30 year trends of temperature are shown to be lower, using well-sited high quality NOAA weather stations that do not require adjustments to the data.
This was in AGU’s press release news feed today. At about the time this story publishes, I am presenting it at the AGU 2015 Fall meeting in San Francisco. Here are the details.
NEW STUDY OF NOAA’S U.S. CLIMATE NETWORK SHOWS A LOWER 30-YEAR TEMPERATURE TREND WHEN HIGH QUALITY TEMPERATURE STATIONS UNPERTURBED BY URBANIZATION ARE CONSIDERED
Figure 4 – Comparisons of 30 year trend for compliant Class 1,2 USHCN stations to non-compliant, Class 3,4,5 USHCN stations to NOAA final adjusted V2.5 USHCN data in the Continental United States
EMBARGOED UNTIL 13:30 PST (16:30 EST) December 17th, 2015
SAN FRANCISCO, CA – A new study about the surface temperature record presented at the 2015 Fall Meeting of the American Geophysical Union suggests that the 30-year trend of temperatures for the Continental United States (CONUS) since 1979 are about two thirds as strong as officially NOAA temperature trends.
Using NOAA’s U.S. Historical Climatology Network, which comprises 1218 weather stations in the CONUS, the researchers were able to identify a 410 station subset of “unperturbed” stations that have not been moved, had equipment changes, or changes in time of observations, and thus require no “adjustments” to their temperature record to account for these problems. The study focuses on finding trend differences between well sited and poorly sited weather stations, based on a WMO approved metric Leroy (2010)1 for classification and assessment of the quality of the measurements based on proximity to artificial heat sources and heat sinks which affect temperature measurement. An example is shown in Figure 2 below, showing the NOAA USHCN temperature sensor for Ardmore, OK.
Following up on a paper published by the authors in 2010, Analysis of the impacts of station exposure on the U.S. Historical Climatology Network temperatures and temperature trends2 which concluded:
Temperature trend estimates vary according to site classification, with poor siting leading to an overestimate of minimum temperature trends and an underestimate of maximum temperature trends, resulting in particular in a substantial difference in estimates of the diurnal temperature range trends
…this new study is presented at AGU session A43G-0396 on Thursday Dec. 17th at 13:40PST and is titled Comparison of Temperature Trends Using an Unperturbed Subset of The U.S. Historical Climatology Network
A 410-station subset of U.S. Historical Climatology Network (version 2.5) stations is identified that experienced no changes in time of observation or station moves during the 1979-2008 period. These stations are classified based on proximity to artificial surfaces, buildings, and other such objects with unnatural thermal mass using guidelines established by Leroy (2010)1 . The United States temperature trends estimated from the relatively few stations in the classes with minimal artificial impact are found to be collectively about 2/3 as large as US trends estimated in the classes with greater expected artificial impact. The trend differences are largest for minimum temperatures and are statistically significant even at the regional scale and across different types of instrumentation and degrees of urbanization. The homogeneity adjustments applied by the National Centers for Environmental Information (formerly the National Climatic Data Center) greatly reduce those differences but produce trends that are more consistent with the stations with greater expected artificial impact. Trend differences are not found during the 1999- 2008 sub-period of relatively stable temperatures, suggesting that the observed differences are caused by a physical mechanism that is directly or indirectly caused by changing temperatures.
Key findings:
1. Comprehensive and detailed evaluation of station metadata, on-site station photography, satellite and aerial imaging, street level Google Earth imagery, and curator interviews have yielded a well-distributed 410 station subset of the 1218 station USHCN network that is unperturbed by Time of Observation changes, station moves, or rating changes, and a complete or mostly complete 30-year dataset. It must be emphasized that the perturbed stations dropped from the USHCN set show significantly lower trends than those retained in the sample, both for well and poorly sited station sets.
2. Bias at the microsite level (the immediate environment of the sensor) in the unperturbed subset of USHCN stations has a significant effect on the mean temperature (Tmean) trend. Well sited stations show significantly less warming from 1979 – 2008. These differences are significant in Tmean, and most pronounced in the minimum temperature data (Tmin). (Figure 3 and Table 1)
3. Equipment bias (CRS v. MMTS stations) in the unperturbed subset of USHCN stations has a significant effect on the mean temperature (Tmean) trend when CRS stations are compared with MMTS stations. MMTS stations show significantly less warming than CRS stations from 1979 – 2008. (Table 1) These differences are significant in Tmean (even after upward adjustment for MMTS conversion) and most pronounced in the maximum temperature data (Tmax).
4. The 30-year Tmean temperature trend of unperturbed, well sited stations is significantly lower than the Tmean temperature trend of NOAA/NCDC official adjusted homogenized surface temperature record for all 1218 USHCN stations.
5. We believe the NOAA/NCDC homogenization adjustment causes well sited stations to be adjusted upwards to match the trends of poorly sited stations.
6. The data suggests that the divergence between well and poorly sited stations is gradual, not a result of spurious step change due to poor metadata.
The study is authored by Anthony Watts and Evan Jones of surfacestations.org , John Nielsen-Gammon of Texas A&M , John R. Christy of the University of Alabama, Huntsville and represents years of work in studying the quality of the temperature measurement system of the United States.
Lead author Anthony Watts said of the study: “The majority of weather stations used by NOAA to detect climate change temperature signal have been compromised by encroachment of artificial surfaces like concrete, asphalt, and heat sources like air conditioner exhausts. This study demonstrates conclusively that this issue affects temperature trend and that NOAA’s methods are not correcting for this problem, resulting in an inflated temperature trend. It suggests that the trend for U.S. temperature will need to be corrected.” He added: “We also see evidence of this same sort of siting problem around the world at many other official weather stations, suggesting that the same upward bias on trend also manifests itself in the global temperature record”.
The full AGU presentation can be downloaded here: https://goo.gl/7NcvT2
[1] Leroy, M. (2010): Siting Classification for Surface Observing Stations on Land, Climate, and Upper-air Observations JMA/WMO Workshop on Quality Management in Surface, Tokyo, Japan, 27-30 July 2010
[2] Fall et al. (2010) Analysis of the impacts of station exposure on the U.S. Historical Climatology Network temperatures and temperature trends https://pielkeclimatesci.files.wordpress.com/2011/07/r-367.pdf
Abstract ID and Title: 76932: Comparison of Temperature Trends Using an Unperturbed Subset of The U.S. Historical Climatology Network
Final Paper Number: A43G-0396
Presentation Type: Poster
Session Date and Time: Thursday, 17 December 2015; 13:40 – 18:00 PST
Session Number and Title: A43G: Tropospheric Chemistry-Climate-Biosphere Interactions III Posters
Location: Moscone South; Poster Hall
Full presentation here: https://goo.gl/7NcvT2
Some side notes.
This work is a continuation of the surface stations project started in 2007, our first publication, Fall et al. in 2010, and our early draft paper in 2012. Putting out that draft paper in 2012 provided us with valuable feedback from critics, and we’ve incorporated that into the effort. Even input from openly hostile professional people, such as Victor Venema, have been highly useful, and I thank him for it.
Many of the valid criticisms of our 2012 draft paper centered around the Time of Observation (TOBs) adjustments that have to be applied to the hodge-podge of stations with issues in the USHCN. Our viewpoint is that trying to retain stations with dodgy records and adjusting the data is a pointless exercise. We chose simply to locate all the stations that DON”T need any adjustments and use those, therefor sidestepping that highly argumentative problem completely. Fortunately, there was enough in nthe USHCN, 410 out of 1218.
It should be noted that the Class1/2 station subset (the best stations we have located in the CONUS) can be considered an analog to the Climate Reference Network in that these stations are reasonably well distributed in the CONUS, and like the CRN, require no adjustments to their records. The CRN consists of 114 commissioned stations in the contiguous United States, our numbers of stations are similar in size and distribution. This should be noted about the CRN:
One of the principal conclusions of the 1997 Conference on the World Climate Research Programme was that the global capacity to observe the Earth’s climate system is inadequate and deteriorating worldwide and “without action to reverse this decline and develop the GCOS [Global Climate Observing System], the ability to characterize climate change and variations over the next 25 years will be even less than during the past quarter century” (National Research Council [NRC] 1999). In spite of the United States being a leader in climate research, long term U.S. climate stations have faced challenges with instrument and site changes that impact the continuity of observations over time. Even small biases can alter the interpretation of decadal climate variability and change, so a substantial effort is required to identify non-climate discontinuities and correct the station records (a process calledhomogenization). Source: https://www.ncdc.noaa.gov/crn/why.html
The CRN has a decade of data, and it shows a pause in the CONUS. Our subset of adjustment free unperturbed stations spans over 30 years, We think it is well worth looking at that data and ignoring the data that requires loads of statistical spackle to patch it up before it is deemed usable. After all, that’s what they say is the reason the CRN was created.
We do allow for one and only one adjustment in the data, and this is only because it is based on physical observations and it is a truly needed adjustment. We use the MMTS adjustment noted in Menne et al. 2009 and 2010 for the MMTS exposure housing versus the old wooden box Cotton Region Shelter (CRS) which has a warm bias mainly due to [paint] and maintenance issues. The MMTS gill shield is a superior exposure system that prevents bias from daytime short-wave and nighttime long-wave thermal radiation. The CRS requires yearly painting, and that often gets neglected, resulting in exposure systems that look like this:
See below for a comparison of the two:
Some might wonder why we have a 1979-2008 comparison when this is 2015. The reason is so that this speaks to Menne et al. 2009 and 2010, papers launched by NOAA/NCDC to defend their adjustment methods for the USCHN from criticisms I had launched about the quality of the surface temperature record, such as this book in 2009: Is the U.S. Surface Temperature Record Reliable? This sent NOAA/NCDC into a tizzy, and they responded with a hasty and ghost written flyer they circulated. In our paper, we extend the comparisons to the current USHCN dataset as well as the 1979-2008 comparison.
We are submitting this to publication in a well respected journal. No, I won’t say which one because we don’t need any attempts at journal gate-keeping like we saw in the Climategate emails. i.e “I can’t see either of these papers being in the next IPCC report. Kevin and I will keep them out somehow — even if we have to redefine what the peer-review literature is!” and “I will be emailing the journal to tell them I’m having nothing more to do with it until they rid themselves of this troublesome editor.”.
When the journal article publishes, we’ll make all of the data, code, and methods available so that the study is entirely replicable. We feel this is very important, even if it allows unscrupulous types to launch “creative” attacks via journal publications, blog posts, and comments. When the data and paper is available, we’ll welcome real and well-founded criticism.
It should be noted that many of the USHCN stations we excluded that had station moves, equipment changes, TOBs changes, etc that were not suitable had lower trends that would have bolstered our conclusions.
The “gallery” server from that 2007 surfacestations project that shows individual weather stations and siting notes is currently offline, mainly due to it being attacked regularly and that affects my office network. I’m looking to move it to cloud hosting to solve that problem. I may ask for some help from readers with that.
We think this study will hold up well. We have been very careful, very slow and meticulous. I admit that the draft paper published in July 2012 was rushed, mainly because I believed that Dr. Richard Muller of BEST was going before congress again the next week using data I provided which he agreed to use only for publications, as a political tool. Fortunately, he didn’t appear on that panel. But, the feedback we got from that effort was invaluable. We hope this pre-release today will also provide valuable criticism.
People might wonder if this project was funded by any government, entity, organization, or individual; it was not. This was all done on free time without any pay by all involved. That is another reason we took our time, there was no “must produce by” funding requirement.
Dr. John Nielsen-Gammon, the state climatologist of Texas, has done all the statistical significance analysis and his opinion is reflected in this statement from the introduction
Dr. Nielsen-Gammon has been our worst critic from the get-go, he’s independently reproduced the station ratings with the help of his students, and created his own series of tests on the data and methods. It is worth noting that this is his statement:
The trend differences are largest for minimum temperatures and are statistically significant even at the regional scale and across different types of instrumentation and degrees of urbanization.
The p-values from Dr. Nielsen-Gammon’s statistical significance analysis are well below 0.05 (the 95% confidence level), and many comparisons are below 0.01 (the 99% confidence level). He’s on-board with the findings after satisfying himself that we indeed have found a ground truth. If anyone doubts his input to this study, you should view his publication record.
COMMENT POLICY:
At the time this post goes live, I’ll be presenting at AGU until 18:00PST , so I won’t be able to respond to queries until after then. Evan Jones “may” be able to after about 330PM PST.
This is a technical thread, so those who simply want to scream vitriol about deniers, Koch Brothers, and Exxon aren’t welcome here. Same for people that just want to hurl accusations without backing them up (especially those using fake names/emails, we have a few). Moderators should use pro-active discretion to weed out such detritus. Genuine comments and/or questions are welcome.
Thanks to everyone who helped make this study and presentation possible.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
I nominate Anthony as Science Adviser for President Cruz in 2017 !!!
“the old wooden box Cotton Region Shelter (CRS) which has a warm bias mainly due to [paint] and maintenance issues. The MMTS gill shield is a superior exposure system that prevents bias from daytime short-wave and nighttime long-wave thermal radiation.”
Published calibration experiments show that even well-maintained CRS shelters suffer a warm bias that varies daily and regionally with insolation and wind speed.
Likewise, the MMTS shelters, though better than the CRS, show systematic biases both day and night — night bias being significantly less.
The net finding is that even Leroy class 1 sites will show systematic biases that will put a permanent and significant uncertainty (~±0.4 C for CRS; ~±0.3 C for MMTS) into any temperature measurement, and that cannot be assumed to decrement away in any large measurement average.
Them CRS units carry their own heat sinks around on their shoulders. Plays havoc with Tmax. Even butts up Tmin quite a bit. Look at the data. Sticks out like a fish in a tree.
Sure, they have that old-world aesthetic charm. But that’s all they’re good for.
I’ll add that it’s not the MMTS units that are going wrong. It’s the CRS boxes that have always been going wrong. For purposes of this paper, we add the jumps for conversion (not the pairwise), but what the job really requires is not jacking the MMTS trend up, but squeezing the CRS trends down. Even if one keeps the offset adjustment, and I’m not yet convinced that offset is even correct.
The MMTS shields are subject to systematic error, too. That’s been shown in several very careful calibration experiments. One of them was discussed in terms of the global average temperature, here (869.8 KB pdf)
Only the new aspirated CRN PRTs are relatively free of systematic error.
Buster
I’ll temporarily pretend to be whatever you need (almost) if you helps with your cognitive dissonance.
How about a priest (i think there’s only one kind) and missionary who offers you absolution if you give me 18% of your income and permission to establish climate saving alternative energy schemes on your property.
I’ll even throw in lessons on safe spaces and how to crush the microaggressor … for the kids .. always for the kids.
I see that you are “all in” my son.
My work is done as I have many other souls to save (absorb into the movement).
Please be sure to tell your story and follow the Pope at http://www.CO2absolution.com
Yes my son.
Fear drives most of the members of the Order of Climate Alarmists.
We are comfortable with our cognitive dissonance as a survival mechanism.
It is why we see you as such a ripe acolyte.
Yet we must endeavor not to respond with scorn. No percentage in it.
Isn’t the main point that we are looking at less than 1 degree C increase since 1880?
Many of the temperature readings have been adjusted and there is uncertainty as to their accuracy. Measurement of temperature in terms of variation from an average rather than plotting the baseline temperature creates the illusion of significant increases and this has happened because of the newness of climate science and the lack of defined procedures.
Does a world average temperature have any real meaning at all?
This possible small increase in temperature has not resulted in any increase in natural disasters so why are we about to spend so much money on a misdirected effort to stop further warming?
+ 10
Indeed, a daubert fan might say …
“the increase in the measured attribute is within the known rate of natural variability, therefore the expert has not presented information to claim a causal relationship”
Then, a more highly paid and connected attorney pleading his case before an equally highly paid and predispositioned judge would say …
“you honor, it is not the place of this court to put the potentially affected populace at risk to such an important hypothesis. we encourage the court to err on the side of caution in making its decision on the validity of this expertise”
Didn’t get an answer to this, maybe because I posted at 1AM:
Evan:
Did you take the arithmetic mean for the summary statistics, or did you take a geographically-normalized mean such as NOAA, BEST etc do?
Because that could be a large source of the difference as well. It would also be apples-oranges comparison. Did you take the same kind of mean for the NOAA stations as well?
thanks. And thanks for replying to a pile of posts here, this has been one of the more fascinating threads in a while
Peter
‘Salright.
We took it direct from individual stations and anomalized them to their own data.
Question for: Anthony Watts or Evan Jones or John Christy or John Nielsen-Gammon:
In your digging, research, work …
Was there any subset of 1,218 Weather Stations that spanned more than 150-years, which might meet less restrictive criteria than set forth in study?
(Looking at what accuracy, at best, “AGW Climatologists”, could have used to calibrate their Proxies with.)
Is so, could you make those stations known when you make data public?
(Thanks, either way.)
Was there any subset of 1,218 Weather Stations that spanned more than 150-years, which might meet less restrictive criteria than set forth in study?
Not a prayer. Unless you go in for inferring metadata. Even then, not likely.
Thanks for reply. An taking time to reply to so many others.
What, actual field research? Physically obvious methodology improvement instead of statistically dubious mummery?
You just don’t want to be taken seriously, do you Anthony?
Oh, yeah? Tell that to me when I’m happily snapping away, lying on my back on a sloping ice-covered rooftop with no retaining wall, five feet from a sheer 40-foot drop to the concrete.
But be prepared to duck; you might rate a snowball or two.
I’m guessing if the US is bad , Africa temp data is armageddon.
I think you missed the sarc tag, Evan.
Um, right.
I also missed the initial comma.
I saw: “Pardon impossible, to be sent to Siberia,” but read: “Pardon, impossible to be sent to Siberia.”
Geoff Sherrington has done some investigation on australian sites. According to his comments at Currys blog. Could that be a guest post at WUWT?
Guidelines for guest posts are listed under Submit Story in the WUWT menu if Mr. Sherrington is interested.
When you look into the the records of well-sited stations, the lack of warming is obvious, as is the effect of adjustments. My study of USHCN stations meeting the CRN#1 standard is here, with supporting Excel workbooks:
https://rclutz.wordpress.com/2015/04/26/temperature-data-review-project-my-submission/
You will find that a disproportionate number of USHCN Class 1 sttions are CRS units. As a result, their trends are even a little higher than the Class 2s. Just shows what lugging its own personal heat sink around does to a sensor.
So, in the future, when the projected warming hasn’t materialized (yet again), the scientists will have to “adjust” today’s data back down to show warming then.
It’s like the data equivalent of whack-a-mole. Pretty soon, NASA and the NOAA will need an entire “Administration of Adjustments” with sub-departments of staff just to keep track of all the changes.
evanmjones,
Your intensive interaction with this thread, as a co-author of the subject research, is very much appreciated by me.
John
Please provide a list of the 410 sites used.
You’ll have to wait until publication. Then, you’ll not only get that, but all the non-complaint sites (plus data), as well, if there are any you think should be included but that we left out.
Reblogged this on gottadobetterthanthis and commented:
During my years of nuclear-related research, my experience convinced me that we cannot measure temperature with a reliable accuracy better than 1.5°C, not under any natural conditions, hardly even under ideal conditions in the lab. Given what Anthony is showing, it is really ridiculous, absurdly silly, to talk about earth warming with the data we have so far.
Sure, we have other evidence, circumstantial evidence, that earth is warming, but our best peleo evidence indicates it has all happened before, even at faster rates, though warmist refuse to acknowledge that obvious fact.
My bottom line is with the people. Acquiescing to the alarmists enslaves people, condemns the world’s poorest peoples to continued starvation and deprivation, and kills people today with all that goes on in the name of “Green.”
Make it personal: Grandma cannot pay the heat bill, and eat, and buy medicine. Left-leaners blame the government and right-leaners, but the the fact is, there is no excuse for the heat bill factoring into this situation. In the West, we can have, and should have, such abundant energy that heating bills are trivial. Instead, Grandma decides it isn’t worth it, and she spends her last night under the stars sleeping in her garden as the frost takes her.
More personal? Think cold, dark operating room with only enough emergency power to keep the vitals monitors running while the doctors open your daughter for a critical, life-or-death surgery, simply because the coal-fired power plants have all been shut down and the wind just doesn’t happen to be blowing.
Remember:
Wind blows, but windmills suck!
Also:
Cold kills. Warmer is better!
+1
No accuracy, no peace
How does what Phil Jones, of the Met Office, said or did apply to a discussion about NOAA measurements?
Abe, I know who he is, and what he did and said, my question was HOW does any of that apply to THIS discussion about NOAA and US weather stations? WHY did you bring Phil Jones up?
The heat-sink hypothesis is an unphysical one. This was pointed out to Evan Jones over a year ago in discussion at Stoat’s. The press release makes no mention of having found a physical explanation. “Heat-sink” in this context is merely a euphemism for: We haven’t found a physical explanation.
Anyone that reflects on what a heat-sink does and how they’re used quickly realizes this is bass ackwards. Heat-sinks reduce trends, not exaggerate them. We don’t put heat-sinks around CPUs in our computers because we want them to run hotter.
I find this whole explanation – or lack of one – especially disappointing because Evan assured us this was easily figured out by their co-author physicist.
First he said, “Our physicist co-author thinks this factor is easy to nail and he does know about the Hubbard paper.”
Later he said, “We will, of course, be hitting it from the physics angle, as well. So it won’t be a statistics-only study. It will be backed by a mechanism that explains why and how (and to what extent) this occurs.”
OTOH, there is a known component of the measuring system that *does* exaggerate highs *and* exaggerate lows – the Dale/Vishay 1140 thermistor used in the MMTS stations. This was documented by Hubbard and Lin, Air Temperature Comparison between the MMTS and the USCRN Temperature Systems (2004).
Since the Menne MMTS Bias adjustments were based on all stations, regardless of microsite, it’s easy to envisage that Menne’s MMTS adjustment isn’t entirely applicable to a subset of the stations. The Hubbard MMTS Bias adjustment is instrument specific – regardless of location or microsite – since it’s just a description of the physical response curve of the sensor itself. But Menne relies on pairwise homogenization while Hubbard & Lin did a year-long side-by-side field study comparison.
While there is nothing wrong with homogenization per se, using the average result from a large group of stations and expecting it to be applicable to all subsets is a leap of faith. It is also unnecessary considering the Hubbard MMTS Bias I adjustment is available. If nothing else, obtaining the same results also using Hubbard would make the results more robust and eliminate the MMTS sensor as a potential physical explanation.
May I point out that the first occurrence of “heat-sink”, much less “heat-sink hypothesis” is in your post. So, is this a red herring? If not, could you provide a link to the “heat-sink hypothesis” to which you object.
Perhaps even the link to that discussion over at “Stoat”. Otherwise I feel this comment is a waste of time.
Stephen Ramsey – I suggest you try reading the main post. Or search this page for “heat sink” if you don’t like “heat-sink”.
After that we can discuss waste of time.
You may be a tad confused on this point. The Menne et al approach applied a custom adjustment to each MMTS transition based on the difference between that station and nearby stations that did not have a MMTS transition. The approach used by Watts et al (as far as I know) applies a constant MMTS adjustment to all stations.
MMTS is somewhat complicated. There is a clear max cooling bias that shows up in most cases, and a min warming bias that shows up in some cases but not all. The max cooling bias is instrumental, but my personal suspicion is that the min warming bias is mostly due to station site changes (since MMTS sensors require an electric hookup, they were often set up closer to buildings than the CRSes they replaced). One thing I’d like to use Anthony and Evan’s work for is to help identify stations that did not move when transitioning to MMTS to help test this hypothesis. Might be a fun little paper in it.
Zeke – Yes, I understand the Menne approach, but that is a totally different approach than Hubbard & Lin’s description of the Dale/Vishay 1140 thermistor results. Hubbard and Lin’s results were based on co-located, side-by-side, field comparison. So their min bias is also instrumental – not due to site changes.
Menne describes the differences in results here: THE U.S. HISTORICAL CLIMATOLOGY NETWORK MONTHLY TEMPERATURE DATA, VERSION 2, where they write: ” As a result, the overall effect of the MMTS instrument change at all affected sites is substantially less than both the Quayle et al. (1991) and Hubbard and Lin (2006) estimates. However, the average effect of the statistically significant changes (−0.52°C for maximum temperatures and +0.37°C for minimum temperatures) is close to Hubbard and Lin’s (2006) results for sites with no coincident station move.”
Considering this is a press release with no reproducible method or data to work with, it’s possible I’m mistaken, but it’s hard to reconcile your interpretation with Evan’s statement:”We do this by applying the Menne (2009) offset jump to MMTS stations at the point of conversion (0.10c to Tmax, -0.025 to Tmin, and the average of the two to Tmean).”
Some comments. First, yes, we apply a constant offset. We do, however, apply it to the month of conversion for each station, so the effect on trend per station will vary widely.
It is a simplification, but not as simplistic as you might think.
Second, MMTS Min warming adjustment is slight. It would make only be +0.0125C to Tmean offset difference to the jump if excluded entirely, which doesn’t stack up to much over three decades.
Third. We are not doing it exactly the way Dr. Menne does it. We are using Menne’s offset numbers as provided by Menne (2009). Menne now does a pairwise homogenization seven years front and back from the point of conversion. Having used that old dodge in wargame design myself (to “simulate” accuracy), all I could do was laugh and shake my head. Not so much because he did it, but because he actually appeared to believe it. And boy-oh-boy did it ever make that nasty CRS issue fade into the shadows, or what. So that’s what he does.
What we do is add the Menne (2009) offset at point of conversion.
One thing I’d like to use Anthony and Evan’s work for is to help identify stations that did not move when transitioning to MMTS to help test this hypothesis.
Hullo, Zeke. Good question.
A few of them did move a bit, but most (by far) were placed in the same spot. You see quite a few old CRS units still in place with an MMTS right next to it. NOAA appears to use a major move as an opportunity to start afresh with an MMTS. But as those stations are considered to be perturbed, this is moot.
If the rating changes as the result of a localized move, then the station is considered to be perturbed and is dropped.
Also, USHCN metadata has greatly improved. They often record station moves by as little as 3 feet. If you want to examine that, look at NOAA metadata online (HOMR) and the surfacestations gallery, along with an active Google Earth.
Welcome to my world, guys. I will go on a bit.
The heat-sink hypothesis is an unphysical one. This was pointed out to Evan Jones over a year ago in discussion at Stoat’s. The press release makes no mention of having found a physical explanation. “Heat-sink” in this context is merely a euphemism for: We haven’t found a physical explanation.
And there I was, thinking it was a euphemism for, “Gosh, those trends sure average a heck of a lot higher when those houses and cementy things are near the sensor. Wow, look at those Tmin numbers. Well it seems pretty obvious why that is.”
As Dr. Leroy put it: the quality of observations cannot be ensured only by the use of high-quality instrumentation, but relies at least as much on the proper siting and maintenance of the instruments.
He refers to “heat sources”, writ large. We refine the observation to distinguish that which generates heat (“heat source”) from that which does not generate heat, but absorbs and re-radiates it (“heat sink”).
Well, anyway, you don’t seem to think much of the term, that’s obvious. Or we wouldn’t still be going on about it after all this time. Is it possible that what you find bothersome about all this is that the words “heat sink” sit so well on the tongue?
Dr. Leroy wasn’t looking at the trends when a station is exposed to “heat source” (which, by his definition includes sources and sinks), but offset. What we do is use his rating system and then look at the trends of the stations thus rated. In your haste to remind be to stick with the trends, I fear you have strayed into the land of offsets a bit, yourself. Besides, being colder does not mean you are not warming faster, as the Arctic guys like to say.
Anyone that reflects on what a heat-sink does
What a heat sink does is reflect.
and how they’re used
Well, in greenhouses, they’re used to take the edge off Tmin and bump up Tmax. That’s the offset effect, anyway. You wouldn’t know how that would affect trend during a warming interval until you measure it, of course. You guys remind me of the story of the dude who got tossed out of the Aristotellian tribe for the crime of instigation to commit empiricism.
quickly realizes this is bass ackwards.
I recommend realizing a little slower.
Heat-sinks reduce trends, not exaggerate them. We don’t put heat-sinks around CPUs in our computers because we want them to run hotter.
You are talking offset. You need to be be thinking trend. I could just leave it at that.
A CPU is a heat source. It is generating its own heat. It is the hottest thing in the room. A CPU is generally located in an enclosed space, and is likely not exposed to get much sun. So the heat sink is taking up energy generated from the computer — a closed and trendless system.
Placing a heat sink next to a computer when sitting outside on a sunny lawn is not going to cool it down. Both the sink and the computer are receiving radiation from both the sun and the surrounding atmosphere. The heat sink is absorbing more energy from the sun than it is from the CPU, then re-radiating some of it back towards the CPU, recorded only at at Tmax and Tmin. Not to mention the general lack of nocturnal/dinurnal variation of a room in a building. When is Tmin inside a closed, artificially controlled environment?
So if anything, the heat sink will be marginally increasing the heat of the CPU at either Tmax or Tmin, which are only times the temperatures are recorded by USHCN. Not that this is much of a practical issue outside a closed room.
I find this whole explanation – or lack of one – especially disappointing because Evan assured us this was easily figured out by their co-author physicist.
I have no doubt that you do. I think I can feel your disappointment radiating off you at Tmin. We never managed to land him, unfortunately. We’ll have to get back to it.
Please note that I was being starkly open about our process, far more than any other paper I’ve seen. Perhaps too open. But the idea is to operate as much as possible in the open. That’s what we do.
First he said, “Our physicist co-author thinks this factor is easy to nail and he does know about the Hubbard paper.”
Well, that work hasn’t been done yet. It will have to wait for followup.
Later he said, “We will, of course, be hitting it from the physics angle, as well. So it won’t be a statistics-only study. It will be backed by a mechanism that explains why and how (and to what extent) this occurs.”
The best laid schemes of mice and men gang aft agley. We can (and do) describe the mechanism, but we are going to need someone to add in the formulas. We’ll address this in followup.
OTOH, there is a known component of the measuring system that *does* exaggerate highs *and* exaggerate lows – the Dale/Vishay 1140 thermistor used in the MMTS stations. This was documented by Hubbard and Lin, Air Temperature Comparison between the MMTS and the USCRN Temperature Systems (2004).
Groovy. We already add an MMTS adjustment offset. When we publish, I will supply a tool that will allow you to drop in whatever MMTS numbers you like better than ours. Either by formula or by swapping in a new MMTS-adj dataset.
Let us know when you do. We would find the results interesting.
But in any event, it won’t be enough of a bump to change things much over what we already did. Maybe 0.01C/decade on the outside.
And speaking of gluteal direction, all you guys think about is how to horsewhip the MMTSs in line with the CRSs. It never seems to occur to you that it’s the CRS units that are the actual problem in the first place — carrying your own personal heat sink around on your shoulders wherever you go will do that. Especially as the paint fades (net).
It’s the CRS units that are giving the spurious results. And, as the MMTS units were calibrated to the CRS units, I see little real justification even for adding in the offset jumps. Either that or the calibrators have some ‘splaining to do. But, being a swell guy, I’ll go along. For now.
It is possible that the offsets should remain — and don’t think I won’t be looking at pairwise to check. But it is glaringly obvious that the CRS trends, esp. Tmax are going to have to be adjusted down. Way down. And that has implications that are going to shake the chain all the way back to 1880.
I think it’s youse guys, not me that have things reversed.
Since the Menne MMTS Bias adjustments were based on all stations, regardless of microsite, it’s easy to envisage that Menne’s MMTS adjustment isn’t entirely applicable to a subset of the stations. The Hubbard MMTS Bias adjustment is instrument specific – regardless of location or microsite – since it’s just a description of the physical response curve of the sensor itself. But Menne relies on pairwise homogenization while Hubbard & Lin did a year-long side-by-side field study comparison.
Just plug in Menne’s data. MMTS adjustment only data is available from NOAA if you care to do that. Or H&L. Besides, a little bigger or little smaller offset isn’t going to matter here. What’s going to matter is the bad CRS bias. You are the ones looking at this backwards.
While there is nothing wrong with homogenization per se, using the average result from a large group of stations and expecting it to be applicable to all subsets is a leap of faith. It is also unnecessary considering the Hubbard MMTS Bias I adjustment is available. If nothing else, obtaining the same results also using Hubbard would make the results more robust and eliminate the MMTS sensor as a potential physical explanation.
There is nothing wrong with homogenization per se, if there is no systematic error in the data. Then it is kindly Uncle H. but when a systematic error is introduced to the data series, Kindly Uncle H goes postal. This is a known thing.
Yet I see no reason you can’t sub in Hubbard’s data. You could even do it station by station. You can be provided with excel sheets that will enable this process when we publish. But even if the bump in trend is double ours, it’s not going to affect our results much.
In Engineering a heat source is at a higher temperature than a heat sink. They are both considered (in the ideal) to have unlimited capacity.
In the context of the paper a “heat source” can supply energy not available from the environment. A heat sink can not supply energy independent of the environment.
What surprises me is that they had to ask Dr. Phil in the first place. Have they no graphs? No Excel? No eyes? What?
THIS Dr. Phil?
I can’t see either of these papers being in the next IPCC report. Kevin and I will keep them out somehow – even if we have to redefine what the peer-review literature is!
Cheers
Phil