How Not To Measure Temperature, Part 64 – Estimating biases and comparing to GISS Homogeneity Adjustments

If you had the task of choosing where to put a climate monitoring thermometer here at the USHCN Climate station of record #469683 in Winfield, WV where would you choose to put it?

Winfield_MMTS_Site_View_South

Certainly the parking lot would not be a good choice. Maybe up in the grassy area behind the security fence? That would be my choice. Winfield is classified as a “rural” station so the grassy area would be a bit closer to the representivity for the area. It would also remove the sensor from the heat sinks of the parking lot and the building.

But then there’s that cabling issue with the MMTS sensor which this station has, it is a bit tough to trench through the parking lot up to the grass. So that leaves only one “logical” choice for placement.


Click image for a supersized closeup view

Surefacestations.org volunteer surveyor Michael Caplinger captured this location in his recent survey of West Virginia stations. As NOAA has already established with their training manual for the Baltimore USHCN station, rooftops are a far less than ideal place, and tend to create new temperature records where none actually exist.

According to the survey form submitted by Mr. Caplinger, he says:

“The new lock and dam opened in 1997. Prior to construction the weather station was possibly located about 100 yards West-Southwest, on land removed/altered for new lock. Reported coordinates appear incorrect for current location.”

According to NCDC’s MMS database, it appears that the MMTS came into being in August,1986, as prior to that they list the equipment type as “unknown”. That’s a good bet for the conversion date from Stevenson Screen, as MMTS did not start being implemented until the mid 1980’s

Also from MMS, and indication of the likely date of roof placement when the lat/lon and elevation changed significantly:

[1999-09-22] 2007-06-10 38.527220
(38°31’37″N) -81.916110 (81°54’57″W) GROUND: 611 FEET N/400/FEET PUTNAM 03 – SOUTHWESTERN EASTERN (+5)
Location Description: LOCK AND DAM, OUTSIDE & 1.2 MI SW OF PO AT REDHOUSE, WV
[1986-08-30] 1999-09-22 38.533330 (38°31’59″N) -81.916670 (81°55’00″W)
GROUND: 571 FEET — PUTNAM 03 – SOUTHWESTERN EASTERN (+5)
Location Description: LOCK AND DAM, OUTSIDE & 1.2 MI SW OF PO AT REDHOUSE, WV

In looking at the temperature record from NASA GISS, one sees what appears to be a step function around 1986, when the station changed to MMTS, seen in the data plot:


Click image for original GISTEMP plot

I downloaded the data, and there is an entire year of missing data in 1986, and the data resumes in 1987. This coincides with the equipment change noted in the NCDC MMS record on 8-30-1986. When I plotted the data and ran some curve fits and baseline value analysis on the two data segments, the differences became more apparent:


Click for a full sized plot

The baseline values between the two curve segments pre and post 1986 differ by 0.51°C, The slopes also differ significantly.

Looking at the GISTEMP plot for Homogenized data, you can note that the data has been shifted upwards a bit in the past, but the step function at 1986 remains:


Click image for original GISTEMP plot

When I plot the homogenized data, it can be clearly seen that there has been no change to the 1987 to 2007 segment of data, but that the 1905-1985 segment has been adjusted such that the early 20th century is a bit warmer, dramatically changing the slope for that segment.


Click for a full sized plot

The baseline difference between the two segments is less, now at 0.31°C

Here is the complete data set, with before and after Homogenization adjustment applied by GISS:


Click for a full sized plot

Note that unlike some other adjustments of rural stations we’ve seen where the past has been adjusted cooler (such as Cedarville, CA) in this case the past has been adjusted to be warmer, resulting in a slight cooling trend for the last century.

It makes no sense to me why GISS would adjust the past warmer. What could account for it? Certainly population growth wouldn’t be a factor, especially for a rural station. UHI doesn’t make any sense either.

Just for fun, I thought I’d try an experiment in data adjustment based on what I know about this station’s history. That isn’t much, but we do know these two dates:

1986 – MMTS installed, and likely moved closer to building due to cable issues
1999 – MMTS moved to rooftop of new locks building, based on lat/lon and elevation change

So based on that history, and having a handle on some other biases I’ve seen at the 500+ USHCN stations I’ve examined thus far, I decided to provide some offsets, based on what I believe a reasonable estimate of the bias might be:

1986-1998 = 0.5°C for MMTS to building proximity
1999-2007 = 1.0°C for MMTS on rooftop

Applying those adjustments and comparing to the GISS Homogeneity adjustment we get this:


Click for a full sized plot

Applying my station history based estimated placement biases as offsets post 1987, I come quite close in slope to that of the GISS homogeneity adjustment. My slope (dark blue) is actually just a tiny bit cooler than GISS. Some might say that my method uses too much “guesstimating”. But how is it any worse really than applying a broad brush algorithm blindly to the data, adjusting the far past, and without dealing with the step function that was introduced when the MMTS was installed? While my method is spur of the moment, it does have something the GISS adjustment doesn’t; adjustments based on known history and known measurement environment. GISS certainly does not know the history or measurement environment in the period that their automated algorithm applied adjustments. NCDC doesn’t have the station history for that period online either.

Looking for another nearby rural station to compare to, the closest I found was Spencer, WV, at 58 kilometers away. It also has a cooling trend, a bit sharper, and most likely has not been placed on top of a concrete building, though the current location is also not the best, at a Water Purification Plant:

While Spencer’s placement at a water plant presently (since 2005) probably would take the “rural” portion out of the record, the previous portion of the station history appears to be truly rural. Up until 1995, it spent most of it’s life at USDA SOIL CONSERV, WITHIN & 0.5 MI SE OF PO AT SPENCER, WV. From experience, I tend to view places such as Ag farms like this as being fairly good sites that don’t get much if any encroachment. This I would tend to believe the Spencer, WV record as showing a true cooling.

So the question is, can we use station photographs and station history, combined with some bias estimates that should be quantifiable either by experiments or direct measurements on site to come up with a more realistic adjustment for USHCN stations? While this is only one example that appears to work, I think the idea bears exploring.

About these ads
This entry was posted in Climate_change, Science, Weather_stations. Bookmark the permalink.

36 Responses to How Not To Measure Temperature, Part 64 – Estimating biases and comparing to GISS Homogeneity Adjustments

  1. Harry Goff says:

    Andrew

    At the risk of sounding a bit heretical. Could you please revert to these type of posts. There are many other blogs providing the other info that you have started to put up. I like the info on the data collection asit stands on its own as to relevence and topic. The other info attracts all sorts of other distracting traffic (such as MA etc) which is not really all that useful to the discussion here. Your blog is great as it is without the distraction. If you stick to this you may not have to get other help as you have requested. Look at CA for a history lesson.

    Apologies for being OT, but I feel that you may get horribly distracted if you do not stick to the knitting.

    best wishes

    Harry Goff (please note I have used my full name – no flames necessary or recognised)

    REPLY: Thanks for the note – Anthony

  2. Mike Bryant says:

    Sheesh!!!

  3. John Goetz says:

    Anthony, this was a thoughtful analysis. I agree that the station history needs to be considered rather than making blind adjustments with a do-all algorithm. I see no reason why your methodology for this particular station should not be favored over the GISS adjustment.

  4. steven mosher says:

    If it’s been adjusted by giss, then its not rural. check nighlights, we may have another case of rural population but dim lights at night ( by the dam ) so,
    if it gets adjusted by giss its because giss thinks it is and urban site or a small town site, not rural

    REPLY: Mosh, on the GISTEMP station selector page (when you type in Winfield as search) they describe it as rural:

    (*) Winfield Locks 38.5 N 81.9 W 425724140010 rural area 1900 – 2007

  5. Gary says:

    Perhaps this is another version of the Turing Test. We’ll know the computers are really smart when they can adjust a temperature curve as well as a human. ;-)

    FYI, the GalaxyZoo project was designed to use minimally-trained human judgment to classify galaxies and quite successfully achieved it’s objective. The results were statistically similar to the judgments of expert astronomers. It seems to me that the same general approach would work here: get several people to propose adjustments to a station and compare their results. See how closely their individual judgments agree. If, say 20 people, work on a couple of stations as an experiment, their results ought to have a range that can be analyzed. Maybe they agree; maybe not. Worth a few hours of time to find out.

    REPLY: Intriguing idea. Something similar was once done with short term weather forecasting in he early 80’s I think. A group of non meteorological people were given basic rules and oitted against the NWS at the time, they did surprisingly well.

  6. Tilo Reber says:

    You have to feel sorry for the warmers. All of the trends seem to be going against them lately. Now it looks like another trend, the sea level, is also starting to change. I plotted the last 3 years of sea level data from the University of Colorado, and there is no sea level rise. So if Greenland and Antartica are melting, where is the water going? If the ocean is warming, why isn’t the expansion showing in sea level rise? Even if only the deep oceans are warming, why isn’t that showing in sea level rise?

    http://reallyrealclimate.blogspot.com/2008/06/university-of-colorado-global-sea-level.html

  7. counters says:

    You asked, “It makes no sense to me why GISS would adjust the past warmer. What could account for it? Certainly population growth wouldn’t be a factor, especially for a rural station. UHI doesn’t make any sense either.” There are a few other reasons begging consideration. Note that they are hypothetical, and there are caveats that exclude certain ones in this case, but they are good possibilities for many of the surface stations brought up in posts across the blogosphere.

    For starters, there is the obvious instrumentation bias. The instruments could be improperly calibrated, could be improperly read, or just plain inaccurate or lack precision. In a similar vein, there is an oft-overlooked bias – timing of measurements. It makes a difference what time of day measurements are taken, and shifting between those times will lead to bias in the datasets. These biases likely do not apply in this situation, but they are still important considerations.

    As for the homogenization resulting in a warmer record, it’s possible that there was an issue with a systematic bias in the temperatures at this station. The micro-climate of the site could just have been a bit off, resulting in a weird temperature record.

    Now, for your assertion that the Spencer sites illustrates a “true cooling.” It’s one thing to plot a negative linear regression, but it takes a bit more to declare that that regression illustrates a cooling. What was the r^2 on the regression? How has the variance trended over time? Just by eyeballing the dataset (my own “spur of the moment” method, if you will), there seems to be hardly any discernable trend going on. Regardless, if we assume that it is justifiable to say the data supports a cooling trend, then is there even any significance to this statement? Does a trend at a point on the Earth’s surface extrapolate to the trend of the atmospheric system in general?

    These are questions that necessitate thorough answers, especially in context of AGW theory.

    REPLY: You can see the GISTEMP plot and download the data (the same data I used for the Spencer plot) here, why not run it yourself?

    http://data.giss.nasa.gov/cgi-bin/gistemp/gistemp_station.py?id=425724140020&data_set=1&num_neighbors=1

    BTW, Time of Observation Bias (TOB) is already in the USHCN data set, having been put there by NCDC. The data is already pre-adjusted for the many issues you cite by NCDC. GISS adds yet one my adjustment, Homogenization. It is that one which I’m questioning.

  8. Bill in Vigo says:

    I am unused to being one of the first to comment on a post. I have great disappointment in our premier scientific organization. (NOAA, NASA, NCDC) I am truly disappointed that the powers that control these organizations would allow such adjustments to go on. Like you Anthony I don’t understand why with the histories of the MMTS systems and their inherent bias due to cabling problems why the adjustments must be for the past when the bias is obviously due to the placement of the measuring device for the most part. I don’t understand why they haven’t changed to the remote data transmission (wireless) that we are capable of today. By your own estimates this would possibly remove in this case 1.5C bias from the modern (last 20+years record) method of data collection. I also have concern that our records from prior to the mid 80’s may have been compromised due to massive amounts of adjustments and estimates of areas of no data. We can’t go back to 1890 and re-measure the data and I can’t imagine a technically correct way to correct the “old” data. I also don’t understand how when you build a new site with a new building and a new parking lot why the forethought isn’t done to put a pipe under the parking lot to Pull the cabling through. I bet they remembered the telephone cables and the electrical cables.

    This may be a conflict of terms but when lives depended on results the military has a rule .

    The rule of the 7 P’s

    “Proper Prior Planning Prevents Piss Poor Performance”

    Perhaps our folks that have oversight of the USHCN and the remainder of our climate/weather data collection assets should implement and enforce this rule.

    Bill Derryberry

  9. truthsword says:

    I have a personal weather station on my roof (just no other place to put it where I live) and it runs 5-10 degrees hotter because of that. I allow for that, if my temp is 97, I know its about 90. So these findings certainly reflect this situation.

  10. stan the man says:

    i would place it on top of the flag pole protected somehow from direct sunlite..

  11. George M says:

    And exactly what would you call GISS “adjustments”? The “data” upon which thay rest seems little more than a guess also. With the Surface Stations information from real observers, rather than numbers churned out of computers with ever more suspect programming, I think you’ve got the upper hand here, Anthony. Carry on!
    George M

  12. Tom Bruno says:

    truthsword: “I have a personal weather station on my roof (just no other place to put it where I live) and it runs 5-10 degrees hotter because of that. I allow for that, if my temp is 97, I know its about 90. So these findings certainly reflect this situation.”

    I think this one statement summarizes the entire average Earth temperature measuring debate. If you say the temp runs 5-10 degress hotter due to this location, how can you “know the temp is about 90″ when the reading is 97. It could be 87 or 92 or anything inbetween. So you really don’t “know”. It’s just a guess maybe even a WAG. I suppose for a individual who wants to know about what temperature it is, this is OK but for anything else it is pure folly.

  13. steven mosher says:

    Anthony,

    The web page uses the rural code of the database which is based on population.
    the ACTUAL gisstemp code uses nightlights for the US. Go review the nightlights posts on CA and you’ll see what I mean.

    For example. Cedarville. It’s RURAL by population, so on the webpage you will see it as rural. BUT, since it has a nightlights = dim, it gets adjusted. Because nightlights says its “not rural”

    Confusing? yup.

  14. Tom says:

    Anthony,

    If I’m reading right, one (or more) of your previous posters suggest that you “stick to the station posts”.

    I just wanted to make sure that the opposing viewpoint was not lost to the “squeeky wheels”. Your blog is one of my daily reads. I find all of your posts to be informative and insightful.

    I think all of the information about stations has to be taken in context – and that’s what I feel the other types of blog entries do.

    It’s one thing to say “the government is screwing up temperature monitoring” (so, what else is new?), but more importantly, “what does this mean to me?”.

  15. Earle Williams says:

    counters,

    if it has a negative slope it is a cooling. Pure and simple. Now is it a statistically significant cooling? Dunno, run the numbers and see. But don’t confuse statistical certainty with fact, nor human perceptions with fact. If it looks to be cooling then it darn well is cooling, stats be damned!

  16. steven mosher says:

    Yup Anthony,

    Winfeild is RURAL by population, but has a nightlights =2. This means
    the Algorithm will consider it to non rural and will adjust it. GHCN says the site is BRIGHT.

    This may actually be a really good case to highlight the problems with using nightlights to determine rural.

  17. AnonyMoose says:

    “why they haven’t changed to the remote data transmission (wireless) that we are capable of today”

    Because someone would forget to change the batteries? Because power cables have to be run to the MMTS anyway so data cables may as well also be used? Because the radio signals don’t travel well into the industrial buildings which are often used at government sites? Because they can’t trust that data won’t be lost when later other radio devices are installed which interfere?

    REPLY: Your argument is weak. Small solar cells run these for years even in cold climes like Alaska.

    http://www.weathershop.com/davis_vantage_pro2.htm

    this would also work:

    http://www.weathershop.com/USB1_temperature_logger.htm

    Given that observers are required to read once a day, I think they could be trained to keep this equpiment running. If they can’t be relied upon to replace a battery or check for operation once in awhile, then they probably can’t be relied upon to get the measurments.

  18. Luke says:

    I stumbled across this site a few months ago and really enjoy the variety of content published. Some have argued for more focus on the misadventures of temperature taking, but I don’t feel like one would get the full picture without your postings on solar cycles and data mining.

    I think one only starts to understand the ruse of AGW when the full picture is presented…

  19. Brendan says:

    Do they use census blocks to help them make their assessment? Or perhaps some of the adjusted population data developed by LANL? Either of those would be better to assess population than nightlights.

  20. Mike Bryant says:

    Anthony,
    Your work is opening eyes all over the world.
    Thank You,
    Mike Bryant

    “Guest Weblog by Dr. Kiminori Itoh

    Dear Readers of Climate Science,

    It is an exceptionally great pleasure for me that I can introduce to you my recent book “Lies and Traps in the Global Warming Affairs” (in Japanese unfortunately),” in this weblog, Climate Science, which I believe to be the most prominent in this field……
    We have described many topics in this book, including inaccurate temperature measurements (e.g., A. Watt’s work), “observations” of climate sensitivity, many climate forcings such as colored-aerosol and vegetation (based on 2005 NRC report as Roger has so many times pointed out), and the effect of solar magnetic activity (including my own work). To discuss these topics I largely depended upon information provided on Climate Science. I really appreciate Roger for his continuous effort to keep the weblog at a high academic level.”

    REPLY: Thanks, I had no idea they were paying any attention in Japan. -Anthony

  21. Eric (skeptic) says:

    By my experience, a roof mounted station has higher highs but also lower lows.

  22. Mike Bryant says:

    Anthony,
    Maybe you should do some presentations in Japan.
    Mike

  23. Gary Gulrud says:

    No offense intended, but yet another network hardware upgrade is like ‘lipstick on a pig’.

    REPLY: Only if you keep the same site, as you can see from my previous post on the USHCN upgrade, they are choosing new locations.

    http://wattsupwiththat.wordpress.com/2008/04/29/what-the-modernized-ushcn-will-look-like/

  24. Mike Kelley says:

    Anthony, this loyal reader says don’t change a thing. Thanks for all the entertainment.

  25. Rod Smith says:

    I’d like to go backward one further step and look at the Federal Meterological Handbook, FMH-1, (easily available on-line) Titled “Surface Weather Observations and Reports.”

    These are essentially US rules for reporting weather, mainly at aerodromes. Although generally not used because of Urban Heating, these are quality observations, hardly like the USHCN product.

    Appendix C of this manual is Titled “Sensor Standards.”

    For temperatures in what I will call the “normal” range, -50C to +50C, the sensor accuracy is specified as plus or minus 0.6C with resolution to 0.1. Accuracy beyond this ‘normal’ range is plus or minus 1.1C with resolution remaining the same.

    THEN, the temperature and dew point is reported in whole degrees Celsius! Unless the original log is available, this .1C is then available to the rest of the world rounded to the nearest whole degree!

    These sensor standards of course apply only to US observations but the reporting standards apply world wide. I have no idea what sensor standards are required of the ROW, nor how much ‘policing’/quality control and/or calibration is routinely done.

    Still it has always struck me that is absolutely absurd to calculate a mean temperature of our earth to .1C using such crude input.

    Then to adjust the basic readings according to some canned formula adds, I think, insult to injury.

    Rant over!

  26. Mark says:

    To the webmaster who runs this site…

    This is a great site but those little pop up boxes that come up when you mouse over them are a royal pain in the rear.

    Mark

    REPLY: You can disable them, just click on the little gear icon in the popup for options

  27. Jeff C. says:

    Anthony,

    To add support to your theory regarding the instrument change in 1986, the NCDC station history file lists the instrumentation as CRS and MN, MX from 1900 through 8/29/1986. It then switches to MMTS on 8/30/1986.

    That squares with the info from NOAA. It also confirms your assumption that a Stevenson Screen was used prior to 1986.

    Also on the NOAA MMS page under the updates tab for Winfield Locks it states “EQUIPMENT MOVED 100 FT N, 10. SHOW LAT/LON TO SECS, CHG REMARKS” under the date of 9-22-1999.

    They sure do make it difficult to piece this info together.

  28. Evan Jones says:

    GISS adds yet one [more] adjustment, Homogenization. It is that one which I’m questioning.

    That’s when you shake up the data so hard all the discernible errors disappear.

  29. Evan Jones says:

    i would place it on top of the flag pole protected somehow from direct sunlite.

    You mean like the politicians?

  30. Evan Jones says:

    I don’t understand why they haven’t changed to the remote data transmission (wireless) that we are capable of today.

    The NOAA/CRN is supposed to do just that. It goes on line in a few months. We’ll have to wait and see. (They have photoed all the stations. We’ll have to see if those photos are available to the public.)

  31. Hoystory says:

    So the question is, can we use station photographs and station history, combined with some bias estimates that should be quantifiable either by experiments or direct measurements on site to come up with a more realistic adjustment for USHCN stations?</cite

    Umm…. only if they confirm the AGW consensus. Otherwise, forget it.

  32. Eric Baum says:

    Just compute the temp gain from the classes of stations individually.
    There are plenty of class 1 sites to give you a good picture of what has actually happened to the temperature, and it would be interesting to see a graph of temp anomaly vs class.

  33. Dell says:

    Anthony

    Just a suggestion to perhaps compare an analysis of the GISS surface temp data to Satellite data and the NOAA US surface temp data to Satellite temp data.

    If I recall correctly, doesn’t the RSS data have numbers for the Continental US? How easy would it be to compare that to the US surface data from the NOAA for the same period..

    It would be interesting to see if there is a significant correlation between troposphere temps and US surface temps, to possibly demonstrate how far out of whack the GISS surface temps are.

  34. KlausB says:

    Hum – sorry, just kiddin’ – now I have an idea for a real good job for Jim Hansen.
    He should join DowJones Newswires and take care of calculating
    the DJIA. Should then be 25,000 by now.

    On the other hand, maybe Jim was already contracted for the birth/death model
    numbers within the unemployment data by the BLS. On a Richter scale of meaningless statistics it tops GISS by far.

  35. fred says:

    Anthony, here is what is hard to understand. You have documented lots of really bad errors in siting, which seem to have a warm bias. Also the adjustments seem to have a warm bias. And yet, the US ground station record as a whole doesn’t seem to have that much of one. Would you not have expected it to have more, and show more warming?

    It more or less corresponds to experience that recent weather is warmer and milder than (say) the fifties. In addition, it shows a degree of warming which is milder and more reasonable, and more in line with balloon and sat data, than the ROW records.

    So how come, if there are all these warming biases built in, that it does not appear to greatly exaggerate the warming? And why does it do so, if it does, so much less than the ROW station data? Or do you take the view that, absent these biases, the US record would actually show static or cooling temps? Seems implausible.

    REPLY: I haven’t taken any view on this yet becuase I’ve been waiting to get a majority of the stations in USHCN surveyed so that we have both spatial distribution and significant quantity of CRN1 and 2 rated stations to analyse

  36. Doug says:

    I love your website, Anthony. Much of the technical talk goes over my head, even though I’m an engineer (electrical). Or maybe *because* I’m an engineer.

    I have to ask this question, though. Why do the officials who gather data, when confronted with obvious “oops” situations, then devise some unproven and often controversial adjustment without ever running some simple experiments to prove their validty?

    AnonyMoose points out a little portable USB measurement stick. Leaving aside the unit-to-unit absolute accuracy (probably +/- 2 or 3 degrees C, but a simple all-at-once calibration would bring them all into line), one could buy a few of these portable devices, and place them in the various locations (on the flag pole, on the roof, on the lawn, on the building, etc.), gathering simultaneous data points at different locations. Gather the data, then actually SEE the temp deltas.

    Instead, we have these fiats from some obscure, anonymous hack with zero data to prove the assertions. If you question one of the government nazis, then you’re “a denier.” Stop guessing and take the data! After all, isn’t this about “science”?

    What am I missing?

    REPLY: Continuing of funding and job security

Comments are closed.