A guest post by John Goetz
[Update: I cross-posted this on CA and in the process added the brief discussion of night lights as well as made some minor text changes. I have reflected all of those changes here]
In May I began a quest to better understand how GISS does its homogeneity adjustment, also known as GISS Step 2. Steve McIntyre took the ball from that scrum and ran with it, producing a set of R tools that nearly replicate the GISS method. Some of the endpoint cases continue to confound those of us trying to understand the source code and how it reconciles – or doesn’t – with the peer-reviewed literature.
As this was going on, Anthony Watts pinged me several times, asking that I look at the Cedarville, CA adjustment to better understand why GISS would apply an urban adjustment to an obviously rural station, a topic which he explored in a previous post. I hesitated, because Cedarville had a lot of “nearby” (as defined by GISS) rural stations, and I wanted something simpler to look at. However, I did not forget his request and I took occasional peeks at the station and its neighbors. At left is an overall site view of the Cedarville station. GISS assigns a “night lights” value of 2 to this station, which is what causes it to go through the homogenization process.
Here is a Google Earth image of Cedarville and the surrounding area with the NASA City Lights image overlay enabled. I am not sure what the NASA sensors are picking up to assign Cedarville the “2″ rating.
Anthony says this in his post about Cedarville: “a place with a good record and little in the way of station moves”. Generally this may be true, but I personally am suspicious about the fidelity of a station’s record when I see Batman lurking in the 1930’s:
OK, let’s assume for the moment that Cedarville’s record is beyond reproach. Let’s further assume that the Cedarville station is urban, and is cursed with the typical frailties of an urban station: lots of asphalt, little vegetation, and placement near an air-conditioner, strip mall, or jet engine. Certainly the surrounding rural stations are of such pristine fidelity that they can be used to remove the urban noise from Cedarville. Let’s take a closer look at those stations and that homogeneity adjustment.
Below is a Google Earth view of all of the rural stations within 500km of Cedarville that are used to completely determine the station’s urban homogeneity adjustment. The Oregon stations seem to be well-represented:
In the next plot I have color-coded the markers to reflect each station’s trend: white is neutral, reds are warming, blues are cooling (the darker the color, the greater the trend). Unfortunately, the red circle I used to indicate the 500km radius gives the white markers a red cast. Orleans and Electra stick out like sore thumbs with sharp cooling trends, while other stations generally exhibit a flat or warming trend.
I then went through the GISS Step 2 process of combining the rural stations into a single “rural record”. This is done by starting with the station with the longest record – Golconda – and combining the remaining stations one by one from longest to shortest record. Without going deep into the details, each station is first adjusted (biased) such that it’s mean matches the mean of the combined record. Then, the station’s record is averaged in with the combined record using a weight that decreases linearly with distance from the urban station at the center (in this case, of course, the metropolis of Cedarville).
The next plot compares the difference between the Golconda record and the final combined rural record. While Golconda has an influence on the final record, it is does not appear overwhelming.
I did notice a big difference when the fourth station, Orleans, was combined (Mina is the second and Willows the third). Comparing the difference between Orleans and the final combined rural record, I saw the slope go slightly negative but close to zero, and the extrema pull in much closer to the final value. I am not sure how to determine which station of the 29 has the greatest influence on the combined record, but my instinct is telling me Orleans is the one.
Here is a comparison of combined Golconda, Mina, Willows, and Orleans record with that of all 29 stations combined. Clearly the first four rural stations of the 29 get us very close to the final solution:
The next part of GISS step 2 takes the difference between the Cedarville record and that of the combined rural stations. Following is a plot comparing those two records:
The difference between Cedarville and the combined rurals is shown in the next plot. Also shown is the adjustment value that GISS calculates from the difference. I would have expected an adjustment that looked less like a lower envelope value and more like an average value. The adjustment result indicates all values before 1910 should be adjusted upward and all values after 1910 should be adjusted downward.
The adjustment shown above in red is then added back into the Cedarville record to produce the homogenized result. The next plot compares the homogenized version of Cedarville with the original. It clearly shows that values before 1910 are adjusted up, and values since are adjusted down.
So what do I make of all this? In the simplest terms, I see Orleans having a rather large influence on the adjustment made to Cedarville. Should Cedarville be adjusted? Well, it certainly is not urban, so the standard GISS urban adjustment seems inappropriate. But the fact that Batman lurks in the 1930s indicates to me that something is amiss and needs to be (as Delbert Grady says in The Shining) “corrected”.
Is Orleans an appropriate adjuster? Certainly it is rural, and the station history indicates it has not moved. However, when I look at the plot of the Orleans data, I see something happened around 1929, and my guess is that it was not sudden global cooling:
I don’t think this is necessarily a situation where garbage in equals garbage out. Rather, I think it is a situation in which a bunch of trimmings are thrown together and mixed to produce a kind an adjustment sausage. It is not necessarily something that accurately reflects the initial ingredients (inputs), but the output sure is tasty, especially after it has been cooked for a while.
I know this is valuable work, but it all seems to me as if we just shoveling sand around. This example might not fall in, but Anthony and others have demonstrated how badly so many of these weather stations are sited. Plus we have missing monthly records (such as these graphs demonstrate), and records that are just plain made up to fit a supposed trend. With Hansen’s stance on global warming seemingly ruling the roost at NASA, what good does much of this discussion do except to further make the point that agenda-driven science is what drives these GISS adjustments? The system in its entirety needs to be examined and fixed before a discussion of the data it provides seems of any relevance. I know this post is a sort of breakdown of the methodology of GISS and its adjustments, but based on those above points it really does not seem that even the data derived now can be trusted. I tend to agree with those who want to leave GISS out of any reasonable discussion about temperature. That whole system is flawed beyond credulity. The only point in this discussion of GISS that I can see is to demonstrate exactly how flawed it really is, though that may after all be the point to this posting.
John, I don’t understand why you’re suspicious of the “batman” warming in the 30s…
In the course of reading this it occurred to me that with every passing year, the “adjustments” to past temperatures really mean less and less. We have satellite data from 1979 till now. Next year will mark 30 years of such data. What is most important from a policy standpoint is what climate is currently doing or has done over the last 20 years or so. If CO2 is the primary driver of “climate change” that will swamp out all other impacts, we should see temperatures tracking with atmospheric CO2 levels. We don’t. These “adjustments” seem to be a lot of hokus pokus designed to make the result validate the models. It wouldn’t surprise me to learn that Dr. Hansen tried several different adjustment mechanisms until he got the desired result.
What temperatures did in the 1930’s or the 1950’s are irrelevant to current policy designed to mitigate a perceived crises that is supposedly happening right now. The trouble is that his data set is the one showing the greatist warming. Everyone seems more focused on what temperatures are now relative to 1968 than they are relative to 1998. It seems like the object is to divert attention away from the fact that temperatures haven’t warmed in the last decade by pointing to relative differences from decades ago. Nothing we can do today can change what happened between 1968 and 1998. Data since 1998 show no great crisis. As the satellite record gets longer and longer, we should rely on that more and on the ground record less. This is simply because it gives better coverage with less error from surrounding land use change.
Frankly, I think he is wasting a great deal of time and effort to make adjustments to past temperatures that are more irrelevant with every passing month.
I lived in Yreka and there are about 40,000 people in the whole county. Even worse is Modoc County, even less people. Cederville is tiny and if they put the station on asphalt it is most likely the only asphalt outside of a fast food or the only asphalt in town.
Nice job John.
One of Kristen’s volunteers did an excellent write up on the Fillnet adjustment using the Maine stations.
http://www.m4gw.com:2005/m4gw/2008/07/makin_up_climate_data_from_jun.html#more
Let’s further assume that the Cedarville station is urban, and is cursed with the typical frailties of an urban station: lots of asphalt, little vegetation, and placement near an air-conditioner, strip mall, or jet engine.
Oh, you mean a genuine GISS Urban Cool Park.
Mike, thanks. From your cite: “The station closed in 1995 but the data continues until 2006”.
If this is true and the data represented in the graph was used in any journal article to support a global temperature figure, this needs to go to Congress.
you used nightlights to classify rural? correct
Steve Mosher: GISS used night lights, not me.
Some of the endpoint cases continue to confound those of us trying to understand the source code and how it reconciles – or doesn’t – with the peer-reviewed literature.
This is where Software Assurance is supposed to enter the picture. Code for which the Software Assurance process has been fully followed can generally be trusted to perform in accordance with its specification.
Based on the preliminary results of my first round of FOIA requests, it appears to me that the process has not been followed.
Sadly, software which has not been through a Software Assurance process can generally be trusted to NOT perform in accordance with its specification.
The points used to construct GISS plots are clearly no longer “data points”. Perhaps we could refer to them as “agenda points”.
Nice work John,
One comment I would make is that GISS adjusts cedarville because it is nightlights = dim. Simply, based on a photograph take from space in 1995, they determined that the area was not rural during its entire history.
Everyone needs to understand the magnitude of that STUPIDITY. So because Cedarville has lights at night in 1995, we can determine that in ,say 1910, its temperature record was UHI infected and needs adjusting. It’s lunacy.
REPLY: I’m reminded of the objection of one snarky Rabett when I first started the surfacestations project when bunny boy said in essence: “pictures don’t matter because they represent only a single point in time”.
Yet Hansen uses a single satellite picture from 1995 to determine the adjustment for the whole record, for the whole USA.
Screwy wabbet, where’s the outrage?
Stupidity indeed Mosh. – Anthony
John,
One last point. GISS throws some of the northern calfornia data before doing its adjustment. See hansen et al 2001.
The strong cooling that exists in the unlit station data in the northern California region is not found in either
the periurban or urban stations either with or without any of the adjustments. Ocean temperature data for the same
period, illustrated below, has strong warming along the entire West Coast of the United States. This suggests the
possibility of a flaw in the unlit station data for that small region. After examination of all of the stations in this
region, five of the USHCN station records were altered in the GISS analysis because of inhomogeneities with
neighboring stations (data prior to 1927 for Lake Spaulding, data prior to 1929 for Orleans, data prior to 1911 for
Electra Ph, data prior of 1906 for Willows 6W, and all data for Crater Lake NPS HQ were omitted), so these
apparent data flaws would not be transmitted to adjusted periurban and urban stations.
Bobby Lane (21:14:20) wrote: “…With Hansen’s stance on global warming seemingly ruling the roost at NASA, what good does much of this discussion do except to further make the point that agenda-driven science is what drives these GISS adjustments?”
It’s actually a waste of time Bobby… if you stick your head in the sand. But if you take information picked up here and at such places as Climate Audit, Climate Clinic, and others, you will be more informed and if hopefully, pass the info onto others in your family and family of friends.
Jack Koenig, Editor
The Mysterious Climate Project
http://www.climateclinic.com
Steve Mosher,
I am not sure the stations are thrown out as they are included in the v2.inv file STEP 2 uses. I’d be interested in seeing where in the code (any step) those stations are discarded. If they are thrown out I can try to redo the analysis without them to see if the Cedarville – Rurals difference is closer to the GISS adjustment.
John, did you see my question above? Why are you suspicious of the “batman” warming of the 30s? Just curious…
As this is followed by examination of rural stations, should this read “surrounding rural stations”?
Reply: Thanks – fixed.
From the station history documents, I think this station installed a CRS in 1931 and there seem to be changes of observers about then too.
Why be suspicious of Batman in the ’20s?
Why it’s the Joker of course. He made off from the local A&W stand without paying for his root beer.
This unprecedented crime wave in the little berg of Cedarville launched the largest man-hunt in Cedarville history. Alas, the locals could not apprehend the Joker.
The sheriff had no choice but to turn on the Bat-signal, which in turn was detected by the satellite, resulting in the misclassification of Cedarville as an urban setting.
Now the Joker has been quoted as saying, “This town deserves a better class of criminal… and I’m gonna give it to them.”
Perhaps the adjuster of GISS data is the better class of criminal?
Glenn,
And that’s the crux of it. But the real dog withmicro climate issues is the whole hom. adjustment process itself. They come up with a corrected temperature by comparing with surrounding stations. It works okay with UHI, station moves and equipment problems but not with many of the things we have found on the surfacestations project. The UHI, relocations and equipment changes are usually associated with large discontinuities and they are well studied and understood. So the hom adjustment catches them for the most part. The microclimate effects on the other hand are not as understood. In addition, you have to wonder how they correct one station (by comparing it to other stations) when 89% of the surrounding stations have simillar problems.
But let;s be clear on one point. The real lesson learned by the surfacestations.org project will not be in how badly the surface temperaure is affected, but how lousy the quality of climate science really is. You would think at least one of these guys would have visited some of these stations. You know that if they did they would have gotten a clue really fast.
That said, the real big tamalle is going to be Anthony’s initial project, the paint. It’s a slow upward creep that even the NCDC admitted will not be caught by the hom adjustment. That and the trees, they grow slow, and the bigger they get the more the local wind is blocked / boundary layers do not mix.
How is the thermal effect of the various wildfires accounted for? This is not the first year California has had this problem, and I wonder how the record might be corrupted?
REPLY: If anything it makes it cooler, due to the heavy amount of particulates in the air blocking sunlight. There were several days when the temp was forecasted to hit 100 and it was uppers 80’s or low 90’s due to smoke cover.
Jeff Alberts, I know that it’s probably not worth mentioning, but I think that John was referring to the shape of the graph between 1928 and 1938. Unfortunately, explaining jokes doesn’t work
Tony, oh I got the joke right off the bat (no pun intended, well, maybe a little). The graph looked like the bat signal, or batman’s headcover. But I gathered he was suspicious of the warming in the 30s, and was wondering why. If I was reading too much into it, I apologize.
I have no idea why Orleans shows significant cooling, but it is probably significantly different than many of the other stations in terms of geography/topography. It is deep in the Klamath River canyon which at that point is quite narrow and incised. Is the station measured by the US Forest Service office there? It may have been moved on the station grounds as the station developed? Just a guess.
This shows, yet again, there is no way to use this old station data for real science.
These adjustments are just hand-waving. While “batman” may look odd, and may be the result of a chnage in instrumentationl, or site, or something else, there is no way to objectively change the record in a way that verifiably makes it more accurate.
EVEN IF you had detailed station history documenting exactly what change might have been made during the 30s, there would be no way to assign an adjustment that was anything more than a guess.
Jeff Alberts, the jump in the 1930s just looked much larger than I would have expected, so it made me suspicious. It is the kind of thing I typically look for when I am looking for trouble 😉
It was my understanding that every temperature series available showed a jump in the 30s…
One thing I’m not sure on, and I hope that someone can set me right with this.
This station had it’s temperature record adjusted. Fair enough. When further adjustments are done on nearby stations, is this station’s history included or excluded? And if it’s included, is the raw data used or the adjusted?
In other words, the adjustment is a biasing of the data, and is this biased data used to further bias other data?
Spending all this time trying to figure out some way to adjust bad data to make it good is not only a complete waste of time, but it gives that much more credibility to Hansen. He’s made a career out of pushing this pseudo sceince as the real thing and out of his supposed wizardry in being able to massage, reinterpret, manipulate, tease out, reconfurbulate and discombobulate the data until it magically takes on meaning. This stuff is not science, it’s black magic and alchemy. It’s like the dark ages except that instead of potions of lizard tongues and bat wings recorded in dusty old books bound in human skin, he uses secret computer algorithms. And like the grand wizards of old, he guards his secrets jealously, thereby maintaining his illusion and gathering around him a host of awestruck apprentices.
I think the best way to put a stop to this nonsense it to declare it as such instead of trying to find the flaws in his methodology and improve on them. It’s not the methodology that’s flawed – it’s the whole concept of thinking that you can transform crap data into good data. But if we continue down this line, it just lends credibility to Hansen and others like him. People end up arguing over whose recipe of turning lead into gold is the correct one, in spite of the fact that none of them actually work and indeed none of them can work.