How not to measure temperature, part 35

One of the emerging patterns that I see again and again as the volunteers and I survey the USHCN climate stations of record around the USA is that many of them have been relegated to back lots with an “out of sight, out of mind” attitude. At California Polytechnic Institute, San Luis Obispo, that attitude seems obvious.

Here we have a USHCN station in the middle of a junk pile. Old storage crates and containers, a rusty metal trailer, and a disabled Winnebago. Most curiously, one piece of junk, a discarded solar mirror, has the potential for heating up the thermometer more than a few degrees under certain conditions. It’s hard to believe, but this is exactly how I found the site.

San Luis Obispo Overview

USHCN San Luis Obispo Cal Poly Looking North

This was my second attempt at surveying this site. In my first attempt, I found three other weather stations, all on campus building rooftops, and one of my readers said he spotted the “official station” in grass plot near a large parking lot. It turned out to be none of those. This station was tucked away behind the maintenance buildings.

Note how the solar mirror is pointed right at the shelter, under some sun angles, I’d wager that this reflects sunlight directly on the shelter. At other times, some localized heating of the ground and objects near the shelter could also occur, heating the air nearby. Here is another view:

San Luis Obispo looking ENE

Catching some rays?

In addition, the Stevensen Screen is wind sheltered on all sides, and a pool chemical test facility was recently constructed about 90 feet south, which adds humidity to the air, possibly increasing Tmin at night.

San Luis Obispo aerial view newer

Click picture for a larger image

And, if that isn’t enough, the Stevenson Screen is at the crest of a small hill with acres of ashpalt and vehicles just below:

San Luis Obispo looking NW from lower parking lot

Given these many microsite biases, it is not at all surprising to see a sharp upwards temperature trend for this location.

San Luis Obispo Poly station plot 

The Cal Poly campus, like many, has seen a lot of building going on. In the last 5 years, two new buildings have been constructed just south of the station, and there have been other land use changes.

It seems a far less than ideal place to measure long term temperature trends. The complete station survey with additional photos is available on my image server.

21 thoughts on “How not to measure temperature, part 35

  1. The elusive SLO is finally trapped. Good job. My apologies for misidentify the instrument package 1/4th mile to the east.
    I’d be willing to wager that those temperature track almost exactly campus enrollment. The planning office has those figures on paper going back to 1960 but is too busy to provide them to me. It would be interesting to plot.
    Doctor Evil asks: “Is it too much to ask to have some friggin’ l-a-s-e-r beams pointed at the Stevenson screen?” 😉

  2. Evan I haven’t rated it yet, and it may not be a 5 since I have no measurements that show the solar mirror illuminates the screen. I didn’t have the opportunity to run tests in that regard while I was there. At best it’s a 3, worst it’s a 5. I have to study it more before issuing a rating.
    Robert, no worries. I initially went to the station in the grass insert you pointed out, but after double checking, I realized after taking photos that the MMS database said “MAX-MIN Thermometers”. I had a more accurate lat/lon fix also in the database so I followed it with my portable GPS until the site came into view. Your starting point got me closer than I was before, down by the Aero building.

  3. All good scientists just love data. Doesn’t seem to matter how they get it. But what’s the point? Everyone seems to be forgetting that all weather related data is useless to anyone other than historians and record keepers.
    The data is fed to computers to find patterns which give us trends which we use to make predictions. Predictions based on the number one overlooked observation in all the history of climatology. WEATHER DOESN’T HAVE A PATTERN. The patterns alleged are a construct of the human brains remarkable capacity for prejudice. We’re pattern finding machines far better than any computer ever built. Here’s an example…
    What’s the next character? Q!
    It’s a repeating sequence of 12323434545678Q12323434545678Q12323434545678Q
    No amount of perfect data, supercomputers, and political/economical influence-free scientist will ever be able to guarantee that a weather pattern from the past is guaranteed to repeat or that a trend will continue in the future.
    I do appreciate what you’re attempting, but I think the best way to assuage the fear of the clamoring masses to once again publicize the certainty “No one can predict the weather” Unfortunately, I’m essentially asking most of the well-meaning, highly trained experts on both sides of the issue to admit defeat.

  4. Anthony,
    I’m feeling a little better that you were fooled as well if only for a few minutes. The screen location just has me crazed. Yeah the junk and shade and reflector are issues but the parking lot. This isn’t just any parking lot, it is a university parking lot. It goes from overflow full of thousands of hot engines today to empty tomorrow (Thanksgiving).
    Another thought. Do the naval stations on San Nicholas Island and San Clemente Island have compatible weather packages?

  5. To keech: of course weather has a pattern! During the winter months cold fronts move from the n.w. to the e. and s.e. Then wamer air from the Gulf pushes n. and n.e. Where the two fronts meet precipitation falls or heavey fog occurrs. There’s a pattern for you. I can predict that mid-February will be colder than mid-Sepember based upon records of weather conditions during many years past. And I’ll be right more often than wrong. Guaranteed!
    Many events in earth’s history may be unique but by looking for patterns and trends is possible to say that climate varies without our imput so every variation is not a cause for concern. Having a reasonable framework for interpretation of information is what is lacking by alarmists not information itself. Certainly throwing ones’s hands in the air and proclaiming that an understanding of the earth’s climate must forever remain ‘terra incognito’ is to cede the whole matter to the necromancers and navel gazers.

  6. To r. chamness: Please don’t misunderstand me. I never said that there are no cycles. However September has not always been autumn historically speaking. The calendar has had to be updated several times through the centuries to adjust to astronomical observations. To say that you know what will happen IF two other things happen does not mean you can predict weather the two other things will happen. Also you use the phrase “I’ll be right more often than wrong. Guaranteed!” A 55% accuracy rating isn’t exactly compatible with the concept of guarantee. I can say that in the northern hemisphere winter is colder than summer. But that’s only guaranteed if nothing new or unpredicted occurs, just because it’s likely not to, doesn’t mean it won’t. Of course taking the term prediction and applying it to a broad enough generality to maintain a favorable likelihood doesn’t make you more accurate it makes you less useful. Don’t get me wrong, I like the weather channel and find it very useful for planning next weekend, filing flight plans, or getting the surf report. I merely want to point out that you very rarely will see a professional give a 100% chance of anything. The climatologists in question (I by no means accuse all of them) are claiming that their computer models can predict a trend to continue for 100 years to accuracy of 3 degrees or less even though they don’t necessarily know what’s causing the observed trend. For further information take a look at the charts of record highs and record lows by state and try to find a pattern there. I’m sure you could, but you’d be a fool to claim that you could make predictions based on it.
    And about those “navel gazers”, I’ve known old men whose trick knee was more reliable then the local meteorologist. I’m sure you meant clairvoyant rather than necromancer. I’ve already gone too long I’ll rant on about the marked similarities between ESP studies and weather predictions later if anyone’s interested.
    I really do like reading these forums and finding out about the inadequacies of our current measurement system please don’t think of this as criticism, merely side-bar.

  7. Keech: A good scientist does care how he obtains his data. It’s the tools he works with. You watch t.v. weather? why, if all weather related
    data is useless except to historians and record keepers? If all weather related data is useless except to historians and record keepers then when you pilot a plane you need not get a weather prediction for the area your flight will cover. It’s of no use to
    you, you not being a historian or record keeper.
    This site showing weather stations makes the point that weather related data really is useless if it’s not accurate as possible and will lead to inaccurate conclusions. gi,go. Climate models fail because all sorts of proxies are used in place of unavailable data, causes and effects are confused, and the desire to reach a certain conclusion. I’m not a fool under any circumstances and a chart that is shaped like a bell could be used to make predictions using record highs and lows. Being correct 55% of the time would earn me a fine living on Wall St. or in Vegas.
    Actually I used exactly the word I wanted, “necromancer”, one who seeks information by divination, not “clairvoyant”, one who sees
    something without using the bodily senses. What a “navel gazer” is I’ll let you figure out. Cheers!

  8. Necromancy often refers to magic involving the dead, not necessarily divination of any kind. Clairvoyants perceive things, often claimed to include the future, without the five well documented senses. This is a semantic argument due to my not defining terms up front which is the key to sensible debate. I apologize.
    Flight relevant data is real time, not prediction. If a broker advertised that he was right 55% of the time he’d be out of a job in no time. 55% isn’t a prediction it’s a chance. You have to actually make predictions and judge your accuracy against the proposed chance before you can begin comparing yourself to divination.
    If you look at a period of time wherein the temperatures gradually increased (usually using a smoothed graph) then you will see a trend. If you increase the range of time the trend disappears. Likely another trend or pattern will be visible, increase the range again and that pattern is toast, ad infinitum.
    I never claimed that given observable wind velocities, observable storm conditions, and observable pressure differentials a competent meteorologist could not provide a very useful range of possibilities. It is however still no guarantee of accuracy and I don’t know of any weather man that would claim it to be. I’ve lived in hurricane prone areas and tornado prone areas all my life. You might be able to give me a projected path after the storm has manifested (though I’ve seen those projected paths err by as much as 180 degrees). You still can’t tell me whether or not a hurricane will develop in the first place. This last lackluster hurricane season is proof enough of that.
    Weather is not mathematic. Applying math to the problem will never yield a magic window that can project history out into the future. I never meant to get this argumentative. And I did not call you a fool. If you voluntarily wish to do something I thought you weren’t foolish enough to do then that’s up to you. I can’t believe you looked at the charts I was talking about if you think a simple bell curve fits in any way. Let’s allow that a bell curve did fit the record of each state’s record high (since modern records start anyway) would you honestly bet the farm on when the next record will hit based on this graph?

  9. I was told way back in junior high band that a “naval gazer” is someone who stands around with his head down all the time rather than looking where he’s going or paying attention.
    I do enjoy intelligent debate such as this and am starved for it at work. I apreciate your taking the time to oppose the viewpoint of an unimportant complete stranger whom you’l never meet. Gotta love the internet.

  10. Hi- It’s been a real pleasure reading your “How not to measure temperature” series . When you consider that national policy is being made on the basis of this nonsense, it’s obviously a very serious matter, but nevertheless, when I see these “stations of official historic record” sitting in the middle of asphault parking lots with AC condensors blowing right on them, I can’t help but laugh out loud. This is abso-frickin-lutely hillarious stuff! This is what I love about the internet. It’s getting harder and harder to pull a fast-one by relying on the laziness and ignorance of the general population on a given topic. Now, there’s always ONE guy out there somewhere with nothing better to do than to fact check and expose fraudulent data. It brings to my mind the guy who recently figured out that net-flix was intentionally shipping dvd’s more slowly to some people than to others to maximize profits. They tried to pull a fast one, and somebody with nothnig better to do caught them and exposed them instantly to millions of people on the internet. The power of information, communication, and collaboration is awesome. The truth simply can’t be hidden anymore, and that’s a good thing if you ask me. Please accept my sincerest thanks for your hard efforts at doing the dirty work of digging into the details that most of us just accept like sheep.Bravo!

  11. Flight relevant data is real time, not prediction. If a broker advertised that he was right 55% of the time he’d be out of a job in no time. 55% isn’t a prediction it’s a chance. You have to actually make predictions and judge your accuracy against the proposed chance before you can begin comparing yourself to divination.

    But with a prediction such as “In Virginia it will be colder in January than in August”, I will be right 100% of the time, or at least 99.99999% of the time. You can predict some seasonal/weather patterns, but not specific events, such as tornadoes, cyclones, thunderstorms, cloud cover, etc.

  12. Jeff: Yes.
    Ty: Yuk it up! Laughter may be the best medicine. But, depending upon the wielder, it is also the sharpest sabre.

  13. Adding my fuel to the fire of prediction…
    Yes, we can all say that the summer is going to be hotter then the winter with a 99.999% accuracy. But who among us can say what the temperature is going to be within 0.1 degree C? That is what ‘they’ are doing. They are predicting temperature X based on past readings.
    They are the same as the fraudsters selling gambling systems. Anyone can make a gambling system based on past events. All you have to do is download roulette spins and construct a system that will win every time based on the past proven numbers. Casinos are happy to give away this information. They know people will look at the data and ‘discover’ a ‘pattern’ and construct a ‘winning’ system. The casinos know that you are going to lose your shirt.
    A casino will kick you out for counting cards (the only real chance you have, outside of cheating). But they will happily SELL SYSTEM BOOKS in the casino for you to ‘win’ with. They know the predictions based on past events are meaningless.
    Why has Gore not made “Truth 2”? With predictions of how his company and their carbon trading and other ‘products’ will save the planet? Surely all he has to do is remove X amount of CO2 from the figures for the planet to be ‘saved’? He could make another $100 million on this sequel “Return of the Killer Truth”.

  14. “They know people will look at the data and ‘discover’ a ‘pattern’ and construct a ‘winning’ system. The casinos know that you are going to lose your shirt.”
    In my field we call the converse of that one the “Texas Sharpshooter” fallacy. Flurry of random shots into the side of a barn. Walks over to barn. Picks the biggest cluster of shots. Draws a circle around it. Says, “THAT’s where I was aiming.”

  15. The station hasn’t always been at that location. Where you see a jump in temp is roughly when it was moved. 1927 was the first move, from town to campus.

  16. That reminds me of a mullen weed I tried to take down with a .22 across the little draw and up the rocky hillside. I eventually shot it down but more importantly, the rock behind it was shot ALL TOO HELL!

  17. Pingback: Beyond the Weather Wall » Jan. 31st-Feb 1st

Comments are closed.