et al.
(*) http://pubs.giss.nasa.gov/docs/2010/2010_Hansen_etal.pdf

Brian D
August 1, 2012 5:58 pm

Looked at another station in my area (211630). Change in TOBS adjustment with no metadata to back up time change. Metadata show 1700 LST until 12/1982. Then from 12/1982 until present its 1630 LST. TOBS adjust dates 1911/1912 – 1/1997, then 2/1997 to present.
What did occur was an update to the lat/long by GPS in 2/1997. Karl’s algorithm must be real sensitive to even the slightest changes to coordinates. Obs times are pm, so neg adjustments are made for the entire record. That change made them more negative. HMMM, wonder what I will find when I come across an a.m. station. They are hammering these p.m. ones pretty good, especially max temps.

August 1, 2012 6:05 pm

Perlwitz says: [ “…” ]
Before being an apologist for the alarmist crowd, work on your 8th grade spelling skillz.☺

Anonymous Coward
August 1, 2012 6:19 pm

It seems like there a lot of comments to the effect of either: “TOBS should average out and have no impact on the data + or -” or “Too little is known about TOBS to make any adjustments” or “The TOBS adjustments that were made are invalid because [insert reason]”.
If you ask me, nobody commenting on this blog has the right to make those statements unless they study the scientific literature on this issue and really understand why the TOBS adjustments were made and how they were made. Saying you feel that you know more than the PhD climate scientists that have spent countless hours studying the surface temperature record and publishing peer reviewed journal studies on it got TOBS wrong because you took an engineering class in 1961 and you just know better than them – well, that is just not a good way to make your case, in my opinion.

August 1, 2012 6:23 pm

And regarding trends, see the unchanging, non-accelerating, worldwide, long term temperature trend as the planet emerges from the LIA. Note that the same global warming pattern has occurred in the past, exactly the same as now – but at a time when CO2 was much lower. Thus, CO2 makes too little difference to even measure. It doesn’t matter, except as a beneficial plant food.
And Perlwitz is still quoting that discredited climate charlatan and serial lawbreaker, Mr. “Coal Trains of Death” James Hansen?? Here is an example of his shenanigans. [look close] No need to try and explain, we know deceptive ‘adjustments’ when we see them.
USHCN plays the same mendacious games. See here.
Conclusion: the truth is not in them. They lie for money, status, and endless paid vacations to places like Bali and Rio. Despicable, no?

Brian D
August 1, 2012 6:30 pm

Should have put the changes down in last comment.
46d 42′ 00″ to 46d 42′ 16″
92d 31′ 00″ to 92d 31′ 31″
That changed TOBS adjust.
Jan
Max = -2.1 to -2.6
Min = -1.7 to -1.8
Feb
Max = -1.5 to -2.3
Min = -1.5 to -1.5
Mar
Max = -1.1 to -1.8
Min = -1.0 to -1.0
Apr
Max = -1.2 to -2.0
Min = -0.9 to -0.9
May
Max = -2.0 to -2.7
Min = -1.0 to -1.0
That’s enough to get the point across. Is this working the same way in the positive direction? Might find out soon enough.

Maus
August 1, 2012 6:32 pm

Perlwitz: “Now, the meme that Anthony Watts had published a study, which alledgedly contradicts other analyses of global warming trends is already being spread, e.g. here”
From your link — “Global warming believer-turned-skeptic Anthony Watts — a former TV meteorologist — posted a new report online questioning the reliability of weather stations in the U.S. Historical Climatology Network …”
So when you said ‘global warming trends’ you meant ‘US global warming’, right? And unforced error of this magnitude is the sort of thing you expect from NASA. That egghead agency that hasn’t been to the moon in ages and is so successful at their core mission that their now making Muslims get the warm-fuzzies and hitching rides with Russians.
“This looks like a variation of the logical fallacy of appeal to authority.”
The porn test doesn’t work here; it is or it is not. And it is not. Nor is it any fallacy at all but color commentary. Perhaps rude, but hardly an attempt to establish his argument by any means.
“Or is this a subtle try to intimidate me by bringing my employer into the argument?”
Yes, of course it’s a threat.
“BTW: Who do you think is my employer anyway?”
Why are you threatening yourself in public? Is this Blazing Saddles? Ah… On second notice I see that Jan P Perlwitz works for NASA GISS. That’s both too obvious and too insulting given what I’ve already stated. So I’m going to go with Burger King. What do I win?
“For the assessment of the magnitude of the global warming trend?”
Assessment of siting issues. Paper is linked here if you wish to engage in peer review.
“These data are a subset of all data from an area that equals only ca. 1.6% of the area of the whole globe.”
Fair enough. What with NASA’s record we should expect that only Americans could cock up basic measurements so badly. But the Russians? They’re golden.
“Different groups use the unadjusted GHCN data and other data sets for the global analysis, and they apply their own adjustment procedures to those data to account for artificial trends.”
Good to know that someone working at Burger King can stumble on the obviousness that including TOBS in the paper is a meaningless exercise. You may have a future in a scientific profession, you know.
“The paper by Hansen et al. (2010) is not referenced in the manuscript by Watts et al.”
Neither were any of my excellent haikus about Leroy (2010). A pity you aren’t able to read any of them. Or Leroy.

August 1, 2012 6:34 pm

Fall 2011: 1221 stations; 1032 rated CRN (SI/final_list; SI/ratings.csv)
Surface Stations: 1068 surveyed, 1007 rated (http://surfacestations.org; progress as of 07/30/2012)
Watts’ 2012 Draft Jul 29: 779 classified per Leroy (2010)

For the purpose of this study, the original site rating metadata from Fall et al (2011), also used in Muller (2012), was supplemented with further refinements and additional station surveys inclusive from June 15th, 2011 to July 1st, 2012, followed by application of the Leroy (2010) site survey rating system to both old and new surveys (Table 1) including both a distance and an area rating component. Any known changes in siting characteristics after that period are ignored. A total of 1065 USHCNv2 stations were surveyed, comprising 87.4% of the 1218 station USHCNv2 network. Of those 1065 stations surveyed, 779 were classified per the Leroy (2010) site survey rating system (Figure 1). As a rule, LeRoy (2010) is less “strict” than Leroy (1999). There is a greater number of Class 1, 2, and 3 stations, and fewer Class 4 stations. There are, however, a greater number of Class 5 stations, as well.

The 779 USHCNv2 stations in Watts 2012 that have been classified using Leroy 2010 is a subset of the current surfacestations.org rated with CRN which in turn is a superset of the stations used in Fall 2011. Oddly though, surfacestations.org currently reports fewer stations classified than Fall 2011 reports.
No one in the ‘public’ knows which 779 USHCN stations have been classified with Leroy 2010. No one knows if that classification has been rigorous. No one can double check Watts’ classification effort since they do not have his Leroy 2010 classification list with which to compare their own results.
Is verification of the Watts’ Leroy 2010 classification effort an important component of “public review” for this paper? In answer I quote Watts’ own draft:

This simple change in the rating system accounts for the majority of 202 differences in the data and conclusions between this study and Menne et al.,(2010), Fall 203 et al.,(2011), and Muller et al.,(2012).

Anthony Watts has chosen to withhold Table 1 from his draft, the very information which accounts for the majority of the difference between his “conclusions” and those of previously published works. And, regarding the list of 779 stations and their Leroy 2010 classification, he has chosen obscurity over transparency; privacy over public access.

Jan P Perlwitz
August 1, 2012 6:35 pm

@Smokey:
Well, thank you for your strong argument on the topic. And not everyone can speak and write all the languages he/she has learned in life in addition to his/her native language, w/o making any mistakes like you can.

A. Scott
August 1, 2012 6:39 pm

Brian D says:
August 1, 2012 at 3:57 pm
I looked at a station near me, and the TOBS adjustment they did to it. The metadata available through MMS and the monthly obs reports I downloaded for the history of the station (218419) which is a part USHCN and GHCN, and they don’t match the changes made via TOBS.
http://cdiac.ornl.gov/epubs/ndp/ushcn/ushcn_map_interface.html
https://mi3.ncdc.noaa.gov/mi3qry/login.cfm
Available metadata from MMS and the forms show unknown time until the start of 1948. From 1/1948 thru 8/2007 reading time is listed as 1500 LST. 9/2007 until present show 1400 LST. Now the TOBS adjustments change in the record more often than that. The dates of change they have via the adjustments are 1895-1902, 1903-1952, 1953-1963, 1964-1967, 1968-2007, and 2008-present. The switch in 2007 is the only match. They must have come across different stuff, and didn’t care to show it or they are using from somewhere else (there is a tendency to smear stuff around). And all adjustments are negative because of pm readings. Max is hit hard, especially after 1968.

This would seem to be a very good station to look at for a number of reasons… not the least of its data goes back to 1948, which its rural, its in the SurfaceStations.org database, and it has two nearby nearby stations (incl. an airport AWOS).
… here is the direct data link:
https://mi3.ncdc.noaa.gov/mi3qry/identityGrid.cfm?setCookie=1&fid=11336
… and the other two stations:
https://mi3.ncdc.noaa.gov/mi3qry/identityGrid.cfm?setCookie=1&fid=31259
https://mi3.ncdc.noaa.gov/mi3qry/identityGrid.cfm?setCookie=1&fid=33396
http://www.airnav.com/airport/KTWM/
https://maps.google.com/maps?ll=47.049139,-91.745222&spn=0.0170,0.0170
This would seem a perfect grouping to compare siting and TOB issues, along with discrepancy between airport and not (although this one is very rural – I’ve flown in and out of it).
One site is in the City of Two Harbors – nother at the airport and the 3rd in the middle of the woods 7 miles NW of Two Harbors (I believe that may be a USFS fire tower location)
Mapping respources include http://www.terraserver.com and the new USGS National Map product is pretty cool as well:
http://viewer.nationalmap.gov/viewer/

CoRev
August 1, 2012 6:46 pm

Unless he has changed jobs, Jan Perlwitz is Director Human Resources NASA, GISS and very knowledgeable, although biased.

August 1, 2012 6:54 pm

“The paper by Hansen et al. (2010) is not referenced in the manuscript by Watts et al.”
Nor did he quote Darwin or Freud… each appropriate in their place.

Jan P Perlwitz
August 1, 2012 6:57 pm

Christoph Dollis, August 1, 2012 at 4:22 pm:

They tried to cram too much, awkwardly, into too short a headline so I’d concede that, but it’s such a minor point and you know it.

Only, that “minor point” is already used, e.g. by the Fox News propaganda outlet, to spread desinformation.
But it wasn’t my main point either. My main point was the obvious contradiction between the conclusions touted in the press release and what is said by Anthony Watts now above. Pielke Sr. has also already been backpedaling w/ respect to “game changer”.

You’re trying to make a mountain out of a molehill. Ally E. in the original Press Release thread came up with what I thought was a better headline, for what it’s worth.

There shouldn’t have been any claims about conclusions or any press release in the first hand, particularly based only on a half-backed analysis.

Jan P Perlwitz
August 1, 2012 7:04 pm

WillR, you wrote

“The paper by Hansen et al. (2010) is not referenced in the manuscript by Watts et al.”
Nor did he quote Darwin or Freud… each appropriate in their place.

Did Darwin or Freud publish a recent study about temperature analysis in the context of climate change, apply methodology to separate rural and urban stations to remove artificial trends, and discuss the USHCN temperature data set and did robustness test with it? Because Hansen et al. (2010) did.

August 1, 2012 7:06 pm

Perlwitz said:
“There shouldn’t have been any claims about conclusions or any press release in the first hand, particularly based only on a half-backed [sic] analysis.”
Tell that to to the fake skeptic Richard Muller.
And the Anthony Watts et al. paper is unquestioningly a game changer. We see it happening already. There are ten times more reports in the mainstream media than ever before. People are beginning to see what many of them already suspected: that climate alarmism is a grant-driven hoax.

Phil
August 1, 2012 7:10 pm
James from Arding
August 1, 2012 7:16 pm

So can anyone help me with how the TOB issue has been handled with the CET dataset?
Does this make CET and other long term carefully measured data sets more relevant for trying to gauge any estimate of temperature trends for making statements about global climate trends?

A. Scott
August 1, 2012 7:19 pm

Ron Broberg says:
August 1, 2012 at 6:34 pm
Fall 2011: 1221 stations; 1032 rated CRN (SI/final_list; SI/ratings.csv)
Surface Stations: 1068 surveyed, 1007 rated (http://surfacestations.org; progress as of 07/30/2012)
Watts’ 2012 Draft Jul 29: 779 classified per Leroy (2010)

For the purpose of this study, the original site rating metadata from Fall et al (2011), also used in Muller (2012), was supplemented with further refinements and additional station surveys inclusive from June 15th, 2011 to July 1st, 2012, followed by application of the Leroy (2010) site survey rating system to both old and new surveys (Table 1) including both a distance and an area rating component. Any known changes in siting characteristics after that period are ignored. A total of 1065 USHCNv2 stations were surveyed, comprising 87.4% of the 1218 station USHCNv2 network. Of those 1065 stations surveyed, 779 were classified per the Leroy (2010) site survey rating system (Figure 1).
The 779 USHCNv2 stations in Watts 2012 that have been classified using Leroy 2010 is a subset of the current surfacestations.org rated with CRN which in turn is a superset of the stations used in Fall 2011. Oddly though, surfacestations.org currently reports fewer stations classified than Fall 2011 reports.

No one in the ‘public’ knows which 779 USHCN stations have been classified with Leroy 2010. No one knows if that classification has been rigorous. No one can double check Watts’ classification effort since they do not have his Leroy 2010 classification list with which to compare their own results.
Anthony Watts has chosen to withhold Table 1 from his draft, the very information which accounts for the majority of the difference between his “conclusions” and those of previously published works. And, regarding the list of 779 stations and their Leroy 2010 classification, he has chosen obscurity over transparency; privacy over public access.

So download the Fall(2011) stations and the Leroy(1999) site ratings, and run each station thru Leroy(2010) standards – determine your own classifications pursuant to Leroy(2010) for each. Then run follow the process and methids noted in the Watts paper and see what your results are, and compare to Watts.
To me this seems the proper way … it doesn’t matter which 779 stations Watts chose – those stations were a “data result” of his process – eliminating that step, the review of all the stations noted – is eliminating a step in the process.

August 1, 2012 7:20 pm

Phil,
Please, enough with government propaganda from 2010. Who is going to watch an hour long screed that begins, “global warming is unequivocal.”
If you’re new here, get up to speed: we know about global warming. The planet has been warming naturally since the Little Ice Age, along the same trend line, whether CO2 is high or low [and by “high” I mean 0.00039 of the atmosphere].
Therefore, the natural warming trend is ipso facto unaffected by the rise in CO2. Any effect from CO2 is negligible. Why? Because scientific evidence and empirical observations prove it. QED

Maus
August 1, 2012 7:21 pm

Jan P Perlwitz: “Only, that “minor point” is already used, e.g. by the Fox News propaganda outlet, to spread desinformation.”
Not disinformation, but to control debate through — typically — emotional appeals. Such as labeling suspect classes or individuals in pejorative terms. Common examples are Blood Libels, War Time Posters, and claiming that a news agency is a propagnda outlet.
But if I accept your emotionally laden and unbacked claim that Fox News is a propaganda outlet that spreads disinformation then we they stated, as the first sentence: “Two new climate studies agree: the planet is definitely getting warmer.”
Then you mean that this is total hogwash. That this:
“Skeptic-turned-believer Richard Muller …”
Is nonsense. And that the following is a complete lie:
“Muller’s study attempts to correct for the quality of the data, in a transparent, repeatable fashion scientists should appreciate.”
Or are you simply engaged in propaganda?

Jan P Perlwitz
August 1, 2012 7:23 pm

Maus wrote:

Perlwitz: “Now, the meme that Anthony Watts had published a study, which allegedly contradicts other analyses of global warming trends is already being spread, e.g. here”
From your link — “Global warming believer-turned-skeptic Anthony Watts — a former TV meteorologist — posted a new report online questioning the reliability of weather stations in the U.S. Historical Climatology Network …”
So when you said ‘global warming trends’ you meant ‘US global warming’, right? And unforced error of this magnitude is the sort of thing you expect from NASA.

What error are you talking about? “US global warming” is a nonsense word. No, here I mean “global warming trends”, because the Fox News desinformation outlet explicitly claimed, referencing Watts:
Watts cherry-picked the well-sited stations and reanalyzed their data; his results show the planet warming at just 0.155 degrees Celsius per decade, rather than the 0.309 C per decade cited by the government.
(http://www.foxnews.com/scitech/2012/07/30/weather-station-temp-claims-are-overheated-report-claims/#ixzz22LoMRylg)

Why are you threatening yourself in public? Is this Blazing Saddles? Ah… On second notice I see that Jan P Perlwitz works for NASA GISS. That’s both too obvious and too insulting given what I’ve already stated. So I’m going to go with Burger King. What do I win?

Fail. You do not win anything. I leave it to the ones who are capable to figure it out.
The rest of your comment is similarly without any substance with respect to my arguments. Just the usual “it’s all cooked” assertion, which false skeptics like to make. How boring.

Jan P Perlwitz
August 1, 2012 7:29 pm

Smokey says at August 1, 2012 at 6:23 pm:

And regarding trends, see the unchanging, non-accelerating, worldwide, long term temperature trend as the planet emerges from the LIA. Note that the same global warming pattern has occurred in the past, exactly the same as now – but at a time when CO2 was much lower. Thus, CO2 makes too little difference to even measure. It doesn’t matter, except as a beneficial plant food.

Apparantly, this is how Smokey thinks science is done, and how scientific theories are being successfully tested or refuted. One shows a few colored pictures together with some stated bold assertions, and that’s it.

Maus
August 1, 2012 7:30 pm

Jan P Perlwitz: “Because Hansen et al. (2010) did.”
Amazing coincidence! That’s the same year that Leroy published a paper on siting classifications, the focus of the current paper. Perhaps you’ve read Leroy?
Or even the validation of Hansen (1988) and the GISS II model?
http://wattsupwiththat.com/2012/06/15/james-hansens-climate-forecast-of-1988-a-whopping-150-wrong/

DavidA
August 1, 2012 7:31 pm

Anonymous Coward “If you ask me, nobody commenting on this blog has the right to make those statements unless they study the scientific literature…”
This isn’t North Korea, we’re allowed to speculate on a blog.

Jan P Perlwitz
August 1, 2012 7:34 pm

Smokey says at August 1, 2012 at 7:06 pm

Perlwitz said:
“There shouldn’t have been any claims about conclusions or any press release in the first hand, particularly based only on a half-backed [sic] analysis.”

Yes, it actually was supposed to say “half-baked”.

Tell that to to the fake skeptic Richard Muller.

Tu quoque argument. I don’t even have anything to do with Muller.

August 1, 2012 7:36 pm

Perlwitz says:
“here I mean ‘global warming trends’, because the Fox News desinformation [sic] outlet…”
As opposed to the disinformation/leftist propaganda outlets of MSNBC, RealClimate, Closed Mind, ABCNews, Pseodo-Skeptical Pseudo-Science, C–>BS, CNN, etc., etc.?
Perlwitz fails miserably when he starts writing about trends. The longest term trends show unequivocally that the planet is warming naturally from the LIA, and that the warming has not accelerated despite a large increase in CO2.
Therefore, even the most casual observer must admit that CO2 is not the cause of global temperature rise. The CO2=AGW conjecture is falsified by Planet Earth herself, so Perlwitz can now stop pocketing the taxpayers’ grant payola that motivates his anti-science nonsense.

DavidA
August 1, 2012 7:37 pm

Anonymous Coward “If you ask me, nobody commenting on this blog has the right to make those statements unless they study the scientific literature…”
This isn’t North Korea, we have the right to speculate on a blog.

Maus
August 1, 2012 7:38 pm

Jan P Perlwitz: “No, here I mean “global warming trends”, because the Fox News desinformation outlet explicitly claimed, referencing Watts:”
Wait wait. Did you mean that the disinformation was the claim that Watts ‘cherry-picked’ or did you mean that the disinformation was your cherry-picking a single misuse of a term by a journalist in an article that otherwise is repeatedly and extraoidinarily careful mention of the US network? I’m unclear as to which propagandist I’m supposed to consider a propagandist here.
“Just the usual “it’s all cooked” assertion, which false skeptics like to make. How boring.”
Indeed, and two naked assertions on your part. I was hoping that some mockery might get your thinking cap on. But hey, we’ve already covered how useless NASA these days. But I am curious as to how you think I’m a ‘false skeptic’. Wouldn’t that make me an agent provocateur and your comrade hitching rides with the Russians?
I’m going to need a citation for that one, Fries.

Jan P Perlwitz