An Easy Airport Heat Island Audit?

It is just now early dawn…

I’ve been out to say “Good Morning” to my bunnies and to assure all is well. Also took a ‘seed jar’ from my seed archive (freezer and fridge on the back patio) as I’m going to start some beans this weekend. Noticed it was “a bit cold” and looked at the patio thermometer… At that time, it was 41 F. BRRRRrrrrr (Hey, you folks in Canada, stop laughing! Now! I mean it!! Look, I’m a native Californian, OK? There are places in L.A. that don’t even HAVE heaters in the apartments… and if your Mocha drops below drinking temperature in under an hour, it’s, er, well, we call it cold…)

So I’m back in the house and I’m pondering what GISS or CRU or NOAA would call this, and thinking of the nearest airport… When I remember that we have Wonderful Wonderful that takes airport codes as locations.

Type type typity type…

SJC is showing 46 F. Hmmm…..

Now my back patio is NOT an ideal measuring location, but the bias ought to be very much to the “warm” side, not the cool. And my thermometer is hardly calibrated worth a damn (about a 20 year old Big Round Dial job). So I get my “photography” thermometer and put it nearby. After a while, It’s reading about 47.x where it’s hard to tell exactly what the x is. A re-check of SJC shows it 48 F at that time.

But this one particular measurement is NOT the point of this article.

What’s The Point?

Each and every one of us can do a simple audit of Airport Heat Island Effect.

Get a good thermometer. Put it in a good location in your yard ( in a Stevenson screen if you can get one). Then compare to the nearest airport code at Wunderground.

Errors in thermometer siting are rarely to the low side. ( It is hard to make something artificially colder). So if you have an accurate thermometer (and I’m pretty sure Anthony Watts at WUWT has nice high / low recording thermometers for sale fairly cheaply) you can provide a “sanity check” on your nearest airport.

I’ll be getting a ‘thermometer upgrade’ for my place as soon as possible (and finding out the specs for a DIY Stevenson Screen ) and then plan to have a periodic check on my local airport. It is only a couple of miles away (same altitude and geography), and Hansen is sure you can compare and substitute temperatures ( with adjustment) from up to 1000 km away (and at different latitudes, geographies, and weather zones), so I’m pretty sure we can have useful information come from such an ‘Audit the Airport Heat Island’ project.

One would expect that the errors would have a more or less ‘bell normal’ distribution, so if we have a LOAD of readings that are lower than nearby airports, that would pretty much show that AHI is very real and give a rough magnitude. If there is little AHI, then there ought to be a more or less normal distribution about the airport temps from nearby observers.

Is there another way to do this study…

Sidebar on Airport Codes

When I first asked for local temperatures, Wunderground gave me the Santa Clara reading rather than SJC (even though I’d put in San Jose as the search term). I suspect it has some zip code info stored or that it ‘makes a good guess’ rather than jump immediately to the airport. When I put in the airport code, I got the airport, and it was a bit warmer.

SJC is presently 48 F while Santa Clara (PWS) is reporting 47.8 F so that’s 2/10 F in a few yards to small miles. (San Jose Airport is on the border of Santa Clara and San Jose). Gilroy is a semi-rural area about 10 – 15 miles south in the same valley. It is reporting 47.6 F at the moment. So as a ‘first estimate’, I’d say we’ve got about 0.4 F of Airport Heat Island. San Martin California, a bit closer, is at 47.7 F.

At times like sunrise, you do need to watch the time stamps. They do not all update at the same time and I expect the slew rate is pretty high at ‘first light’.

There is the very real possibility of identifying matched sets of airport code thermometers and “nearby” non-airport code thermometers and doing an AHI study from your living room or office without ever going outdoors (!) or owning a thermometer. This would take a bit more ‘desk work’ to find COOP or similar stations “near” each airport, but ought not to be too hard and is a lot less work than installing your own thermometer.

So, to be sure, you ought to use the actual airport code for getting the airport temperature. Then you can try putting in some nearby city names and see if there is a good reference set near your particular airport.

Also, FWIW, one study I saw on AHI said it was strongest on days with no wind (the idea being that the wind brings in fresh cooler air) so you might also want to pick stations in each major wind direction from the airport and compare them too. If the heat is drifting on the wind, there ought to be 3 cooler stations, and medium warm ones at the airport and just down wind, during modest breezes.

Why does this matter? Because the Global Historical Climate Network, GHCN, is presently 92%+ at airports in the USA and France is “all but one” airports (and one reader asserted the ‘not an airport’ was miscoded and really IS an airport…) so any AHI directly contributes to a bogus “Global Warming” factor. Many airports are also coded as “Rural” when they clearly are NOT a rural environment.

My Thermometer Location

Just so folks know, the ‘bias’ factors that might impact my thermometer are that it is over a flagstone patio (that is shaded) and under a white painted metal awning (but attached to the pole holding up the North edge of the awning, so right AT the edge). The awning itself tends to be shaded by the house from southern exposure; however, in the late afternoon, the sun shines directly in at a low angle onto the backside of the thermometer. Readings during that last 2 hours or so of sunshine are to be highly suspect of being high side. It hangs about 4 feet from the little office sized fridge and 3 foot tall mini-freezer that are my seed archive. They consume very little power and run intermittently, so I think that they are not much of an issue, but need to be assessed.

There is also a dark wood fence that DOES get some southern sun on it about 5 feet from the backside of the patio, so again we have a sunny bias to high side mid-day.

During the evenings / night, the proximity to the flagstone, and the house (stucco) ought to induce a mild warming bias. Basically, it takes a while to cool down masonry. There is also the obligatory BBQ (rarely used) about 20 feet away made of brick. This masonry might also slow the rise time in the morning. Also it is a fairly large and old ‘dial’ thermometer that may simply be wrong due to age, weathering, and the complete lack of calibration.

And I’m in the middle of a 20 mile wide basin that is Silicon Valley, so it’s a large UHI basin all by itself. Cars and pavement galore for a good long ways, then you reach the fringing mountains that keep it all in the bucket.

Given all that, I’d expect any reading I get to be biased high via the Urban Heat Island and specific locational issues.

San Francisco and SFO

Putting in “San Francisco, California” gave me SOMA San Francisco (South Of MArket street, near downtown San Francisco) while SFO gave the airport. SFO had a similar 6:53 AM reporting time as SJC, so I suspect there may be a ‘once an hour’ update near the hour. SOMA gave a current time. SFO is reporting 50 F (odd, another ‘whole degree F’, so I’m wondering if they always report whole F? And do they round up?) SOMA is reporting 49.8 F at 7:46 AM, so almost an hour later and much more ‘sun time’ and it’s still colder.

For those not familiar with the area, SFO is a few miles south of San Francisco and on the edge of the bay. (Some runway approaches are over water right up the end of the runway, so you used to get the ‘water ditch’ instructions on approach. But that seems to have mutated into a general ‘your seat is a floatation device’ nag recently.) SOMA is also near the water, just a bit further north, but much closer to a major urban area and generally down wind of it.

I would have expected SOMA to be warmer than SFO.

UPDATE: At 7:56 SFO updated to 52 F. SOMA is 50.7 at that same time. Monte Verde San Bruno (near SFO to the East) is 50.9 F so it’s looking to me like an easy way to spot about 1 F of AHI at SFO as the sun hits.


About E.M.Smith

A technical managerial sort interested in things from Stonehenge to computer science. My present "hot buttons' are the mythology of Climate Change and ancient metrology; but things change...
This entry was posted in AGW Science and Background, Favorites and tagged . Bookmark the permalink.

9 Responses to An Easy Airport Heat Island Audit?

  1. E.M.Smith says:

    LAX is 61 F while “Downtown LA” is 59.3 F. So we have a very large UHI location in “Downtown” that’s 1.7 F cooler than the nearby Airport. And the airport is from about 1/2 hour earlier at 7:53 AM.

    I think I’m seeing a pattern here…

  2. Mike Davis says:

    12:35 EDT! Airport at Knoxville 73f Thorn Hill 73F
    White Pine Elementary school 73f
    All locations overcast!
    I know when my flowers are going to bloom as the bloom 2 weeks later than in the city.

  3. Mike Davis says:

    With this area being the Ridge and Valley geographical formation in East Tennessee and Knoxville Airport being in Alcoa all 3 locations are some what rural with my site being just short of Forest, Well maybe not even short as the average home lot is in the neighborhood of 40 acres of forest with a small garden and house!

  4. E.M.Smith says:

    SOMA is running about 1/2 F below SFO in the afternoon too.

    Downtown LA is lightly warmer right now that LAX.

    Denver Airport (DEN) is running 86.9 F (partly cloudy) while Denver I-70 and Kipling Arvada is 86.2 ( ZIP 80002) FOR 0.7 warmer.

    It’s looking like the places that are sunny are ‘warmer at the tarmac’ (with some stochastic character) while full overcast is leaving things more even over an area (as a first guess).

    Looking at Phoenix it’s 90 at Sky Harbor airport (PHX) and 88.7 at (85001) on 36th and Camelback. 1.3 F higher.

    At Chico (CIC) municipal airport, it’s 58.8 F while downtown at ( 95926 ) it’s 60.1 F with winds 6 mph and some 14 mph gusts. (Thought the airport says calm with 8 mph gusts). So that puts the airport 1.3 F cooler. Oddly, Oroville Airport, another small muni airport, is 62 F. It’s about 14 miles away and ought to be nearly the same. There is darned little between them other than flat and farms. Red Bluff, 30 miles north is 59.4 F. All in all that CIC temp looks like an error of a small amount. Colusa, to the south west of Chico brings up this station:

    RAWS SAC NWR QD CA US, Princeton, California (PWS)

    at 62 F. No matter what way you go from Chico, it gets hotter than at the airport. Odd. Paradise, California is ‘up slope’ from Chico a short ways away and is at 56.5 F Wonder if we’ve got a bit of downslope from Paradise to Chico…

    Downtown Salt Lake City (84101) is 47.8 F while SLC is 55 F for a whopping 7.2 F bump! Wow. But the downtown is from about 50 minutes later in the evening that the airport. Still…

    So as a “reasonable thesis” it looks to me like on average airports are running a degree or so more, but that local specifics can override and make a city warmer than the airport. On averae, I’d expect a net heat gain of the airport over the nearby station, but for any given station, possibly even depending on who is under a cloud at the moment, it is possible for the airport to be cooler.

    The effect looks stronger in places with bright sun and low winds. Semi-rural airports under overcast have now or little delta temp. It will be interesting to see how things cool down in the evening.

  5. Roger Sowell says:

    Great post, E.M.! I live in Marina del Rey, just a mile or so up the coast from LAX. Our temperatures are usually just a bit lower than LAX, but not much really due to all the wind. It is almost always breezy here, (guess that’s why they built an airport…)

    But downtown Los Angeles is a different story. The measuring station presently is on the USC campus, in a grassy area but surrounded by parking lots, tennis courts, and a major street. It is also about 10 miles northeast of LAX. For those who are familiar with USC campus, it is located about 400 feet due West of third base at Dedeaux Field, the baseball stadium.

    For what it’s worth, here’s a website that gives the current plus previous 24 hours readings on airports from NOAA:

    One can simply edit the URL where “KLAX” exists, and input the last three letters for the appropriate airport, thus KSJC gives the info for San Jose International Airport. Note that the letters must be all caps.

    This might help to interpolate between readings where wunderground has a different time stamp.

  6. vjones says:

    You know, with some cleaned up historical data and a little bit of coding, it would be perfectly possible to quantify AHI in this way, taking account of things like wind direction and weather.

    It is the sort of problem wherein neural nets would be valuable for discerning multiple non-linear relationships. I have some experience with the value of using these for complex data modelling, but unfortunately only as an end user of a model created by someone else.

  7. oMan says:

    Highly interesting. Talk about a good application of citizen science and the creation of a “climate” of accountability. You could build a network with a few hundred individuals willing to spend a little dough on getting a “good enough” thermometer, and willing to spend a little time following standardized methods to install, calibrate, maintain it; and read/record/report what it says. Combine that with an organizer/curator/spokesperson function, and you could call BS on the “official” weather system if it misbehaved. Collect enough data over enough places over a long enough time, and you could vindicate Anthony’s argument that the current system is out of whack. Do this responsibly (with bulletproof methods and data) and pretty soon the data-cookers and hysterics will be discredited or at least forced to temper their claims and work a lot harder and more honestly to make their case.

    Wonderfully asymmetrical in its effect, possibly reshaping policy decisions worth hundreds of millions or billions of dollars. The net at its best. Thanks…

  8. Rod Smith says:

    Also, please DO NOT forget elevation.

    If you do a WunderMap of your local county (or other local area) you will find readings from many local thermometers. I haven’t counted, but for Okaloosa County, Florida I see a good dozen or more plus 4 large airports, Eglin and Hurlburt AF bases, Destin, and Crestview.

    Generally METARS from airports are taken just prior to the hour. And you can find on the net at NOAA the last 24 hours worth of Metars for many US stations.

  9. Mike Davis says:

    An Update: At 5pm EDT Knoxville Airport 88 and Thorn Hill 87.
    As far as elevation the Airport shows 962 and my GPS says 1045.

Comments are closed.