Plus answers to yesterday’s Fun puzzle: Name these official stations.
Given that California Governor Jerry Brown has recently setup a website at the governor’s office basically telling skeptics to “shut up” I thought this would be a good time to publish this.
This is a paper that was presented at a climate conference by Jim Goodridge, former State Climatologist of California, titled Population and Temperature Trends in California at the Pacific Climate Workshop, in Pacific Grove, CA March 22-26, 1987.
In this paper, Jim presented what I believe to be the very first photos bringing attention to the issue of station siting. Yesterday, I published both of those photos on WUWT here: Fun puzzle: Name these official stations.
The answer to the first photo was correctly made by commenter “Hoser”:
Hoser says:
August 15, 2012 at 10:32 pm
It’s been a long time, but the top one might be Mt. Hamilton, Lick Observatory. That might be the astronomer’s dormatory behind the car. Yikes, 25 years since I’ve been there.
Yes, the official temperature at the Lick Observatory is measured on a concrete slab rooftop where cars can park and there’s a chimney nearby:
Surprisingly, that station is still in operation today. It has been converted to MMTS electronic thermometer, but from what I can tell, still appears to be at the same location as before. Note the walkway bridge and chimney shadow:
Interactive source map: http://binged.it/PscDx2
NOTE: Perhaps one of our WUWT readers in the Bay Area can make a trip up to the Lick Observatory this weekend to advise with a photograph if the station still exists on the same spot or not. You’d think that on such a hallowed grounds of science, they’d know enough to put the thermometer away from the chimney and concrete. Let’s see if they’ve figured it out in 25 years since then.
As for the other station near the incinerator, that is a Taylor max-min thermometer used by the Quincy, CA Highway Department, now since closed. Nobody got that one, but there were some good guesses. 
Siting issues aside, Jim made some important discoveries in this paper where he looked at rural -vs- urban temperature trends. He only has a paper copy left, as the Mac disks this was done on have long ago been lost. I took the paper copy to Staples and had it scanned into a PDF file, which is presented in full below.
This page 9 of graphs below, figures 4 and 5 tell the story for California Surface Temperature data:
Mind you, this is data that Jim used prior to the big range of adjustments that have been applied by NCDC. Jim provides all that data in the paper. It might be interesting to compare the data then and now to see what has been done to it. Another important distinction of note is that this paper was presented over a year before NASA’s Dr. James Hansen went before the Senate in June 1988, and touted his science and model predictions, deeming it so solid that they had to turn off the air conditioning in the hearing room for “theatrical effect”.
Figure 6 and 7 on page 10 are also instructive:
But this set of graphs from page 12 is really interesting:
The trend for rural stations is interesting, because Jim found a correlation for it:
Here are figures 15 to 18:
And here is Figure 19. Indeed the similarity is remarkable.
The other conclusion to Jim’s paper is that there is a correlation between population trend and temperature trend for inland urban stations, as seen in this graph:
Jim eventually went on to publish a letter in the Bulletin of the American Meteorological Society in 1996 on this issue. This one graph from that letter was a “light bulb moment” for me:
The reply from Kwang-Y Kim, published next to Goodridge’s letter is an interesting admission:
Kim had co-authored a CO2 regional modeling paper with Gerald North in 1995, suggesting that temperatures were on the rise to CO2, but Goodridge in his letter had suggested their base temperature data had been polluted:
I have to wonder, if somebody had put Goodridge’s 1987 paper in front of Jim Hansen in 1987 or early 1988, would it have made any difference in his claims made in June 1988 before the Senate?
Probably not, because as we’ve seen, there seems to be an unwavering belief system that climatic scale temperature is controlled only by Carbon Dioxide concentration, and anyone who presents a contrary view is immediately denigrated and labeled. For example, Hansen’s CRU compatriot Dr. Phil Jones already had formed a strong opinion of Goodridge’s work, which we see thanks to Climategate 2 (bold mine):
file 4789.txt
date: Tue, 25 Apr 2000 09:25:14 +0100
from: Phil Jones <p.jones@xxxx>
subject: Re: CA climate
to: Tom Wigley <wigley@xxxx>,Mike Hulme <m.hulme@xxxxx>
Tom,
Bryan Weare is at US Davis. He would know about some of the things you
mention. The jerk you mention was called Good(e)rich who found urban
warming at all Californian sites.
I’m away until today until May 5 in Nice and Geneva. I hope you can do
the temperature plots yourself and that Mike can do the precip ones.
Mike has the data as 5 degree grid boxes, so the it would be good if
you could define these for him. I think he’s back tomorrow.
It would be possible to use the 0.5 degree grid boxes but we’d have to
get Mark New to do that for us.
Cheers
Phil
At 12:13 PM 4/24/00 -0600, Tom Wigley wrote:
>Phil and Mike,
>
>I have to attend a meeting organized by EPRI and the California Energy
>Commission on June 12, 13. The focus is future climate scenarios and the
>implied impacts. It will include discussions of GCM results and
>statistical and LAM downscaling. They want someone to address observed
>climate (homogeneity problems; E-W and N-S contrasts; ENSO effects;
>changes in circulation — such as increased offshore cyclogenesis, changes
>in storm tracks; etc.), but they don’t have anyone invited yet. Chuck
>Hakkarinen (EPRI) says there is someone at UC-Davis who is an “expert” on
>CA climate. Who is this? Do you know any other Californians who are in
>the observed climate game and who you respect? (From memory, there are
>some nitpicky jerks who have criticized the Jones et al. data sets — we
>don’t want one of those. Wasn’t one of these guys called Goodrich?)
>
>For myself, I would like to have some monthly time series for the CA area
>average. I can possibly do this for temperature, but certainly not for
>precipitation. Is there any way you two could send me time series within
>the next day or so (before I leave for Australia)? For the regions, I’d
>like results for the following separate areas (as near as you can do it):
>(1) 32-36degN, 115-121degW
>(2) 36-42degN, 118-124degW
>(3) 32-42degN, 114-124degW
>(4) 36-42degN, 106-114degW
>The last one represents the headwaters of the Colorado River.
>
>Finally, if you had some PDSI time series for the region, I’d very much
>like these too.
>
>Many thanks,
>
>
>Tom
>
>
>
>**********************************************************
>Tom M.L. Wigley
>Senior Scientist
>ACACIA Program Director
>National Center for Atmospheric Research
>P.O. Box 3000
>Boulder, CO 80307-3000
>USA
>Phone: 303-xxxx
>Fax: 303-497-xxxx
>E-mail: wigley@xxxx
>Web: http://www.acacia.ucar.edu
>**********************************************************
>
>
Prof. Phil Jones
Climatic Research Unit Telephone +44 (0) 1603 xxxx
School of Environmental Sciences Fax +44 (0) 1603 xxxx
University of East Anglia
Norwich Email p.jones@xxxxx
NR4 7TJ
UK
—————————————————————————-
Tom Wigley and Phil Jones are some piece of work, aren’t they?
The entire 1987 paper by Jim Goodridge is available here as a PDF: Goodridge_1987_paper (16mb)
We owe Jim Goodridge some thanks, not only for the work he has done, but also for the abuse he’s suffered alongside us all from “The team”.
.9 R!
Can’t possibly be climate science, way too accurate.
Excellent post. Bookmarked and flagged. Thank you.
From page 2 of the paper:
So what this paper is saying is that if you divide the weather stations into those with a low slope and those with a high slope, and call the high slope set “urban,” you will find that urban stations have a high slope.
That might explain why in his map of “urban” and “rural” sites many of the rural sites are sandwiched between or right next to urban sites. It also explains why
His method shares something in common with your surface station audit, Anthony: he is not blinded to the trend when he categorizes whether a station is rural or urban. But at least his method is objective, which can’t be said for yours
REPLY: You are welcome to demonstrate why the method we use, Leroy 1999 and Leroy 2010 in station categorization is biased. Just saying it is, doesn’t make it so.
Read up on it here:
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 http://www.jma.go.jp/jma/en/Activities/qmws_2010/CountryReport/CS202_Leroy.pdf
-Anthony
meanwhile, down under, CSIRO have a new big scare:
17 Aug: ABC: Conor Duffy: Climate change sees tropical fish arrive in Tasmania
The CSIRO is warning climate change is having a big impact on the country’s oceans, with tropical fish turning up as far south as Tasmania.
A major report on oceans and climate change, to be released today, says the damage under the sea is much clearer than when it released its last report on the subject three years ago.
As well as causing a southern migration, climate change is causing a decline in some temperate fish stocks and ocean acidification is beginning to affect shellfish…
http://www.abc.net.au/news/2012-08-17/climate-change-sees-tropical-fish-head-south/4203830?section=tas
Good research and comparative analysis, Anthony. Three atta boys and one excellent job. Hopefully this will shut the monkeys up (haha).
Or how about this little gem from page 3:
Wow! That statement really inspires confidence. No wonder this ‘paper’ never got published.
REPLY: It was never submitted for publication, it was presented at a climate conference. Please do try to pay attention. And observation of the station situation is in fact a way to quality control the metadata. For example, explain why the San Francisco station (on a rooftop downtown at that time) was classified as rural. – Anthony
Oh, the stupid…it burns, doesn’t it Anthony? You could try reading the articles, first. Just a suggestion. I mean, this is the second blog post today that you’ve completely whiffed on (the first being the helium article).
Feel the burn, bro.
REPLY: I only feel the haters, such as yourself and faustus. Hate on dude. You are welcome to answer the question I posed to him. – Anthony
RE: faustusnotes says:
August 16, 2012 at 4:26 pm
The comparison between counties with populations over a million and those under a hundred thousand are not so “arbitrarily selected.”
You are just looking for a reason to deny the facts glaring you in the eye. Sad.
These facts have been obvious from the beginning, as has been your ( and Hansen’s) dedicated will-to-ignore-the-facts.
Twenty-five years! A quarter century!
That’s a long time to stay stupid. Sad.
Pity he didn’t separate rural agricultural from rural non-agricultural, as Christy found California agricultural areas were warming, while nearby non-agricultural areas showed a slight cooling trend. Christy ascribed the agricultural area warming to albedo changes from increased irrigation.
http://journals.ametsoc.org/doi/abs/10.1175/JCLI3627.1
The Team Says:
“His method shares something in common with your surface station audit, Anthony: he is not blinded to the trend when he categorizes whether a station is rural or urban. But at least his method is objective, which can’t be said for yours”
My,my, my!
You seem to have hit a raw nerve Anthony. Just typical National Socialist Warmista personal attacks. They are a [snip] that can’t really debate the issues. But then we don’t really expect much from them in the way of informed debate, sadly.
Thanks for sharing Jim’s interesting work.
Rich
One possible hypothesis for the warmer SST stations is their relationship to sewage water outflows, storm drain outflows, or power plants outflows.
[snip]
lol – yeah, probably right, but gezzz.
faustusnotes, you should have read past page 2. On page 3, Jim Goodridge wrote that, “In general the classification of the temperature records as urban or rural is fairly close to reality as the writer knows it from viewing most of the sites. (my bold)”
JG also mentions that both San Francisco and Vancouver BC are “well ventilated” coastal cities, in which local SSTs and air temperatures are well-correlated.
Even a cursory inspection of the station map in Figure 3 shows that your comment, “many of the rural sites are sandwiched between or right next to urban sites.” is wrong. Also, anyone familiar with the folded topography of California would not worry much about nearby spots on a map. The top of San Bruno Mountain, for example, is about 4 miles from the center of San Francisco, is almost completely rural (apart from broadcast repeaters), 1000 feet high, is constantly breezy and would overlap the spot for SF on that map.
Having spent some time in Cloverdale, this commenter can attest that it is indeed “rural”. (But don’t forget those orchard heaters!). It does suggest that the old-timers were right — it really was hotter in the ‘thirties.
I have always been suspicious – skeptical – of data acquired for a purpose, acquired with an objective in mind, for the potential of confounding corruption. This convenient data, predating our controversy, is doubly valuable.
I’m disappointed that I doubted my impulse to Lick Observatory for the architecture, but I honestly do not remember that spot, born and raised in Santa Clara Valley.
Phil M, visually inspecting the station is central to making the urban/rural distinction and to validating the temperature record. Your dismissal betrays ignorance. The other critical validation, of course, is regular field calibration of the instrument, but that’s another issue.
Pat Frank says:
August 16, 2012 at 5:34 pm
‘faustusnotes, you should have read past page 2. On page 3, Jim Goodridge wrote that, “In general the classification of the temperature records as urban or rural is fairly close to reality as the writer knows it from viewing most of the sites. (my bold)”’
Sorry, Pat, but the Warmists will not be able to understand what you have written. For example, you quote: “the writer knows it from viewing most of the sites.” The vast majority of Warmists will assume that the reference is to viewing photos on the internet. It will never occur to them that Goodridge traveled to the sites or that he could classify them by visiting them. It is all too practical for the Warmist. /sarc off
Figures 10 (urban) and 11 (rural) have the exact same abscissa values. The latter appears to show rural stations with 22 and 26.5 million populations.
[I’ve lost this comment twice, already, when I clicked on a chart while trying to verify something, and was sent off to yaya land. I hope it didn’t end up posted.]
Reblogged this on Is it 2012 in Nevada County Yet? and commented:
This is an important paper for all California readers. It clearly shows that rural areas are cooler than urban areas.
You’ve just convincef me that AB 32 will indeed fix global warming in California. As the cost of doing business rises, many businesses and their employees will leave and the state population will decline. The de-urbanization should lead to lower temperatures. Those folks at CARB are brilliant!
“from viewing most of the sites”
yep.. like actually removing your butt from in front of the imagination model on the computer, and viewing reality.
Maybe a warmist… just one, somewhere, will try this…. eventually……maybe, perhaps
nah.. not a chance..
The PDF of the “Comments on Regional Simulations of Greenhouse Warming Including Natural Variability” – Goodridge (1996) is available here:
http://pt.scribd.com/doc/67524224/PDF-1996-Goodridge-Comments-on-Regional-Simulations-of-Greenhouse-Warming-Including-Natural-Variability
and Fartsus.. if you want to prove the categorisation is wrong, then go and do a history and site check on each and every one of the stations used.
Until you do that.. you got NOTHING !!!
The small-population counties in the “light bulb moment” graph shows a distinct cooling trend if one removes the first 15 years of data.
This, alone, completely falsifies the warmists’ cherished belief that increased atmospheric CO2 causes warming.
If the warming-due-to-CO2 is valid, it must be consistent. It cannot act arbitrarily nor capriciously.
I always ask the warmists, “How does it know which cities to ignore (or counties, or states)?”
Thus far, there has been no reply.
So Anthony, no comment on the methods of the paper? Do you agree that the classification method makes the argument tautological, and if so are you going to update your post to include at least an acknowledgement of the tautology?
Pat Frank, “viewing most of the sites” is not objective. No doubt the US Census Bureau or the Post Office (or the CDC) holds a database of urban/rural classifications for every settlement and location in the USA, and that data could be used to classify the stations as rural/urban. Note also he says “most” so he doesn’t make clear how many of the sites don’t match his (regression-based) classification. Even his chart of population-based classifications isn’t objective, since it doesn’t use established definitions of urban/rural and it appears to rely on pooling data from multiple sites, some of which could be rural and some urban.
Richdo, there is no personal attack in my comment, which is purely a criticism of methods. It’s ironic that your own comment containing this complaint had to be snipped because of … a personal attack.
Goodridge has clearly divided the data set into two at the median regression slope (this is why there are 39 sites in each group). The median regression slope is >0, so that means that at least half of the stations had a warming trend in 1987. This means some of those classified as “rural” by his (regression-based) method probably have a warming trend. Given that Goodridge’s arbitrary classification system places some rural stations into the “urban” category, this means that warming trends were being observed in both urban and rural stations in 1987.
I would be interested to hear why anyone thinks that the station shown in the picture would be contributing to a misleading understanding of global warming. If AGW is not happening, by what mechanism is that station showing a warming trend that would mislead naive scientists into thinking warming is happening? And how could that mechanism be replicated across a majority of such sites in the contiguous USA?
REPLY: “Goodridge has clearly divided the data set into two at the median regression slope (this is why there are 39 sites in each group).”
Umm nooo, you can’t read or do simple math apparently. Figure 4, 35 Urban stations, Figure 5, 39 rural stations. Note that on page 2, he identifies 74 stations in his dataset.
39×2= 78, not 74.
35+39 = 74
So much for your argument. – Anthony
Excellent Post, Anthony. What a revelation! And so early on that he cannot possibly be called a “denier”, I mean it resonates with pre-Hansenite scientific innocence.
Can I suggest that a Goodridge Study be carried out on all sets of stations world wide, I guess along the lines of your study in the US, a Wiki-Goodridge evaluation to coin a term. We could then aggregate the results to get a truly GOOD estimate of global temperature because clearly the BEST one is not what it seems. It would of course be termed the Goodridge Temperature in the man’s honour.
The very fact that he CAN select 39 stations that have, on average, a negative temperature trend, makes a TOTAL MESS of the term Global Warming..
gees, it isn’t even State Warming,
just URBAN warming.
Still no comment on the fundamental tautology or on the subjective nature of the classification process?
So he hasn’t divided the stations on the median slope. So can you explain why he chose to divide them at that particular slope value? Why not 0? Is it the mean slope? There’s no objectivity in his method at all.
M Seward, if you used the same method on all stations world wide you would get the same result as Goodridge because it’s tautological. The method assumes the result.
Imagine if a climate scientist did the same thing: “stations were defined as ‘accurate’ if the regression slope of temperature was greater than 0.01. Accurate stations showed a significant warming trend, while inaccurate ones showed cooling. This is clear evidence that the world is becoming warmer.”
You guys would have a field day with that, wouldn’t you?
REPLY: It’s a simple prescreening testing for an early result, to present at a climate conference of some peers. This was his first ever test of the issue IIRC, so yes the sample size is small. Remember this is one man working alone in spare time, not a funded project. His later work used a much larger set of stations, and it still holds up.
For example, from a post I made in 2007:
LaDochy, S., R. Medina, and W. Patzert. 2007. Recent California climate variability: spatial and temporal patterns in temperature trends. Climate Research, 33, 159-169 You can download the paper in PDF format in its entirety here: ca_climate_variability_ladochy.pdf references Goodridge’s work. So it must not be as terrible as you try to make it out to be.
Here’s an updated chart from 2009, using hundreds of California COOP stations.
You are so busy trying to shoot it down you can’t even bother to look around. – Anthony
faustusnotes says:
August 16, 2012 at 6:26 pm
….I would be interested to hear why anyone thinks that the station shown in the picture would be contributing to a misleading understanding of global warming. If AGW is not happening, by what mechanism is that station showing a warming trend that would mislead naive scientists into thinking warming is happening? And how could that mechanism be replicated across a majority of such sites in the contiguous USA?
==============================================================
Where I work we have data from 2 remote air temperature sensors. A few weeks one gave a reading of 106*F. The other read 89*F. I checked with the airport and it reported 88*F. So what was the temperature here at that time? 88? 89? 106? 94.33?
Well done Anthony. You are probably aware of the Torok et al paper “Urban heat island features
of southeast Australian towns” in Aust.Met.Mag 50 (2001) They found the relationship
ΔTu-r(max) = 1.42 log(population)-2.09
and say the UHI effect is likely to be smaller in Australian towns and cities than in North America and Europe. The article says Tapper, N.J. 1982. Atmospheric infrared radiation over Christchurch,
New Zealand. Ph.D. thesis, University of Canterbury, 384pp. found similar result in NZ. Oke (Oke, T.R. 1973. City size and the Urban Heat Island. Atmos. Environ.7, 769-79.) did a study including Hamiliton Canada and found higher UHI.
It is amazing that so-called climate scientists make upward adjustments of temperature of badly located weather stations rather than downward.
I have an outside temperature gauge on my SUV and just about every day I notice the higher temperatures 1 to 3C (day and night) at the town centre compared to the semi-rural outskirts where I live.. In rural & forested areas further out the temperature is about 1C lower.
Anthony, I didn’t comment on the sample size. He could have a trillion weather stations and it doesn’t change the fundamental problem in his classification method – a problem that persists in your chart. How is aggregated temperature at the county level indicative of rural vs. urban heat effects? You have to analyze that using station-specific rural/urban classifications, which was not done in the original paper and is not done in the post you link to. And you need to do it with an objective classification of rural/urban that is independent of the data series – something that neither Goodridge nor you have done. Classify the stations according to the US Census Bureau’s definition of their local area and try again. Repeating Goodridge’s analysis with another 10 years of data doesn’t make it right – just wrong for longer.
I note your charts show a warming trend across all the stations, just stronger in the most heavily populated counties. Though of course you don’t give measures of uncertainty, as ever, so we can’t tell if the trend is significant. It’s possible of course that the largest urban areas are situated in areas most likely to warm fastest (flat lowlands, coastal areas, and low altitude locations?) so even then the difference in trends isn’t necessarily indicative of a UHI effect. You guys are always going on about correlation not equaling causation. This is a classic example of that problem…
REPLY: You’ll have to be more specific when you say “I note your charts show a warming trend across all the stations…” not a mind reader, please reference it. Several have been provided. If you want precise answers, ask precise questions.
BTW what do you do at the University of Tokyo? Are you James Annan who works from there and blogs here? or maybe this guy http://faustusnotes.wordpress.com/ who is caught up in the gaming scene? I just like to know who I’m talking to when somebody gets so wound up.- Anthony
faustusnotes, viewing the station sites is as objective as viewing atomic emission spectra, or any other data. You’re implying a categorical dismissal of objectivity. There’s no reason to suppose that Jim Goodridge did not make a professionally valid examination of the sites.
Theo Goodwin is correct — actually traveling to the sites is the only way to validate their records in terms of sensor placement. That’s what Jim Goodridge did. Your criticism of him on that account fails.
RE Pat Frank,
“faustusnotes, viewing the station sites is as objective as viewing atomic emission spectra, or any other data. You’re implying a categorical dismissal of objectivity. There’s no reason to suppose that Jim Goodridge did not make a professionally valid examination of the sites.”
Quite so. Jim was the State Climatologist for California. It was his job to know and visit all the weather stations that provided him with data.
*** MODERATORS ***
Would you please revisit:
“The entire 1987 paper by Jim Goodridge is available here as a PDF: Goodridge_1987_paper (16mb)”
I don’t want to be nitpicking, but as it is written, the “16 mb” would read “16 mili bits” (ie. 16/1000 of a bit), which is very different from intended “16 MB” reading “16 mega bytes”, ie. 16,000,000 bytes (and yes, I know that these are not parts of SI, but still…).
Please, don’t take me wrong. Mistakes such as this are commonly tolerated in ordinary daily newspaper articles, but this is, after all, a science blog – and the best one we have.
Thanks,
Bob
I have a rather strong memory that in January of 2006 a WUWT post was about the rural temperatures in inland California and showed that there was no warming. I have not been able to find this again and the archives do not cover this time. Also, memory could be fuzzy.
Would someone know of this?
Well, whether intended for publication or not, it’s flawed and needs to be corrected. Peer reviewer faustus has pointed out a genuine mistake in methodology. Peer reviewers are not obligated to rewrite a paper or to prove that the methodology results in incorrect results. It falls upon the authors to make the necessary corrections.
AndyG55 says: “The very fact that he CAN select 39 stations that have, on average, a negative temperature trend, makes a TOTAL MESS of the term Global Warming..”
Not quite. We know that the Earth has warmed from the Little Ice Age (LIA), so there must be a general upward trend, since, yes? Yet we clearly observe negative trends in many areas. Neither the observations nor the “climate models” have established that putative global warming or real, post-LIA global warming will result in similar trends in all locations. There will still be microclimate–chaotic, dynamic, inscrutable.
Such disparate results in a single State may (or may not) be statistically unlilkely, as Andy alleges. Goodridge’s anecdotal, hand-waving assessment of the validity of his rural/urban station assignments makes me suspect that his results are correct. But anecdotal assessments, based on “viewing most of the sites,” are not science, especially after the fact. Confirmation bias, you know?
The paper needs to be rewritten after applying a non-arbitrary, independent algorithm for rural/urban assignments. At a minimum, the author needs to document his validity test or, better, have the assignments done by others in a blind test. I hope no one is offended by this. Let’s all step back and take another look.
REPLY: The simpler thing to do is put up one of Jim’s later papers. As I mentioned, this was just a first effort for a conference presentation. I doubt he’ll re-write it because he’s got later versions that dealt with the problem using a larger sample size and different methods. – I ask for one of those to put up the next day or two. Bear in mind that Jim was the state climatologist, and it was his job to know about the stations that provided him data. He visited most every station in the state, something I wonder if other state climatologist bother to do today. So if he says the station is rural or urban, he likely has firsthand knowledge of the fact.
After he retired, I recall going on a road trip with him (spring 1992) to visit a station called Four Trees in the Feather River Canyon. A 4wheel drive was required to get there, and I gladly obliged since I had one. He was interested in this station because it showed the largest rainfall value in the state, and he wondered if the equipment had malfunctioned. He wanted to know firsthand.
You can look at data all you want, but it won’t tell you if a tree branch has dipped into the rain gauge. Only firsthand observation will tell you that. I find it comical that so many people miss this important part of the data quality process, and get so bent out of shape over his choices. I know the man, and he knows just about every weather station in the state. – Anthony
Following up on Pat’s comment: August 16, 2012 at 4:52 pm
“meanwhile, down under, CSIRO have a new big scare:
17 Aug: ABC: Conor Duffy: Climate change sees tropical fish arrive in Tasmania
The CSIRO is warning climate change is having a big impact on the country’s oceans, with tropical fish turning up as far south as Tasmania.”
Very interesting.
Up here at 19° South we have had a decidedly cool winter in some ways. Local BoM puts it down to lower SST. So – cooler up here, but so warm down there the fish are moving south?
How come the fish know the sea 3000km south of here is warmer? I suppose it’s possible …
Or maybe somebody got the idea from “Finding Nemo”?
Looks like a lot of people think that actually going into the field and making first hand observations to collect data is wrong and that sitting behind a desk and writing ” adjustment ” algorithms is the correct way. Welcome to the weird world of computer modelling and statistics without any science behind it.
Gunga Din says:
August 16, 2012 at 7:07 pm
faustusnotes says:
August 16, 2012 at 6:26 pm
….I would be interested to hear why anyone thinks that the station shown in the picture would be contributing to a misleading understanding of global warming. If AGW is not happening, by what mechanism is that station showing a warming trend that would mislead naive scientists into thinking warming is happening? And how could that mechanism be replicated across a majority of such sites in the contiguous USA?
==============================================================
Where I work we have data from 2 remote air temperature sensors. A few weeks one gave a reading of 106*F. The other read 89*F. I checked with the airport and it reported 88*F. So what was the temperature here at that time? 88? 89? 106? 94.33?
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
Here’s what I first had typed before I decided to take out any info that might give an indication conditions of the sites. The point I was trying to make (the hard way, perhaps) is that without knowing the details of the sensors location, all you have is a number with no way of knowing if it’s really telling you what you think it is.
“An indirect answer. Where I work we have data from 2 remote air temperature sensors. One is a fairly new installation. It is on the south side of a brick building, in full sun over concrete. The other is about 4 miles away on the control house of a dam. A few weeks ago the newer gave a reading of 106*F. The dam read 89*F. I checked with the airport (about 4 miles in the opposite direction of the dam from the 106 sensor) and it reported 88*F. So what was the temperature here at that time? 88? 89? 106? 94.33?”
If the data from the sensor is only as good as the location and integrity of the sensor. Hansenites would have said the temperature was 94 or maybe even 95. Read Watts et al. Some “Climate Scientist” aren’t naive. They have an agenda or a dollar to make.
Is Al Gore a climate scientist? Do you consider him to be any kind of authority on the environment?
Mann. A scientist in any field may be personally disapointed if his hypothesis is shown to be wrong. But does he hide his data and his methods so that his hypothesis can’t be challenged?
Hansen. Does a scientist pretend his 100 year “projections” are right when they’re already way off just a decade or so out? Is “Coal Trains of Death” a scientific or a PR term?
Take a step back and think about it. The Earth can spare a few moments.
jorgejafkazar: “Peer reviewer faustus has pointed out a genuine mistake in methodology. ”
Ah yes, the Heckler’s Veto. But then I object to your peer review of peer reviewer faustus’ peer review so your post isn’t fit for publication. And if you have any decency you’ll remove it post haste. This is absurd of course. But the problem is that your post is capricious, non-responsive, and makes no attempt to establish the ‘genuine’ nature of any of alleged mistake.
Nor does Faustus do any better in kvetching about ‘arbitrary’ partitions. Zero degrees celsius is arbitrary as is zero degrees farenheit. As well 1 inch, 1 meter, or 1 second. So is the choice of any baseline period for ‘normal’ in climatic terms; be it 1945, 1954, when the Vikings were settling Greenland or during the last ice age.
The entire complaint raised in peer reviewer faustus’ peer review is a red herring to distract from the issue that temperature trends correlate with population. If one wants to object to drawing conclusions about UHI then the only appropriate objection is that correlation is not causation. And that it may simply be that biological organisms prefer warmer climes over colder ones. Thus it is the anomalistic temperatures themselves that caused the population increases; and thus explain the correlation.
But irrational and unthinking nonsense about the arbitrary notions chosen for clarity is itself arbitrary. And if arbitrariness is dispositive then peer reviewer faustus’ peer review is unfit for purpose. And your even more arbitrary sycophantism is even less fit still.
Venter;
Welcome to the weird world of computer modelling and statistics without any science behind it.
>>>>>>>>>>>>
I was about to make a rude and sarcastic remark about being able to build a computer simulation that would produce a global warming result no matter what the data. Then I thought…. Oh Mann, better not, someone might take me seriously and actually do it.
“There will still be microclimate–chaotic, dynamic, inscrutable.”
Yep.. Its called the Urban Heat Effect.. and it DOES affect urban thermometer readings….. A LOT,
but it is basically discounted by the likes of BEST and GISS, neither of whom has even BOTHERED to go out in the field to check the historical site changes.
Yet when someone does actually do some real research, gees, don’t the warmists get uppity !!!
faustusnotes wrote, “You have to analyze that using station-specific rural/urban classifications, which was not done in the original paper and is not done in the post you link to. And you need to do it with an objective classification of rural/urban that is independent of the data series – something that neither Goodridge nor you have done.”
The plain statements in the paper show that you’re wrong on both counts. Let’s note JG’s comment page 3 again: “In general the classification of the temperature records as urban or rural is fairly close to reality as the writer knows it from viewing most of the sites. (my bold)”
In the context of the paper, Jim Goodridge has an a priori knowledge of site quality, from having inspected the various stations as part of his job as CA State Climatologist. He then collected temperature records, and assigned to “rural” those he, “arbitrarily selected as those with a slope (B) of the regression analysis less than 0.0125.
We’re doing science here. The selection procedure amounts to a hypothesis that rural stations have a smaller slope. Choose the stations that have a smaller slope. Check whether those stations so-chosen are, in fact, rural. What does JG find? That, “the classification of the temperature records as urban or rural is fairly close to reality” based upon his direct knowledge of the stations.
What does “fairly close to reality” mean? Quantitatively, we don’t know. If we trust JG’s professional acuity and honesty, we can surmise the he means a correlation — between low slope and rural standing — strong enough to validate the subsequent analysis and conclusions.
JG then compared the urban/rural categories against SST, which are not subject to urban heating or artifact-producing heat sources, and which strongly influence land surface temperatures. He found a good correlation between SSTs and the rural category stations, but not between the SSTs and urban category stations. That amounts to an independent confirmation of his categorical distinction.
At this point, JG’s analysis passes the test for a valid study. Your objections are mooted.
You may wish to withhold granting JG the professional acuity and honesty to apply his a priori station knowledge validly, when he decided his slope-determined categorical choices. But if you do that, then you’re required to provide objective evidence demonstrating what would be an explicit charge of dishonesty and/or incompetence. You’ll have to give factual evidence of false science on JG’s part. Failing that, your dismissals have no standing.
Great thread that seems to have attracted some new eyes. Good to see them here. It really increases the value of WUWT.
This paper is interesting and certainly should act as a cautionary. But it would seem not.
The fact that the nit-pickers are dropping discussion-nits all over this effort to quantify simple historical sampling error / bias tells you all we need to know about the current state of the data.
The nit-pickers are getting soaked by their tempests in teapots. But I can not blame them entirely, since the data sucks and nobody seems to be bothered by its lack. Good replicate random sampling data would have flattened these idiotic objections by the nit-pickers.
Until an adequate number of randomly selected sites with replicated instrumentation weather stations are created, we will NEVER be able to determine statistically accurate temperature and other environmental values with known error values, sources of error, and variance.
Adequate data should never need the “adjustments” that further taint the vast majority of current data sets with N = 1. Since there would be enough replicates to simply discard tainted samples.
Garbage In equals Garbage Out.
BioBob: “Adequate data should never need the “adjustments” that further taint the vast majority of current data sets with N = 1. ”
Amen. It’s a further interest that adjustments are provided, and considered, crucial in terms of fully working diffusion and heat equations for temperatures measured from wooden buckets on deck. (SST) But that minor considerations such as incinerators, asphalt, heat pumps and other unquestionable macroscopic features are considered happenstance nonsense that do not alter the record in any manner at all. (Leroy, Goodrich, Watts) It’s not simply GIGO, but self-refuting GIGO.
Anthony, I’m not James Annan and my identity and job is unimportant to my contribution to this thread. I didn’t put my blog link in my header because I didn’t want a flurry of attention from this site, but you are right that I am the latter blogger. Also, incidentally, I’m not “worked up” nor am I a “hater.” I’m just raising specific and quite simple points about the basic research methodology.
I can’t refer to specific charts because you don’t label them with Figure numbers, but I think you can infer which chart I’m talking about – it’s the one you just put up in comments.
Pat Frank, Goodridge’s statement about his having checked the sites is full of qualifiers – “fairly closely” “most of the sites.” This is not an objective criterion. Classification requires valid and objective criteria, not just some guy’s opinion – even the state climatologist. I don’t know if there is a gold standard for a well-sited weather station, but I strongly suspect it is not “Jim Goodridge said so,” and I know that most Census bureaus have an objective standard for identifying whether a place is rural or urban – the British ONS has 7 categories, for example, and I’m sure that the US has categories too. Would you object to those categories being given more weight in the definition of what is “rural” than the opinion of a single state climatologist?
I don’t need to give factual evidence of false science on Goodridge’s part because I’m not accusing him of that. I’m accusing him of using a poor criterion for identifying urban vs. rural weather stations, which presupposes the conclusion of the paper and thus does not answer the question he is claiming to answer.
I’m still waiting for someone to explain the mechanism by which the pictured station would give a warming trend if there is no AGW…
REPLY: Ah, OK. There’s a station list and data at the end of the paper, it is pretty easy to check against which ones are rural/urban using a number of criteria. NOAA’s MMS database, NASA GISS database, etc. Even a Google Earth inspection can help. The question is: would you trust a state climatologist with firsthand knowledge of the stations to make the correct rural/urban distinctions a year before “global warming” was launched into the public consciousness by James Hansen? I would.
On that last note: temporal inhomogeneity. That will be the subject of an upcoming post. It is easy to understand when you’ve looked at as many weather stations as Jim and I have. Stay tuned. – Anthony
Folks might find this interesting: http://rankexploits.com/musings/2012/uhi-in-california/
The urban/rural classification method employed in this paper does not support the paper’s claims. This criticism that the analysis is tautological is correct. Sites with low trend are compared to sites with high trend, and the sites with high trend are found to have higher trend than the sites with low trend. This is not a result, it is a recapitulation of the methods.
It is similar in nature to Hansen’s most recent paper, in which he cherry picks a base period with low mean and low variance, compares it to a different period with higher mean and higher variance, and “finds” that the distribution of the latter is higher and wider than the former, and further “finds” that more of the data from the higher, wider distribution fall above an “extreme” criterion than do the values from the lower, narrower distribution from which that “extreme” criterion was derived.
Pat is correct when he says “The selection procedure amounts to a hypothesis that rural stations have a smaller slope.” but that wasn’t how it was analsyed, wasn’t how it was presented, and it certainly isn’t how it is being interpreted here.
It would be interesting to revisit Goodridge’s data, and check to see what differences in siting exist between the low and high trend groups. I would drop the “urban/rural” bit though. It introduces confounding that can hide real effects. Sure, there are urban heat islands. There are also rural heat islands, and rural thermometers are often found in them. Apply comprehensive site quality criteria, and see if Goodridge’s subjective classification that works for “most” sites can be verified quantitatively.
See also the reference to Goodridge 1985 Urban and rural differences in long range temperature trends in California recorded in this paper by G. Kukla, J. Gavin, & T. R. Karl
Urban Warming Journal of Climate and Applied Meteorology, Vol. 25, No. 9 (September 1986) pp. 1265-1270.
“it is pretty easy to check against which ones are rural/urban using a number of cirteria. NOAA’s MMS database, NASA GISS database, etc.”
I don’t know that i would trust either of those sources overly much. maybe they would help a bit, but let me see a written history and analysis of all changes close to, and in the surrounding area.
faustusnotes, JG’s criteria were objective, as indicated by the sites he illustrated as suffering from artefactual temperatures (Mt. Hamilton and Quincy).
Your description of “fairly closely” and “most of the sites” as ‘criteria’ is incorrect. They’re not criteria. They’re descriptive statements. JG’s criteria for site choice must have been those of a professional climatologist. He didn’t list them, but he did provide the noted illustrations of sites subject to artifacts; and the sources of artifacts at Mt. Hamilton are explicitly described on page 3. We can surmise, therefore, that JG applied professional criteria in his correlation of rural/urban vs. temperature slope. JG’s known professional career as state climatologist supports that surmise.
Further, you have not addressed the significance of the correlation of SST with rural land surface temperature, and the lack of that correlation with the urban land surface temperature. That correlation/non-correlation is an independent confirmation of JG’s distinctions and data.
Your construct rests entirely on an implicate insinuation that JG is a careless dilettante. In terms of actual content, you don’t have a case.
Philip Mulholland, you’re right. Kukla, Gavin and Karl 1986 cite Jim Goodridge’s results and take them as valid. There they are, supporting data in a peer-reviewed climate scienceTM paper. Who’d dare gainsay them now.
JJ wrote, “Pat is correct when he says “The selection procedure amounts to a hypothesis that rural stations have a smaller slope.” but that wasn’t how it was analsyed, wasn’t how it was presented, and it certainly isn’t how it is being interpreted here.”
JJ, in terms of the rural/urban hypothesis is exactly how the data were analyzed. JG made an explicit comparison with SST. That comparison amounts to a test that could have falsified the hypothesis, except that it did not. JG even states, page 4, that SSTs are not subject to “near by incinerators and other waste heat sources.”
JG may not have explicitly written in terms of Popperian criteria for science (deduction followed by a test that risks falsification), but it’s very clear that his study incorporated them, and presented them. That is exactly hypothesis-testing and its presentation.
How JG’s work has been interpreted here is functionally irrelevant to its scientific standing. The only relevance here is whether a given person is interpreting JG’s paper in terms of its content, or not. Faustusnotes is not. Your suggestion of tautology is explicitly refuted by the comparison of rural/urban trends with SST.
Likewise, JG’s last paragraph cites the 1986 paper of H. Diaz, who compiled the temperatures at rural stations across northern Canada, Alaska, and around Greenland. Figure 19 in that paper shows anomaly trends from 1881-1981 from these rural stations. There is essentially no trend over the entire 20th century portion of those data. They are similar, in other words, to JG’s rural station trends, as JG pointed out.
The correspondence of Diaz’s northern North American/Greenland rural trend with JG’s California rural trend, but not with JG’s urban trend, constitutes another independent and potentially falsifying test; weaker than the test using the West Coast SSTs but nevertheless a good and independent test that produced a confirmation. These independent tests wreck your claim of tautology, and faustusnotes’ as well.
To summarize, JG divided his temperature data into rural/urban by referencing the slopes of the time-wise trends. He then tested his distinction of rural/urban by matching the station lists with what he professionally knew of their physical site quality. Site quality is an independent and objective criterion. JG then did his comparative analysis. He followed that by testing the results against SSTs all along the West Coast. Local SSTs are a second independent and objective criterion. He finally drew attention to the positive correspondence of his results to the rural trends published by H. Diaz. That is a third objectively independent criterion. There is no tautology in JG’s study.
You and faustusnotes plainly have no case.
Figures 4 & 5 from Goodridge in Anthony’s lead-in above, are similar in many ways to Warwick Hughes’ 1992 compilation of Australian rural and urban temperatures.
http://www.geoffstuff.com/Warwick%20graphs.jpg
These graphs were created in part to rebut some of the early denial of UHI by Phil Jones, whose work in places has been shown to be so bad that the law has been brought to bear on some data, e.g. the China part of the Nature 1990 letter that included Australian data similar to these. Strange, I’ve not heard an apology from Phil – but it’s only 22 years after the event.
Further relevant reading is at http://www.warwickhughes.com/papers/90lettnat.htm
Pat Frank, that’s incorrect. Goodridge’s criterion for site choice is stated clearly in page 2 of the paper, I quoted it above: if they have a temperature trend greater than 0.0125 per year, they are defined as “urban” and if it is less than that it is “rural.” He did not define them directly in terms of whether they are actually urban and rural. Even if he had, Goodridge is a climatologist, not a geographer, and typically geographers or demographers define whether a location is “rural” or “urban” (see e.g. the ONS’s definitions of “hamlet” and “village” in the UK – these are specific categories). Goodridge gives no indication in the paper of the standard by which he judges whether a site he “mostly” visited is rural. There are objective criteria for these kinds of things.
So once again: Goodridge’s definition of urban/rural is not objective, it presupposes the result, it is not blinded to the temperature trend in the station (as it is defined from that trend) and he is not an expert on definitions of rural/urban in any case.
This question Anthony is interested in can be answered, of course, by finding the site definitions and matching them to an objective criterion of rural (I would suggest a census bureau or mapping organization’s definition, or as someone else has suggested the existing categories in NOAA’s database). Anthony went to great pains in the past to do this analysis using low and high density population counties, but that is a weak proxy (since it pools urban and rural station sites within the counties). A better approach would be to label the individual sites explicitly and then analyze them.
This won’t get at notions of causality though. It may just be that the biggest settlements are more likely to be placed in areas that are vulnerable to warming. Correlation does not equal causation.
Maus says: “…your post isn’t fit for publication. And if you have any decency you’ll remove it post haste….your post is capricious, non-responsive, and makes no attempt to establish the ‘genuine’ nature of any of alleged mistake…red herring…irrational and unthinking nonsense … your even more arbitrary sycophantism is even less fit still.”
You call my “post” non-responsive, but it’s impossible to respond to others before they say something. Now that you have opened your mouth, I shall respond: Your comment is rude, ill-mannered, false, and deliberately insulting. Faustus established the nature of the procedural error most clearly: Data was first sorted by trend and then used to show that one bin had a higher trend than the other. No further “attempt to establish the genuine nature” of his statement was necessary. Constructive suggestions for fixing the paper were made, and Anthony replied in kind.
To those ‘nit-pickers’ – I would say that you have completely missed the point of JG’s work. If this had been taken seriously at the time, (which it perhaps slightly was, by a few, as the above citations confirm) – then some more serious analysis of the data (at that time) could have been undertaken and corrected way before all the autocorrection/algorithmic computing done today and all the subsequent data concerns (Nasa/giss adjustments, etc, etc) avoided !! – (or at least reduced – can’t get rid of the political human ‘adjustments’, I guess!)
My take on the paper, is that it was simply one mans ‘view’ of the validity of station data and, upon closer inspection – some correlation with urbanisation seemed apparent. Now, whether this was actually published or not, is totally irrelevent. Similarly, if there are, shall we say, supposed or intuitive decisions in assigning data to ‘classes’ for analysis – that is different to current methods – then those who say this invalidates the findings really need to re-assess/rewrite using the current methods using the original data and see if there is still any correlation. Conversely, if this showed he was essentially ‘right’ – then the current method needs re-evaluating. Harping on about one guys ‘choices’ and calling them unscientific is rather crass, IMO.
This kind of historical analysis is, and should be seen as, quite valuable. If I were not so busy, I would do the comparison to the currently ‘held’/’used’ historical station data myself – I waonder just how much adjustment has been made………
JJ says:
August 16, 2012 at 11:20 pm
I partly agree – but I don’t see that the grouping of high and low trends seperately is necessarily completely invalid. For example, there are likely to be urban stations, sited in inner ‘rural’ (e.g large parks?) locations which could potentially ‘match’ proper ‘rural stations (especially over the time period we are talking about, less A/c’s, cars, etc). Similarly, there could be so called rural stations, as in the example, located in poor sitings and affected by closeby ‘urban’ effects.
The grouping may seem arbitrary, for his analysis, but it is not necessarily ‘wrong’ and presumably his use of his own staion siting knowledge would have helped that process?
No doubt about it though.. Global URBAN Warming exists !!
Very interesting comparison of urban and rural temperatures by Jim Goodridge. It seems that he was simply looking for the truth based on accurate data from weather stations in California. It may not have been a peer reviewed paper but it clearly demonstrates the UHI effect.
The paper has been cited by H. Akbari who wrote ‘Energy Saving Potentials and Air Quality Benefits of Urban Heat Island Mitigation’, which was published by the Lawrence Berkeley National Laboratory.
http://escholarship.org/uc/item/4qs5f42s#page-1
Mind you Hashem Akbari’s solution to global warming is to paint your town white. The BBC interviewer suggests painting the moon black. :o)
http://heatisland.lbl.gov/publications/hashem-akbari-speaks-bbc-radio-world-service-about-global-cooling
http://heatisland.lbl.gov/
I don’t know if you’re interested in this, but I ran across a 1971 index of California climatological stations, compiled by Goodridge, at http://archive.org/download/jk4climatologicalst165calirich/jk4climatologicalst165calirich.pdf
faustusnotes says that ‘correlation does not equal causation.’ which is quite true….except for rising levels of CO2 causing rising temperature, of course. Hence the Warmist panic now that the data is showing falling temperature as CO2 continues to rise.
Pat Frank, you have now made two references to the correlation of rural trends with SST, but I think you’re barking up the wrong tree. My understanding of AGW science (admittedly limited) is that SST is predicted to warm more slowly than land temperatures. If so, then the correlation with rural temperatures is just an artifact of Goodridge’s method. He has divided the data set into the fast-warming and slow-warming stations. If AGW science is true then it is natural that the SST will correlate with the slow-warming stations, since the SST is itself slow warming. Goodridge has chosen to call the slow-warming stations “rural” but remember, that’s not how he selected them.
The same artifact applies to the correlation between “urban” (that is, rapidly warming) series and population. If you divide a group of time series into “increasing linearly” and “not increasing” then the former will be correlated with any other time series that is increasing in the same dimension (time). A good way to think about this is to imagine that the warming trend is due to a solar cycle, i.e. completely independent of population. Because temperature and population are increasing in a linear fashion, they will be correlated.
What is it you guys are always saying? Correlation does not equal causation? Need to be careful about that …
Incidentally, correlation coefficients are not a good way to compare time series. You need to apply a common differencing function to render both series stationary, and then use cross-correlation functions. That Goodridge didn’t do this is a bit of an oops, but Box and Ljung had only written their work about 10 years previously (I think), and it hadn’t filtered through to a lot of people, so I suppose he can be forgiven. As I understand it (I could be wrong), Mann made no such mistakes in his hockey stick work.
Incidentally, if Goodridge has found an urban heat island effect, then the only solution for handling historical temperatures, since we can’t measure them again, is an adjustment. Which climate scientists apply all the time. But that’s considered to be a big bad no no around here, isn’t it? Even though it’s been going on in every field of science for time immemorial. I guess you guys must all be pointing on weight hand over fist, because you refuse to zero your scales in case you get accused of tampering with the data …
Also Anthony, your original post includes private communication between two scientists that contains the single word “jerk,” and which you seem to think makes them bad people. Since I started posting on this thread, a quick review tells me that I have been called a Nazi, accused of personal attacks, called a heckler, called a hater (by you), called stupid (for 25 years!), told I’m a warmista who can’t understand what someone writes, been accused of getting “uppity” and “worked up” (by you), accused of “idiotic objections by nit-pickers” (ironic, don’t you think, in light of your response to the exact same complaint in those private emails), and you’ve had to censor one commenter for using a bad word about me. All in the space of what, 60 comments? In a public forum, as opposed to a private email. Also, even though I clearly don’t want my identity revealed, you’ve tried to do that, including publishing my place of work, in a forum full of hostility.
Do you think that might provide some perspective on that part of your post, and do you think maybe your cohorts here might need a bit of a reminder of the boundaries of polite debate? The behavior in the last 60 comments makes it clear that the scientists you quote in those personal emails are way, way better behaved than a good third of your commenters.
You should all think on that.
REPLY: Those are now public emails, known worldwide now. I make no apologies for printing public domain information. As for publishing your place of work, that is public information too. Read the policy page “know your opponent”. I’ve seen so many like you that fit the mold you represent here that attack here, so maybe I’m a bit sensitive. I’m just not too worried about it if you are upset. – Anthony
faustusnotes,
Adjustments are not the problem. The problem is the incessant ‘adjustments’ made to the temperature record with no explanation of how they were arrived at, what the methodology was, or what the original raw station temperatures were. And the ‘adjustments’ are always of two types: either lowering past temperatures in order to sho more rapid warming, or ‘adjusting’ current temperatures upward. You don’t seem to have a problem with those shenanigans. Why not?
There is an enormous amount of money at stake. Without complete transparency of all data, metadata, methods and methodologies, the presumption is that they are lying for money. The simple fix is to open the books: post everything on line. Instead, they kep it hidden from the public that paid for it. Why? The reason is clear. They are trying to alarm the public, because that brings in the money.
My second comment just got eaten, so I’m going to post it again. You have put up two private emails between scientists (which is in itself a pretty shameful act) in which one calls Goodridge a jerk, and you seem to think that means they’re bad people. Since I’ve posted on this thread I’ve been called a Nazi, accused of personal attacks, called a hater (by you), stupid, accused of getting “uppity” and “worked up” (by you), accused of “nit-picking” (ironic in light of the emails you’ve published), accused of being incapable of understanding what others write, and you’ve had to censor one comment for saying bad things about me. Also, even though it’s clear that I don’t want my identity revealed, you’ve tried to do so and have even published my place of work, in a very hostile forum.
In light of that, don’t you think that your complaint about what two scientists said to each other in a private forum is a bit … weak? And do you think you should perhaps be reminding your cohorts of how they should behave in polite debate? Because the behavior in this thread suggests to me that at least a third of your commenters are just as bad as, or worse than, the scientists in the quoted emails. I think you and most of your commenters should think about your manners a little.
REPLY: You are quite mistaken, and quite wrong. There are no comments in this thread calling you a “Nazi” The only existences of the word are in this comment, and the one that got “eaten” recovered above from the SPAM filter which catches such phrases. You seem to think that is some sort of conspiracy against you, when it is just simple spam protocol.
I asked questions because I like to know who is challenging me. “Know your opponent” is my motto. You seem to think you are the only one allowed to ask tough questions. For all I know you could be another person trying to make an exploit/hack as has been done at other skeptic blogs this week. I’m well within my rights to ask those questions. But you are free to hide behind the comfort on anonymity like so many anonymous cowards who frequent here and hurl from behind the curtain.
Mostly what I’ve learned about you is that you are a complete waste of time, since you ignore the larger picture to focus on attack points, and concede nothing. To that end, I won’t waste any more with you. But I will post up Jim Goodridge’s response to your one question.
Be as upset as you wish, and thank you for your consideration. Cheers! – Anthony
Agenda 21 1992 (UN) Earth Summit, “there has to be a way found to take the cash from the rich and give to the poor” our duty to impoverish the rich countries..
179 nations including USA signed up, probably none of them knew what they were doing. But how do you do it? you pick something like CO2, make it a bad guy, and tax the hell out of everything that produces it. AGW is working for them for now, but the truth is emerging.
Bet you have an ILCEA rep in your back yard.
Should be ICLEA (apologies, typo)
faustusnotes says:
August 17, 2012 at 12:52 am
Pat Frank, that’s incorrect. Goodridge’s criterion for site choice is stated clearly in page 2 of the paper, I quoted it above: if they have a temperature trend greater than 0.0125 per year, they are defined as “urban” and if it is less than that it is “rural.” He did not define them directly in terms of whether they are actually urban and rural.
I suggest you read what Pat Frank actually said.
You are correct – the initial discrimination was as stated on page 2 based on the temperature changes. Well done – GOLD STAR – don’t hold it too tightly though.
Subsequently, this arbitrary split of sites into urban and rural was VALIDATED. First by local knowledge (something that you do not have). Secondly, by use of SST data and showing that these known rural sites correlated with SST data whereas the non-rural sites do not. Finally by citing a peer reviewed paper from Canada. So you can give that gold star back as you have disregarded the validation.
Note also as you were told many times – this was a presentation to other people interested in climatology, when it was somewhat arcane and a poor-man’s hobby and Croesus like riches were not available.
As these things are so important – I find it totally unforgivable that none of these overfunded agencies have put in place any configuration management. Their quality control only improved after Anthony’s Surface Station work stung them – and I think even then it would have been ignored except someone saw it as a funding opportunity to get some nice shiny new technology.
There are not that many surface stations in the world. Each site should have its own record with complete description and status. Any changes to the reported observations from the site should not be glossed over as ‘homogenization’ or ‘adjustments’ by algorithms implemented by people who have never left the computer lab. Instead each change of observation should be separately recorded with the reason for doing so and signed off by the person doing it and their supervisor. After all these were formally recorded observations – what right has a someone in a remote computer lab several decades later to say they were incorrect or should match another observation site – also never visited – up to a thousand kilometers away?
The lack of governance and configuration management is startling. One can only surmise that it is intentional.
Pat Frank says:
JJ, in terms of the rural/urban hypothesis is exactly how the data were analyzed.
No, they weren’t. To test the hypothesis that dividing the sites between high and low trend will correlate with urban vs rural siting, one would arbitrarily divide the sites by trend (check!), investigate each site and rank against objective siting quality criteria (not so much), and report the analysis which quantitatively demonstrates the difference in siting quality that was hypothesized (not at all). Only when that was demonstrated would one then proceed on to compare the full trends and make conclusions about what the difference means (mostly what this paper did).
I’m sorry, but saying “viewing most of the sites” and finding the characterization “fairly close to reality” is not a sufficient quantification of the results to test the hypothesis, and the talking around of the obvious busts in the characterization that Goodridge notes (urban in rural and vice versa) may be on target, or it may be ad hoc and special pleading. There is no way to discern from what was done.
I’m not saying that Goodridge’s claims aren’t valid. I rather suspect they are. They just aren’t demonstrated by what he did. And that is the “science” part. There is value here, and he should get props for doing what he was able to do on his own dime (Hansen produces absolute shit masquerading as science, and he has literally billions at his disposal) but it isn’t complete.
Your suggestion of tautology is explicitly refuted by the comparison of rural/urban trends with SST. .
Nope. Neither does the comparison to someone elses rural/urban trends. Just two more ways of “finding” the low trend that was the method for dividing the sites in the first place. If your hypothesis is that dividing the sites by trend will effectively divide them into urban/rural siting classes, you can’t demonstrate the hypothesis by looking at trend. You have to look at siting.
You and faustusnotes plainly have no case.
On that particular point, we very much do. On other matters, it is not helpful to lump us together, as we are quite far apart …
Anthony, thanks for bringing this paper to the fore. I remember some years ago when I first saw the rural vs. population densities wrt temperature trends here at WUWT. It still stands as good empirical work.
Just as a stopped clock is still correct twice a day, one of the websites referred to by Anthony (http://julesandjames.blogspot.ca/) has many beautiful “nature” photos from Japan, including one of Mt. Fuji that is quite unlike any others that I have seen and which looks more like a woodblock print than a photograph.
IanM
I just took a quick look at Mr. Goodridge’s paper. His reasoning is not immediately clear to me. If, as Faustusnotes suggests, he simply grouped rising-temperature sites in one pile (A) and falling-temperature sites in another (B), and then defined the former as ‘urban’ and the latter as ‘rural’, then the UHI conclusion would indeed be tautologous, ‘by definition’.
However, if he first defined the groups A and B by temperature trends, and then looked for an explanation for the difference, and after observation proposed that UHI accounted for most of the station values, that’s a perfectly valid method.
Since Mr. Goodridge, according to Anthony, had actually visited all of the stations in his sample, he would have easily been able to validate the hypothesis that the difference between the two groups A and B was caused by UHI. He would have also been able to suggest alternative explanations for stations that did not fit the hypothesis, e.g. in San Francisco.
My guess is that this, rather than simple-minded tautologizing, was indeed his method. The lesson, if there is one, is that, besides transparency of data and analysis, one needs to clearly show the steps and progress of one’s scientific reasoning.
/Mr Lynn
REPLY: I’ll simply ask him and post his response. Note Goodridge’s note about San Francisco. Clearly he is aware of that station, and as I’ve said he has visited most stations in California under his office. He doesn’t like to get involved in comments (like the majority of readers here), due to the trolling angry people that sometimes inhabit them, but he will respond to an email.
Anthony
I laughed when I saw the bathroom scales analogy. Calibrating the scales and then weighing myself naked is ridiculous. I like to weigh myself with all my clothes on and then make an ‘adjustment’ to reflect what I think I should weigh. :o)
What the Goodridge paper demonstrates is that papers like BEST that say there is no UHI effect have likely missed something.
Papers like BEST that find no UHI make a classic mathematical error. They are comparing apples to oranges. They compare the rate of change in temperature with absolute population levels and conclude there is no UHI.
What they fail to compare is the rate of change in temperature with the rate of change in population. It is the change in population that gives rise to the UHI. People moving into and out of a region change the temperature. When population is static, temperatures are static. When population changes, temperatures change.
California is an obvious place to demonstrate this, because of the large change in population over the past century. The Goodridge paper does not prove the UHI. What it does show is that that there is something going on that is related to population that has been missed by much of climate science.
The BEST fallacy can be easily demonstrated with you car. Find a spot of level ground. Hold the gas pedal steady. The vehicle speed will remain steady, regardless of whether the gas pedal is pressed a little (rural) or a lot (urban). From this BEST concludes there is no correlation between the gas pedal and your vehicle speed.
The speed of a vehicle on level ground only changes when the gas pedal changes. So,if you are looking for the UHI (change in vehicle speed) you need to look for change in population (change in gas pedal location).
BEST, Jones, they missed the UHI because they didn’t measure the right thing. They looked in the wrong place, and from this they concluded there was no UHI. A few days ago I lost my car keys. I looked for them in the wrong place and didn’t find them. From this I concluded that my keys did not exist. The IPCC used this very same logic to conclude CO2 causes warming.
More than changes in albedo, the changes in relativee humidity in comparisons between rural agricultural and rural non-agricultural I would expect to be substantially more influential. The forested areas can be quite dry in comparison to farmland under irrigation and direct insolation. Water vapor being a strong influence, the changes in time periods with irrigation versus without irrigation should show up quite strongly in the temperature and humidity observations. The urban aeas should also exhibit some of these changes as well with the inceases and descreases in the irrigatoin of urban landscaping and golf courses. A subset study of stations located in proximity to the golf courses past and present could be interesting.
Goodridge has shown that there are two different types (populations) of weather stations in California. One set synchronizes with SST, the other shows a rising trend.
That information we all agree on. At that point the REASON for the two distinct types of temperature trends have to be determined AND accounted for in the math or all the homoginizing, griding and adjusting done by the “Team” is just so much garbage.
Applying any type of mathmatics/statistics to a bunch of numbers shown to be from two different populations is an absolute no-no in statistics and that is what has been done. Goodridge’s little study invalidates ALL of the current temperature data sets except for the satellite observations from UAH and RSS. ESPECIALLY WHEN, despite FOIA requests and even lawsuits the methods for adjustments remain hidden. link and link and link
Here is the NASA GISS temperature graph (2008) and here is the graph of GISS from 1980 vs from 2010 (Satellite observation for 2009 marked with * )
and here is the satellite UAH Global Temperature
Discussion about the divergence on WUWT and here
At this point I really do not care if it is a “Rural” vs “Urban” issue or not what matters is there IS AN ISSUE and so far I have seen nothing that shows the issue has been addressed except for the tossing out of stations causing an increase in the % of stations at airports.
Any trust I had in Climastrologists when out the window after reviewing their abuse of temperature data.
And just for yucks here is an article on the error in the temperature measurements.
You are on the right track, but need to expand your treatment of the problem a little bit.
The UHI effect is very, very strongly dependent on the first steps of the total change from a “pure rural” environment (dirt roads, single family farms randomly scattered about with less than 8000-14,000 people per county, no air conditioning and central heating and almost no manufacturing) into a massive urban setting (massive regional energy use, large reflective changes with concrete and roads and buildings, large wide-spread populations using central air conditioning and heating, commercial energy and transportation use.)
We measure and talk about temperature changes every day of 2 deg C to 4 degree C between “suburbs” and “the city” around every city in the country and worldwide. Therefore, to pretend UHI doesn’t exist – that the change in pristine, rural temperature trends over time is “proved” to the same as the change in urban temperature trends over time is foolish. Dead wrong.
But … WHEN did the temperature change we recognize as UHI occur? (More accurately, when did the majority of the temperature change occur, since it is an on-going effect as urban industries and regions change, as suburbs grow, and as “rural” areas change from grasslands and hardwood forests to farmlands to pine forests back to irrigated farmlands and as farmland use changes (no irrigation, plowing, spray irrigation, flood irrigation, leafy to grassy crops, etc.)
The UHI “signal” (the correction actually for the total UHI effect over 200 years of measurements) will be inversely logarithmic with respect to population, building+road area, and energy use in the region. That is, a change from 15000 people in a 100 km x 100 km “county” in 1900 to 35,000 in that same “county” in 1920, and then from that 35,000 to 150,000 in 1980 will be much more effective in changing measured temperature from 1920 to 1980 than a change from 1,500,000 to 2,500,000 in the twenty miles around Central Park between 1940 and 1950.
The UHI effect saturates quickly, and that saturation effect in most areas of the US was achieved rapidly in the cities. The change from what used to be remote “rural” to close-to-city-UHI’s for previously “suburban” temperature stations is NOT being accounted for in today’s “models” needed by the government to prove their CAGW extrapolations.
faustusnotes says:
Even if he had, Goodridge is a climatologist, not a geographer, and typically geographers or demographers define whether a location is “rural” or “urban”
Uh, not for climatological purposes. This sort of silliness is why the “urban/rural” designation is initially intuitive but ultimately obstructive to the goal. A proper examination of non-climatological effects on temperature records will undoubtedly turn up an “urban/rural” component, the full effect of which won’t be seen by looking at just that, due to confounding. The proper proceedure would be to classify site quality (as Anthony has recently done elsewhere – your attempt to slur that by association was despicable) and watch “urban” and “rural” fall out of the results.
This won’t get at notions of causality though. It may just be that the biggest settlements are more likely to be placed in areas that are vulnerable to warming.
According to “climate science”, that is not possible.
Incidentally, if Goodridge has found an urban heat island effect, then the only solution for handling historical temperatures, since we can’t measure them again, is an adjustment.
Only? How quickly you toss out the option of tossing out the data. That is the only proper way of ‘handling’ data that are irretrievably corrupted. Yes, it is a question to determine whether or not the data are permanently buggered, but it is a question that must be asked before it is answered …
faustusnotes: “My understanding of AGW science (admittedly limited) is that SST is predicted to warm more slowly than land temperatures.”
I’ll give you points for your stamina at dissembling in the face of adversity. I’ll simply reiterate the long and short of what’s in Goodrich and what you’ve finally been drug to kicking and screaming.
Based on a wholly arbitrary partition of the sets — without statement as to the how, why, what of that partition — we have one set correlate with SST and the other with population growth / total county population. And this, you finally, have the pained honesty to admit.
There are only two statements that are valid or invalid at this time. And only two for you to refute properly or be drug into sanity against your will:
(1) If this is prone to the Texas Sharpshooter fallacy — and that’s a legitimate concern — then so to the rest of the notion of ‘Global Average Temperature and Trends’. Specifically and especially if you start putting your thumb on the scale to ‘adjust’ and ‘correct’ for possible biome, human, instrumentation differences, or other nonsense.
(2) This warrants further research. The paper itself, that you’re screaming about, did nothing but sift existing data and look for various correlations. They were found, and without the help of a massive mathematical religion spewing tablets from a supercomputer. If this is seen as disconfirming or disconcerting to one religion or another? Suck it up. Duke it out like men that fancy pocket protectors and replicate or fail to replicate in differing data sets.
Kev-in-UK says:
I partly agree – but I don’t see that the grouping of high and low trends seperately is necessarily completely invalid.
Neither do I. The invalid part is what came next.
Group by “urban/rural”, look at trend – AOK.
Group by trend, look at “urban/rural” – AOK.
Group by trend, call trend-based groups “urban/rural”, look at trend – not so much.
The grouping may seem arbitrary, for his analysis, but it is not necessarily ‘wrong’ and presumably his use of his own staion siting knowledge would have helped that process?
Had his own station siting knowledge been used rigorously, instead of as a hand waving exercise, you bet.
What we have here is a fellow saying “I have a hunch”, and quantifying that hunch. Essentially, he is saying, “If I am right about this, the effect might be this big.” Left out is the part where he demonstrates he is right. There is some value in quantifying the potential effects of an untested hypothesis (I share his hunch), but that isn’t a hypothesis test.
JJ says: “…I’m not saying that Goodridge’s claims aren’t valid. I rather suspect they are. They just aren’t demonstrated by what he did. And that is the “science” part. There is value here, and he should get props for doing what he was able to do on his own dime … but it isn’t complete.”
I agree 100%. I suspect Goodridge’s conclusion is qualitatively correct, but his methodology was less than rigorous.
BTW, did you follow the Scribd link to Kim’s response to Goodridge? It’s unbelievably silly:
“…Further, considering the small area of significant urbanization compared to the total global area, local urbanization will not seriously contribute to the warming in
the global average temperature.” –Kwang-Y Kim
Except that thermometers within cities are used to characterize the temperature of surrounding areas, which are not small.
It is interesting that the two data sources with the longest period of a zero slope are ones that are not influenced by UHI.
Sea surface temperatures: since January 1997 or 15 years, 7 months (goes to July)
RSS: since December 1996 or 15 years, 8 months (goes to July)
RSS is 188/204 or 92.2% of the way to Santer’s 17 years.
See: http://www.woodfortrees.org/plot/rss/from:1996.9/plot/rss/from:1996.9/trend/plot/hadsst2gl/from:1997/plot/hadsst2gl/from:1997/trend
(P.S. WFT only goes to March with Hasst2, but the last four months are known and will not make the slope positive. With the sea surface anomaly for July at 0.386, the average for the first seven months of the year is (0.203 + 0.230 + 0.241 + 0.292 + 0.339 + 0.351 + 0.386)/7 = 0.292. This would rank it 11th.)
Goodridge 1987
This date was before the Catastrophic Global Warming scare hit the fan so to speak where activists would try to convince the world that 1 C per century would destroy all mankind.
Looks like Goodridge just did a reasonable overview at the time to see if the data (current at the time) was showing any kind of trend by checking out the reliability of the data. Kudos to him. He checked and documented available data and its source and reported his finding. How was he to know the upcoming madness of 1C per century would be purported to destroy Earth (CAGW).
Now that it has been presented on WUWT some are presenting arguments that would be more suitable to a failed butt tuck (e.g. “it’s not what I want it to be”).
Butt that’s the way it is. 😉
Anthony, there is a clear comment up above – that you yourself snipped – that contains the phrase “National Socialist Warmist” – aimed at me. You do understand what that means, right? And to that now you’ve added “coward.” You’ve insulted me three times in this thread, having put up a complaint about rudeness in private emails by scientists. What do you think that is telling everyone?
REPLY: You want precise answers from me, yet seem to think that when you use the phrase “Nazi” and it doesn’t appear in this thread, that somehow that is precise enough for you. LOL! Be as upset as you wish, but I’m just not going to play your games anymore. It isn’t possible to insult anonymous cowards, or to libel them, because there’s no damage to a named person. So I’m just not concerned. – Anthony
Someone please hand faustusnotes a hanky. He needs something for his sniveling.
Smokey, I’m not sniveling. I’m pointing out the hypocrisy of complaining about climate scientists calling others a jerk, in a thread where you call someone a coward and a hater, and your mates call him a national socialist and an idiot.
It’s especially hypocritical when the insults in the thread are exactly the same as those in the emails (right down to specific words: “nit-picking.”) It’s real shoe-on-the-other-foot stuff, this is: you are laughing at climate scientists for getting “uppity” when someone “nit-picks” their work, but then when someone does exactly the same thing to an extremely bodgy analysis you like, you get really angry at them for it and use lots of rude words.
Doesn’t reflect well on the commenters …
Someone pass faustusnotes a hanky.☺
@ Gail. “At that point the REASON for the two distinct types of temperature trends have to be determined AND accounted for in the math ”
But the GUSH and HADCRUD do account for it.. They assume all the ones with no trend are wrong, and “adjust” them so the match the trend of the ones that do have a large trend (ie the Urban sites).. See, now nice and homogeneous.
Nice to see a few more people stating that it is necessary to look at the CHANGES that have occured at a site. The 1970’s to 2000’s was generally a period of rapid urbanisation. Rapidly expanding urban settings even in so called rural areas. ANY site that was swallowed up by this urban expansion MUST experience a UHI effect, and without know the complete history of developement even within a couple of kilometers of the site, there is no way that you can account for possible urban warming effects.
And when I see BEST and others saying that there is no apparent UHI effect, then all I can assume is that they haven’t been competent enough to figure out how to find it, or have deliberately avoided finding it, to further their agenda and/or to keep the funding coming.
faustusnotes says:
August 17, 2012 at 6:43 pm
It wouldn’t be nit-picking, if we we talking about a peer reviewed paper – but we aren’t – this is more about some guys curiosity and tentative enquiring and playing with data. The nit picking misses the interesting point of the paper, which is that UHI (and poor station siting) was beginning to be readily apparent in the data back in the early 80’s. And yet, despite this (or because of?) the keepers of the data (Hansen, etc) used it as a political tool to further their agendas/careers.
Now, I don’t suppose you have a magic wand to reverse time and I don’t know if you have any science background – but if you could return to that period, review this work and initiate a ‘proper’ and robust measuring and recording system, having seen the flaws in the system – we would have now had a few decades of real, genunie reliable raw data – which would actually be of use. Instead we have a seriously fecked up and adjusted suite of datasets with very little understanding of what/why/where/how it’s all been put together. This data is now being spoon fed to us – via computer models – to predict catastrophes which aren’t happening!
Everytime someone looks at a particular station (esp. rural ones) it seems, more often than not, to demonstrate that all adjustments result in an upward trend. Most sceptics simply want to know why – when good data seemingly does not show such a trend. But anyway, if you want to ‘believe’ the current data and models, when clearly they are suspect – thats up to you!
I have a computer model/program to play blackjack – trust me, it’s very robust – would you like to place a bet and I’ll tell you if you win? You won’t/can’t see the cards being played, but I’ll be honest and tell you if you win…..Doesn’t this strike you as something only stupid folk would do? But, if you think about it – at a basic level – this is exactly what the climate boys have been doing for the past 30 years! – and you wonder why they (or their silly muppet supporters) aren’t trusted???
JJ, I suspect we’re approaching a rapprochement.
You wrote, “To test the hypothesis that dividing the sites between high and low trend will correlate with urban vs rural siting, one would arbitrarily divide the sites by trend (check!), investigate each site and rank against objective siting quality criteria (not so much), and report the analysis which quantitatively demonstrates the difference in siting quality that was hypothesized (not at all).”
We agree on your first point. Your second, “not so much” we almost agree, except that I’m willing to concede JG’s professional judgment.
For your “not at all,” his Table 1 lists all 74 stations rural and urban stations. Table 2 again lists the stations, with the urban ones bolded, and gives location (lat., long., elev.) and regression slope. Three maps are given allowing correlation of site with population (Table 3). His two final figures demonstrate poorly sited stations.
Let me go a little further: Roger Pielke Sr. published a paper in 2002, with eight co-authors [“Problems in Evaluating Regional and Local Trends in Temperature: An Example from Eastern Colorado, USA” Int. J. Climatol. 22: 421–434 (2002)] discussing siting problems at 11 surface stations.
That paper showed zero site photographs. Table 1 was decadal populations (1910-1990) of the 11 station locales, like JG Table 3.
Table 2 was site description – lat., long., elev. years of data, and a short qualitative description of the area; like JG’s Table 2, though JG gave no description.
Here’s how Pielke Sr., et al., describe their sites: “The 11 sites selected are listed in Table I and their locations are shown in Figure 2. The station elevations ranged from 1033 to 1638 m (Table II). Nine sites were predominantly rural, with surrounding areas of rangeland, pastures, or cropland, and with historically low populations less than 10 000. Fort Collins was the only major urban site, with a population close to 100 000. Five sites are in northeastern Colorado, and six sites are found in the Arkansas River basin in southeastern Colorado.”
That constitutes their entire description of sites, station quality, and criteria of choice.
Figure 2 is a topographical projection of eastern Colorado, with the sites plotted on it. It conveys no information about site quality.
How is JG’s paper of lower quality than that peer-reviewed paper?
You wrote, “I’m sorry, but saying “viewing most of the sites” and finding the characterization “fairly close to reality” is not a sufficient quantification of the results to test the hypothesis…”
Agreed, which is why I noted that we have to surmise JG’s professional acuity in making his judgment. Recall that Philip Mullholland pointed out that JG’s manuscript had been cited by Tom Karl, et al., in their 1986 paper on “Urban Warming.” The citation was to 1985 “Extended Abstracts, Third Conf. on Climate Variations and Symp. on Contemporary Climate: 1850-2100, Los Angeles, Amer. Meteor. Soc. 158-159.”
So, JG’s manuscript is an extended abstract of the talk he gave at a professional symposium. We don’t know what he might have mentioned in his talk concerning the content of his judgments of station quality. However, one can’t expect an extended abstract to contain the sort of details one expects in a full paper. One also doesn’t expect hand-waving arguments to fly at a Society-sponsored scientific meeting.
Hence my willingness to concede JG’s expertise. I agree that details are missing. However, in this case I do not agree that implies a likelihood of sloppiness or bad science.
I wrote, “Your suggestion of tautology is explicitly refuted by the comparison of rural/urban trends with SST.”
And you replied, “Nope. Neither does the comparison to someone elses rural/urban trends. Just two more ways of “finding” the low trend that was the method for dividing the sites in the first place.”
You are denying the independence of SSTs. There is no reason to think that SSTs would correlate with any arbitrarily divided set of temperature trends. There is plenty of reason to think that SSTs will correlate with unbiased local land-surface temperatures. The correlation of SSTs with one trend, but not the other, validates the rural/urban distinction, in that the SST correlation implies stations that are free from urban heating. Such stations were found to be JG’s rural-designated stations. That finding was not imposed by his prior choices.
The same conclusion applies to the similar rural trends found by others. Published rural trends are independent of JG’s choices. There is no reason to think that known rural temperature trends will correlate with any arbitrarily chosen set of station trends.Their correspondence was found later, a posteriori. That makes the correspondence a valid corroboration of JG’s rural/urban distinction.
You wrote, “If your hypothesis is that dividing the sites by trend will effectively divide them into urban/rural siting classes, you can’t demonstrate the hypothesis by looking at trend. You have to look at siting.
The known-to-be rural trends (i.e., Diaz) have had their siting verified. Their correspondence with JGs trends implies the correctness of his distinction. Let’s note that JG wrote that he had inspected the sites. He just did not give any specific details. We can both regret that. He did provides maps, locations, and populations, however.
My association of you and faustusnotes was meant to refer only to the tautology claim. Sorry to be unclear on that.
for those of you that have your skivvies in a knot about the “incinerator” in picture two, that might not be an incinerator sitting beside the thermometer in the cute little house on the stand. in the long gone day and age when outdoorsmen were really into fixing their own provender there was a practice of preserving meat and fish known as “smoking”. guys would catch lots of fish, kill a deer or two and maybe an elk and place the prepared meat inside a structure that would expose the meat, or fish, to the smoke of a low fire for several weeks. this thing has the look of one of those smokers.
another reason to doubt that it is an incinerator is the lack of tiny bits of white paper that congregate in the area of the loading door/gate to the thing and on the ground around said door. also where are the foot prints and other evidences of ground traffic at the door.
also why is the thermometer up on a stand where the attendant has to climb a ladder to get to it. as i remember those “weather stations” were about chest height rather than what, eight feet off of the ground.
however:
if this station is in an area of extreme snow fall (as in donner summit where the houses had covered walkways for the railroad crews walked back and forth between the turntable/roundhouse and depo) then i might be fulla !@#$.
C
pk says:
August 18, 2012 at 12:01 pm
You just smoked your own skivvies.
Where there’s smoke there’s fire. The rust would suggest fire with breather holes at bottom. I see two sides of the incinerator. Do you see the other two? Knotted skivvies much?
nahhh just hoof in mouth desease.
C
faustusnotes, let me reiterate: JG implicitly but obviously presented the slope-based division of stations as a hypothesis.
Implicitly because he did not state so outright. Obviously because immediately (pp 2-3) after making the distinction in terms of slope he noted the non-intuitive distinctions of the San Francisco and Mt. Hamilton stations, which were automatically but correctly divided by his classification. JG then went on to note that his division of sites by slope produced lists that corresponded to what he personally and professionally knew about the site quality of the stations.
You’ve continually missed that point. And that point alone is enough to refute your claim of non-objectivity.
You also wrote that you’re “accusing him [JG] of using a poor criterion for identifying urban vs. rural weather stations,” except that you don’t know what his criteria were. His criteria were not restricted to slope, because he immediately tested his slope-based distinction against whether the stations were, in fact, rural or not.
If he’d found that larger and lesser slopes did not correlate at all with his knowledge of the rural/urban siting of the stations, he’d have realized his distinction was erroneous and could not have proceeded with the analysis.
His base criterion, therefore, was obviously whether the siting of rural/urban corresponded to his distinction by slope.
Obviously JG didn’t specify how he made his rural/urban judgments. However, I’ve already dealt with that here, pointing out that he must have brought his professional acuity to bear.
You disputed his acuity, suggesting without evidence that as a “climatologist, not a geographer” JG is somehow unable to discern a rural from urban site. But then his two pictures display a detailed appreciation of artifactual circumstances. This vitiates your profession-based objection. JG is clearly able to distinguish site quality. Most observers would surmise that he possessed this capacity, as that sort of distinction is basic meteorological practice.
I was glad to see you write that, Correlation does not equal causation. 🙂 We definitely agree on that point. The correlation between the 20th century rise of CO2 and the rise in surface air temperature, for example, says nothing whatever about a causal relationship between them.
faustusnotes, the correlation of land temperatures with SST has nothing to do with a claim of AGW. The very high heat capacity of the oceans means that SSTs generally govern air temperatures.
However, land temperatures rise and fall to far greater extremes than SSTs because of the low heat capacity of air, and also, often, because of the albedo of the land surface.
JG’s comparison between SST and unbiased rural land air temperatures is very standard in climatology. See James Hurrell’s 1996 paper discussing exactly these correspondences between the Atlantic and European temperatures, and the Pacific and North American temperatures. Here is another paper — 2001, Bonsal, et al. — showing a century’s worth of influence by ENSO, PDO, and NAO oscillations on temperature trends across Canada.
Such long range influences are called teleconnections — not to be confused with the control of global temperatures by Michael Mann’s California bristlecone pines.
You wrote, “Goodridge has chosen to call the slow-warming stations “rural” but remember, that’s not how he selected them.” But, of course, that is how he validated them.
You wrote, “Incidentally, correlation coefficients are not a good way to compare time series. … That Goodridge didn’t do this is a bit of an oops, but Box and Ljung had only written their work about 10 years previously (I think), and it hadn’t filtered through to a lot of people, so I suppose he can be forgiven.”
Nice try, but JG didn’t correlate any temperature series. He merely showed the visual similarity of their trends. With that comment, you’re beginning to stoop a little, faustusnotes.
You also wrote, “As I understand it (I could be wrong), Mann made no such mistakes in his hockey stick work.
Michael Mann merely published work he knew was a statistical farce, using a method he knew mined for hockey sticks. Was that a mistake?
You wrote, “Incidentally, if Goodridge has found an urban heat island effect, then the only solution for handling historical temperatures, since we can’t measure them again, is an adjustment.”
Not correct. The only way to handle a prior urban heat island effect of unknown (and unknowable) magnitude is to estimate an uncertainty and propagate that into any temperature trend. No climate scientists apply that any of the time.
Pat Frank says:
JJ, I suspect we’re approaching a rapprochement.
There is a convergence.
We agree on your first point.
There it is. 🙂
Your second, “not so much” we almost agree, except that I’m willing to concede JG’s professional judgment.
Most people around here say “In God we Trust, all others show data.” Being Agnostic, I’m a notch tougher than that. 🙂
For your “not at all,” his Table 1 lists all 74 stations rural and urban stations. Table 2 again lists the stations, with the urban ones bolded, and gives location (lat., long., elev.) and regression slope.
None of which speak to urban/rural or the associated siting quality issues.
Three maps are given allowing correlation of site with population (Table 3).
They may allow correlation of site with population, but that wasn’t done. Sure, if a reader wanted to, they could take the station location data, and investigate each for surrounding population and other siting quality issues. Then they could run some stats and determine if the sites assumed to be “rural” based on their temp trend were in fact rural. They could do the same thing for the sites assumed to be “urban”, and they could publish those results, completing Goodridge’s paper for him. He could have done that, too. But he didn’t, and that is my point.
His two final figures demonstrate poorly sited stations.
Which is interesting in its own right, but those and the other examples given in the text are provided to demonstrate that those sites don’t fit the “urban/rural” classification assigned to them by the arbitrary trend break. What other sites also don’t fit? We don’t know.
Let me go a little further: Roger Pielke Sr. published a paper in 2002, … How is JG’s paper of lower quality than that peer-reviewed paper?
Leaving aside for the moment the logical fallacy of exonneration by association … a few things:
1) As you point out: “Table 2 was site description – lat., long., elev. years of data, and a short qualitative description of the area; like JG’s Table 2, though JG gave no description.”
How is JG’s paper of lower quality, with respect to urban/rural assessment of each site? Well, based on what you said, Pielke gave a site assessment for each site in his study, and JG didn’t . And what you (for some inexplicable reason) didn’t say is more damning still. Pielke’s Table II includes the current population for each site. Oh, and Pielke’s Table I gives the population history for each site, for the entire 20th century. Shame on you Pat.
2) Pielke’s paper is not about “urban/rural” differences in trend. It is about regionality and to a lessor extent seasonality. Pielke’s paper does not compare “urban/rural” groups of stations – or any grouping of stations, for that matter. His paper’s thesis is that individual sites are not representative of regional trends, so he compares all of the sites in a region against each other. Population isn’t even mentioned in Pielke’s abstract, and only appears in the text as an ancillary speculation on a couple of odd observations. It has no bearing on his results.
Soooooooooo … Pielke et al did a better job of documenting a factor (“urban/rural”) that has no impact on his results than did JG, who didn’t provide any documentation for the factor that was the basis of his thesis.
3) Pielke’s reasoning is not tautological, JG’s is. Comparing individual sites to each other and finding significant differences, and finding significant differences between individual sites and the average of all the sites in a region, is interesting. Finding that the trend of sites chosen because they have low trends is lower than the trend of sites chosen because they have higher trends is …
Recall that Philip Mullholland pointed out that JG’s manuscript had been cited by Tom Karl, et al., in their 1986 paper on “Urban Warming.”
So? Appealing to authority is a logical fallacy. Appealing to Tom Karl’s judgement is, well … not something that I would make a habit of.
So, JG’s manuscript is an extended abstract of the talk he gave at a professional symposium. We don’t know what he might have mentioned in his talk concerning the content of his judgments of station quality. However, one can’t expect an extended abstract to contain the sort of details one expects in a full paper.
If JG were truly taking the tack that you are relying on – that dividing stations on trend and looking for expected differences in “ruralness” was the hypothesis being tested, then the quantification of those differences in “ruralness” would have been the results of the paper. One does expect a minimalist abstract to present the results, let alone an extended abstract. Kind of goes to the definition of an abstract.
One also doesn’t expect hand-waving arguments to fly at a Society-sponsored scientific meeting.
Well, I have seen all sorts of nonsense flying at scientific conferences, expected or not. Too, we don’t know if what was presented at the Society sponsored meeting did fly. It may have been panned. Either way, it says nothing about what was presented. The only thing we have before us that descrbes the content of the talk is this paper, and this paper is lacking.
You are denying the independence of SSTs.
No, I am denying the independence of the method of dividing stations by trend from the “finding” that the trends of the stations so divided differ according to the method of their division.
There is no reason to think that SSTs would correlate with any arbitrarily divided set of temperature trends.
Sure there is. Don’t be silly. Given the heat capacity of the ocean, SST trends are expected to be more or less flat, regardless of your position on ‘global warming’. Pick a group of *anything* based on the flatness of its trend, and you would expect it to correlate with SSTs. And you won’t be wrong.
There is plenty of reason to think that SSTs will correlate with unbiased local land-surface temperatures.
You are assuming “unbiased”, but you aren’t grouping based on documented bias. You are grouping on trend and calling the group based on the trend that you would expect of an unbiased group “unbiased”, and calling the group based on the trend that you would expect of a biased group “biased”. Upon examining what you have done, you then “find” that the high trend group has a higher trend than the lower trend group, and you pretend that this says something about bias. Sorry. No. You’re just regurgitating your assumptions.
The correlation of SSTs with one trend, but not the other, validates the rural/urban distinction,
No. That line of reasoning is a form of confirmation bias.
“… in that the SST correlation implies stations that are free from urban heating.
No, the SST correlation confirms stations that are free from *any* significant net heating, “urban” or otherwise. Given that the stations in question were chosen because they didn’t demonstrate net heating, this is not surprising. It is tautological.
Such stations were found to be JG’s rural-designated stations. That finding was not imposed by his prior choices.
Uh, yes it was.
The same conclusion applies to the similar rural trends found by others.
I concur. Comparison to other “rural” trends is similarly flawed. Essentially, you are saying that the “findings” of JG’s paper are not in JG’s paper, but that his findings are to be found in other people’s papers. Silly.
As I stated above to another poster, there are a couple of different ways to legitimately skin this cat:
1) Divide stations based on some measure of “ruralness”, and compare the trends between those groups for evidence of difference. In this case, differences in trend are the results.
2) Divide the stations based on a measure of their trend, and compare the “ruralness” between those groups for evidence of difference. In this case, differences in ruralness are the results.
Either of those works, but JG didn’t do either of those. And it isn’t an ala carte deal. Using the methods from 2) with the results from 1) – which is what JG actually did – is simply not valid.
These sort of circular reasoning fallacies can sometimes be difficult to detect. That really isn’t the case here. It is pretty obvious.
JG said that he divided his groups based on the magnitude of their temp trend. He says he called the low trend sites “rural” and the high trend sites “urban”. Listen to him when he says that. If you want to clearly see what he did next, just go thru his paper and everywhere you see a term like “rural site” or “rural temperature” in reference to his data, replace it with “low trend site” or “low trend temperature” and every time you see a term like “urban site” or “urban temperature” so used, replace it with “high trend site” or “high trend temperature”. So stated, the ‘results’ aren’t so compelling.
The known-to-be rural trends (i.e., Diaz) have had their siting verified. Their correspondence with JGs trends implies the correctness of his distinction.
What correspondence? JG doesn’t demonstrate any correspondence with Diaz. What known-to-be-rural trends in Diaz? JG doesn’t even make that claim. It wouldn’t matter if he had, for the reasons shown above, but he didn’t even do that. You are seeing things that aren’t there.
Let’s note that JG wrote that he had inspected the sites. He just did not give any specific details.
Lets note it correctly. JG said that he had “viewed most of the sites”, and he said that the classification of that subset of the sites was “fairly close to reality” based on his recollections of having viewed that subset. Nothing objective or quantified. “Fairly close” is not only subjective, but a very weak commitment at that. Lipstick does not a prom date make. It simply annoys the pig.
He did provides maps, locations, and populations, however.
Well, maps and locations is the same parameter, and he did not provide site populations. If he had, people might reasonably have expected him to have used those populations as his “urban/rural” criterion.
He did provide county populations, but he did not use them to determine “rural/urban” for his sites. To the contrary, his list has many instances of both urban and rural designated sites within the same county, rendering county population irrelevant to the question of “urban/rural” insofar as JG’s paper is concerned.
As it stands this paper is seriously flawed. It isn’t necessarily beyond fixing, tho. Goodridge didn’t do it right, because he didn’t have the resources to get the data needed to do it right. Anthony probably has those data, having had access to a resource that Goodridge did not: free labor sourced from the internet. It is likely that the sites referenced in JG’s paper have been visited and classified by the Surface Stations project. It would be interesting for someone to start over with JG’s “implied hypothesis”, split the sites on trend, and use Anthony’s data on siting quality to determine if there are differences … it wouldn’t be the first paper that took 25 years to finish.
Thanks for the conversation, JJ. It’s good to air out these things. I’ve just seen your 2012 August 19, 2:19 pm comments (Sunday evening), and will post a reply in a day or two.
Lazarus Add-On. Lose nothing. Ever. Save yerself scadoodles of time an’ aggervation. Promise.
D/L’d the paper; reads very clearly as a careful and competent effort by a true professional (and local expert with first-hand knowledge).
The Jerk’s Revenge! More broken glass for the Team to chew on …
JJ, putting aside where we already agree:
I wrote, “Your second, “not so much” we almost agree, except that I’m willing to concede JG’s professional judgment.”
To which you answered: “Most people around here say “In God we Trust, all others show data.” Being Agnostic, I’m a notch tougher than that. 🙂”
Regarding professional judgment, the data are Jim Goodridge’s education and years of employment as a climatologist. Here is a record of his Special Recognition Award at the 2005 California Extreme Precipitation Symposium. JG’s credentials to make a professional judgment are beyond dispute. Nevertheless you dispute them, as is your right.
The central question is whether the sites themselves are in fact urban/rural or not. You trust that Pielke Sr. et al., 2002, wrote the truth about their stations; so do I. But you will not grant the same trust that Jim Goodridge wrote the truth when he noted rural/urban correspondence with his direct inspection of “most of the sites.” Pielke Sr.’s descriptions are not data, e.g., untampered photographs. Nevertheless, you credit his statements. Therefore, it seems that your trust is inconsistently given.
Regarding the stations listed in Tables 1 and 2, you wrote, “None of which speak to urban/rural or the associated siting quality issues.”
Rural/urban is designated; you merely deny JG to have made a professionally credible judgment. As you note JG did not describe specific site quality. Pielke Sr., et al., 2002, also did not describe specific site qualities, but did describe areal features, which you see fit to not deny.
Regarding the maps and Table 3, you wrote, “They may allow correlation of site with population, but that wasn’t done.”
Except Table 3 does records County population. Pielke Sr., et al., 2002, record the population of the nearby urban area. The same effort would be required if one wished to independently verify the qualities of the sites listed in either paper.
This is true of any paper in science. One trusts that the report is accurate. Lacking trust, or if the reported data are critical, one makes the effort to repeat the observation or experiment oneself. You’re supposing that JG’s paper is uniquely bereft by requiring such verification. However, it’s not. It’s typical. Specific verification of any paper requires that sort of effort.
You wrote, “What other sites also don’t fit? We don’t know.”
We don’t possess that level of assurance about any report that we have not personally verified. What should we do about that? Should each of us re-verify science starting with Galileo’s experiments? If you trust that Pielke Sr. et al., 2002, were telling the truth about their sites, there’s no reasonable way to reject JG’s statement of direct urban/rural knowledge. Unless you can prove him incompetent.
“exonneration by association” That was a demonstration of adherence to professional practice. Science is method.
You wrote, “And what you (for some inexplicable reason) didn’t say is more damning still. Pielke’s Table II includes the current population for each site. Oh, and Pielke’s Table I gives the population history for each site, for the entire 20th century. Shame on you Pat.”
Pielke, Sr., et al., give the populations of the associated urban areas. JG gives the populations of the associated counties; less resolution for sure. On the other hand, JG gives the full temperature histories of each site. Pielke Sr, et al. do not. You didn’t mention that. Should you shame yourself?
JG Figures 13 and 14 show that the California temperature trend positively correlates with population at the urban-designated sites but is independent of the increasing rural population.
Although he didn’t mention it, JG’s Figures 12 and 13 show that the total trend in temperature and the urban trend alone, have the nearly same slope of 0.03 F/decade. The rural trend is about zero F/decade. That is, JG’s data show the entire trend in California temperature is correlated with the growth in population of his pre-designated urban sites.
Regarding SSTs you wrote, “Pick a group of *anything* based on the flatness of its trend, and you would expect it to correlate with SSTs.”
The correlation of SST with local unbiased temperatures is a deduction from physical theory. Your objection ignores the physically valid scientific context of the correlation.
You wrote, “You are assuming “unbiased”, but you aren’t grouping based on documented bias. You are grouping on trend… etc.”
JG validated his grouping by direct knowledge of the site quality. You merely deny his standing to do that. Having denied that, your subsequent claim of lack of validity in the data is merely tendentious. I.e., dependent on your presumption of JG’s lack of professional acuity.
You wrote, “You’re just regurgitating your assumptions.” Thanks for the amusing irony.
You wrote, regarding the SST rural/urban distinction, “No. That line of reasoning is a form of confirmation bias.” Not correct. The SST data are independent.
You wrote, again regarding the SST correlation, “<em.Given that the stations in question were chosen because they didn’t demonstrate net heating, this is not surprising. It is tautological.”
This is your central issue, and faustusnotes’. It rests entirely on your denial of JG’s professional standing to have observed, judged, and known the urban/rural distinction for “most” of his sites, where “most” means a large enough fraction such that JG himself was professionally confident in the results.
The tautology, therefore, is a consequence of your disbelief, and not inherent in what JG did.
The credibility of your claim of tautology, therefore, resides entirely on whether one wishes to credit your rejection of JG’s professional acuity, or not.
That makes your claim, and that of faustusnotes, indistinguishable from a polemic. It requires selective credit of your standing to deny JG’s standing.
You have no such standing here. Nor does faustusnotes. Nor do I, for that matter. However, unlike you and faustusnotes, I’m prepared to honor the evidence of JG’s professional career.
And that’s the baseline issue. There’s really not much more to be said.
You wrote, “Essentially, you are saying that the “findings” of JG’s paper are not in JG’s paper, but that his findings are to be found in other people’s papers. Silly.”
You appear unaware of the power and meaning of a corroborating result.
The rest of your comments are along the same lines. You deny JG’s professional standing to have made the urban/rural distinction. You disallow the deductive physical context that links the meaning of SSTs and unbiased rural land air temperatures. You don’t credit independent published corroboration.
That’s your position as you’ve apparently stated it. It’s without merit. You’re welcome to the last word. I sincerely doubt it will substantify your case.
Pat,
I had expected that we were breaking away from the circular reasoning of the earlier discussion. Instead, you have doubled down and introduced a brand new fallacy, attempting to prop up the tautology with an appeal to authority.
Very disappointing. Particularly so, given that such is being posted here on WUWT, which earned its place in the blogosphere when Anthony rejected the official pronouncements by “people of acuity, prestige and standing” that the siting of temperature stations was just so. Anthony then quantified the extent to which they just weren’t. Now you’d reverse the roles?
I’m going to disentangle the fallacies, and address each in turn, starting with the new one.
I am not disputing JG’s credentials, or denying his professional standing, or rejecting his professional acuity, or any of the other ways that you restate the same false ad verecundium argument. Professional standing does not substitute for data and analysis. No doubt that JG is eminently qualified to perform an objective and quantified siting assessment, if he chose to do so. He didn’t. He would likely be quite capable of quantifying the difference in siting assessment between two populations of stations so classified, if he chose to do so. He didn’t. Had he done those things, then he would have had scientific results to report and he could have rearranged the logic of his paper in the manner you suggest to avoid the tautology. He didn’t, didn’t, and didn’t.
The central question is whether the sites themselves are in fact urban/rural or not.
No, that isnt even a question. We know that the sites are not in fact urban/rural. JG tells us that. A substantive issue is that we do not have a quantification of urban/rural for each of the sites. JG tells us that, too. Nor do we have a quantification for any of the other siting issues that JG appeals to in his ad hoc justification of the instances he mentions where the urban/rural classification was incorrect.
The central issue is that, lacking those quantification data there are no results from which to test what you argue was JG’s “implied hypothesis”, and lacking that ability to test that hypothesis JG chose instead to assume that hypothesis, creating a tautology.
You trust that Pielke Sr. et al., 2002, wrote the truth about their stations; so do I. But you will not grant the same trust that Jim Goodridge wrote the truth when he noted rural/urban correspondence with his direct inspection of “most of the sites.”
Nonsense. It trust that if JG said that he visited most of the sites at some point in his long and distinguished career, then he in fact visited most of those sites. Most is not sufficient. Subjective recollection is not sufficient. Unquantified assessement based on subjective recollection of most is not even in the same room as sufficient. The problem is not that I don’t trust what JG says he did. The problem is that I do.
If we trust that JG did what he said he did, then he created a tautology. You want to rescue JG from that faux pas, by pretending he did something other than what he says he did, your “implied hypothesis”. But if we trust that JG did what he said he did (and I do), he has no results from which to derive conclusions regarding the validity of that hypothesis.
Prove this to yourself. Attempt to answer these very basic questions regarding the results that would be necessary to evaluate a test of the ‘implied hypothesis’ that you are pretending was the thesis of JG’s paper:
* Of the 74 sites that were used in this paper, how many were visited by JG?
* When it is stated that some of the 74 sites were ‘fairly realistic’ WRT the urban/rural classification that was not based on urbanity or rurality, what does ‘fairly realistic’ mean? i.e. what is the urbanity standard that JG is applying to arrive at ‘fairly realistic’?
* Of the 35 stations classed as “urban” by the trend selection method, how many were visited by JG?
* Of the 35 stations classed as “urban” by the trend selection method that were also visited, how many were actually “urban” by his urbanity standard?
* Of the 39 stations classed as “rural” by the trend selection method, how many were visited by JG?
* Of the 39 stations classed as “rural” by the trend selection method that were also visited, how many were actually “rural” by his urbanity standard?
* Of the stations that were incorrectly classified as urban or rural by the temp trend based method, how many if any does JG think should remain in the incorrect classification? For what reasons?
* Of the stations that were correctly classified as urban or rural by the temp trend based method, how many would be switched if the reasons for switching classification were applied?
No test of the “implied hypothesis” could proceed absent the answers to those very basic questions, which would constitute minimal results for such an examination. From those answers, one could begin to quantify the hypothesized differences in rurality between the two populations. From those quantities, one could apply a hypothesis test, perhaps a comparison against a chi-square distribution, for example.
But you cannot answer those questions, because those data are not included in the paper. And they weren’t included, because JG didn’t have those data. He had the (standing, acuity, credentials, prestige, glowing halo, lack of original sin, whatever) to have properly collected those data and quantified those results. But he did not.
And lacking those data, he had no results for the “implied hypothesis”, so he did not present it. Instead, he created a tautology.