Author Topic: ASOS/AWOS Accuracy  (Read 3470 times)

0 Members and 1 Guest are viewing this topic.

Offline drew1021

  • Senior Contributor
  • ****
  • Posts: 229
Re: ASOS/AWOS Accuracy
« Reply #25 on: December 15, 2023, 07:12:23 PM »
I had the same scenario at our local ASOS station KINT. However the bias was on the positive side.
It showed up in late winter of 21. Up until then my readings were pretty much spot on with the site. Afterwards whatever they did there was a 1-2 degree Fahrenheit bias to the plus side. I let the techs know and they were going to check it but they didnt. About a month or two ago its been reporting much more consistently compared to local pws and airport 17 miles east. Still seems to rise quite rapidly especially on sunny days. Guess its possible they moved the ASOS assembly.
VP2 with 24 hour FARS. WU: KNCLEWIS2. CWOP/APRS: DW4712, COCORAHS: NC-FR-7

Offline Buick

  • Senior Member
  • **
  • Posts: 99
    • KWABELLI466
Re: ASOS/AWOS Accuracy
« Reply #26 on: December 15, 2023, 07:42:53 PM »
I hadn’t considered that the KBLI ASOS had been moved. It still appears in the usual location on the current Google satellite imagery though that may change. On the other hand, my station is 2mi away and always reported very similar if not identical avg temps so it’s kind of hard to imagine that just moving it to another location on the airport property would have any effect on average. I checked the KBLI data for November and December 2022 against my data and the monthly avg temp at my location was 0.1F colder than at KBLI during November, while the monthly avg temp at my location was identical to that of KBLI during December. During November 2023 the monthly average temp at my location was 1.1F warmer than at KBLI and for the first 14 days of December 2023 the monthly average temp at my location is 0.9F warmer than at KBLI. I’d wager that by the end of the month it’ll be 1.1F warmer on avg.

It’s also hard for me to believe that two sites 2mi apart which don’t have any substantially different characteristics that would contribute to differences in sensible weather would have average temp differences of ~1F. KSEA has an average November temp 2F warmer than KBLI and is much more affected by urbanization. Its average annual temp is 2.5F warmer than KBLI. Anyone with local knowledge of interior Western Washington knows that that 2.5F makes a world of difference in terms of the sensible weather and climate of those locations. On the one hand, KSEA has essentially the prototypical climate of interior Western Washington and on the other you have Bellingham, the icebox of the region. 2F on average is a big differenceand half of a big difference is still not even close to small.
« Last Edit: December 15, 2023, 07:56:05 PM by Buick »
VP2 (6163) w/ 7345.425 & 24hr FARS
6410 @ 26' AGL, 12' above rooftop
6415 @ 12' AGL
7714
Rainwise 111
Stratus
WiFiLogger2
WeatherLink Live

Offline DaleReid

  • Forecaster
  • *****
  • Posts: 2002
    • Weather at Eau Claire, WI
Re: ASOS/AWOS Accuracy
« Reply #27 on: December 15, 2023, 08:07:01 PM »
The fact that  you've seen relatively close tracking of your numbers vs. the airports is interesting, but I'm of the opinion that a 2 mile difference is huge and you were lucky to see those values be as close as they historically have.

I know with my less than idea siting (you can only do so much on 5 acres with trees and buildings the microclimate changes are significant.  Airport siting of the stations is very controlled and trees and such are some distance away.

If  you have blacktop roads, trees, water and buildings, a predominate breeze over them can sure bring temperature changes.

If your station was on an ideal siting location and the ground prep and nearby areas was exactly the same, I'd still be surprised to see close tracking.

I couldn't find the paper on line, but a couple years ago I read a university research level study of an airport with an siting that met the published standards, and then something like 4 or six temporary stations with identical calibrated sensors scattered about the location.  There was often a 2 or 3 degree temperature difference (my recollection is poor for precise numbers) depending upon time of day, sun angle and wind direction which were predictable and repeatable.

I think your experience borders on the same experiences of those with multiple rain gauges (even COCORAHS (sp?)  types) with variations in tenths of an inch even on the same plot of land.

I would need to review your initial comments about your station, but unless it is Vaisala, RM Young or Campbell Scientific, calibrated and recalibrated sensors, there is far from any expectation of being closer than a couple of degrees.

I think your experience of having 'close' following was a very fortunate circumstance, and not something that would be predictable.  And frankly with hobby grade stations, the weak link is our stations, not the commercial research grade instruments at FAA installations.

My two cents worth again. Dale.
ECWx.info
&
ECWx.info/t/index.php

Offline Buick

  • Senior Member
  • **
  • Posts: 99
    • KWABELLI466
Re: ASOS/AWOS Accuracy
« Reply #28 on: December 15, 2023, 08:20:09 PM »
As I’ve said, the issue isn’t that a difference exists, it’s that a strangely predictable difference has materialized over the past couple of months where no particular difference has existed for nearly a decade. Nothing has changed about my site and as far as I’m aware nothing has changed about the KBLI site. It’s interesting that you mention discrepancies between multiple rain gauges as I run 3 at the same location: the VP2 tipping spoon gauge, a Rainwise 111 and a CoCoRahs gauge as a reference. The Davis gauge is clearly inferior and nearly always underreports but the Rainwise gauge is excellent and, on average, agrees pretty much perfectly with the CoCoRahs - the agreement between the two for the month to date is 99.5%. The thing about the 2mi distance that’s relevant is that there are no physiographic features sufficient to produce any kind of actual different microclimates. Obviously two different locations 2mi apart may frequently have different weather conditions but over long periods the average conditions should be essentially equal if all else is held equal and in this case all else is held equal.

For what it’s worth, I’m pretty meticulous about maintaining my equipment. I replace my temp/hum sensors annually and I think you may be underestimating the quality of Davis’ products. Obviously one can spend as much as one cares to spend on weather measuring equipment but there’s a rapidly diminishing return on accuracy once prices get well into the 4 figures and beyond. Would I gladly accept a gift of a USCRN observing station? Yep. As long as it comes with tech support

I also run two SHT31 temp/hum sensors at the ~same location but in different radiation shields and they nearly always agree on monthly avg temps to within 0.1F. The fact that theyre in different shelters but report the same conditions on avg is a good indicator that QA/QC is very good wrt to precision. WRT accuracy, this indicates that they are either equally inaccurate, which is certainly a possibility, but I think the likelier possibility is that they are not only very precise but also very accurate because it seems more probable that they both read accurately than that they both read inaccurately but equally so. Theres more reason to assume the former than the latter because if it was true that they were both inaccurate it would be more improbable that they would report similar values than it would be if they were both accurate because inaccuracy is a defect so its more reasonable to assume that defects are associated with other defects.

That is to say, this defective sensor which reports inaccurate values is equally defective compared to this other sensor which reports the same inaccurate values.

Or: both of these sensors are accurate and thats why they report the same values.

Intuitively, one clearly has a higher probability of being true, all else equal.
« Last Edit: December 15, 2023, 08:54:18 PM by Buick »
VP2 (6163) w/ 7345.425 & 24hr FARS
6410 @ 26' AGL, 12' above rooftop
6415 @ 12' AGL
7714
Rainwise 111
Stratus
WiFiLogger2
WeatherLink Live

Offline CW7491

  • Senior Contributor
  • ****
  • Posts: 205
Re: ASOS/AWOS Accuracy
« Reply #29 on: December 15, 2023, 09:34:14 PM »
Doesnt correspond to the shift youve seen, but it looks like they did move the ASOS at KBLI in November 2021. Google maps shows it at the new location.

https://www.ncei.noaa.gov/access/homr/#ncdcstnid=20027993&tab=LOCATIONS

Offline Buick

  • Senior Member
  • **
  • Posts: 99
    • KWABELLI466
Re: ASOS/AWOS Accuracy
« Reply #30 on: December 15, 2023, 09:40:42 PM »
Awesome info! Really appreciate the knowledge I can glean here. The first time I actually decided to locate it it was in its current location.
VP2 (6163) w/ 7345.425 & 24hr FARS
6410 @ 26' AGL, 12' above rooftop
6415 @ 12' AGL
7714
Rainwise 111
Stratus
WiFiLogger2
WeatherLink Live

Offline CW7491

  • Senior Contributor
  • ****
  • Posts: 205
Re: ASOS/AWOS Accuracy
« Reply #31 on: December 15, 2023, 09:42:02 PM »
And looks like there is a plan to potentially move it again see page 6-45 figure 6-21

https://www.portofbellingham.com/DocumentCenter/View/8962/Bellingham-Master-Plan-2019

Offline Buick

  • Senior Member
  • **
  • Posts: 99
    • KWABELLI466
Re: ASOS/AWOS Accuracy
« Reply #32 on: December 15, 2023, 09:52:38 PM »
Wow, man, the information you have is a resource to be treasured. Thanks, again!. Im not totally surprised that the ASOS is being relocated as theres been considerable construction at KBLI over the past year.
VP2 (6163) w/ 7345.425 & 24hr FARS
6410 @ 26' AGL, 12' above rooftop
6415 @ 12' AGL
7714
Rainwise 111
Stratus
WiFiLogger2
WeatherLink Live

Offline Buick

  • Senior Member
  • **
  • Posts: 99
    • KWABELLI466
Re: ASOS/AWOS Accuracy
« Reply #33 on: December 15, 2023, 10:00:41 PM »
In any case, theres absolutely no reason that a wx station located anywhere on KBLI property should have noticeably different avg temps than at my location especially given the history of close correspondence. I mean, if you had adjacent locations with substantially differing mean temps youd expect corresponding wind velocities due to the pressure difference driven by the temp difference. So given the proximity of the locations in question youd expect to see nearly constant offshore wind regardless of how stable weather conditions were because the pressure gradient driven by the temp difference would be more-or-less unrelenting. This is most certainly not the case wrt to the locations in question and in fact the opposite prevails.
VP2 (6163) w/ 7345.425 & 24hr FARS
6410 @ 26' AGL, 12' above rooftop
6415 @ 12' AGL
7714
Rainwise 111
Stratus
WiFiLogger2
WeatherLink Live

Offline CW7491

  • Senior Contributor
  • ****
  • Posts: 205
Re: ASOS/AWOS Accuracy
« Reply #34 on: December 15, 2023, 10:12:20 PM »
Yeah, while I certainly understand the tolerance logic, its still crazy to me that the NWS will go out to the ASOS with a reference thermometer and call it good if its within 5F! Wed be pulling our hair out.

Offline Buick

  • Senior Member
  • **
  • Posts: 99
    • KWABELLI466
Re: ASOS/AWOS Accuracy
« Reply #35 on: December 15, 2023, 10:19:21 PM »
No shizz! Having recently dealt with a temp/hum sensor that was not reading as expected by an amount that might seem perfectly OK to the non-enthusiast, I put a not inconsiderable amount of effort into identifying and rectifying the problem. I knew from experience that what was occurring was anomalous and indicative of a problem due to its persistence and predictability.
VP2 (6163) w/ 7345.425 & 24hr FARS
6410 @ 26' AGL, 12' above rooftop
6415 @ 12' AGL
7714
Rainwise 111
Stratus
WiFiLogger2
WeatherLink Live

Offline Buick

  • Senior Member
  • **
  • Posts: 99
    • KWABELLI466
Re: ASOS/AWOS Accuracy
« Reply #36 on: December 17, 2023, 03:50:21 PM »
Decided to reach out to Cliff so well see if he deigns to respond...
VP2 (6163) w/ 7345.425 & 24hr FARS
6410 @ 26' AGL, 12' above rooftop
6415 @ 12' AGL
7714
Rainwise 111
Stratus
WiFiLogger2
WeatherLink Live

Offline CW7491

  • Senior Contributor
  • ****
  • Posts: 205
Re: ASOS/AWOS Accuracy
« Reply #37 on: December 17, 2023, 10:25:16 PM »
Let us know if he does

Offline Buick

  • Senior Member
  • **
  • Posts: 99
    • KWABELLI466
Re: ASOS/AWOS Accuracy
« Reply #38 on: December 19, 2023, 11:16:28 AM »
No response, as expected
VP2 (6163) w/ 7345.425 & 24hr FARS
6410 @ 26' AGL, 12' above rooftop
6415 @ 12' AGL
7714
Rainwise 111
Stratus
WiFiLogger2
WeatherLink Live

Offline TheBushPilot

  • Senior Contributor
  • ****
  • Posts: 151
  • Calibration Technician
Re: ASOS/AWOS Accuracy
« Reply #39 on: December 19, 2023, 12:40:19 PM »
Yeah, while I certainly understand the tolerance logic, its still crazy to me that the NWS will go out to the ASOS with a reference thermometer and call it good if its within 5F! Wed be pulling our hair out.

5F???

That's insane. Why the lax tolerance?
"There is always more than one thing influencing anything we are trying to measure." ~ Sherman Fredrickson
"Do it right or don't do it at all."

Met Instruments Project
Central Chasers

Offline CW7491

  • Senior Contributor
  • ****
  • Posts: 205
Re: ASOS/AWOS Accuracy
« Reply #40 on: December 19, 2023, 03:09:29 PM »
Agreed. It doesnt make much sense. The Washington Post article I took a screen shot of earlier in the thread cites that the tolerance of the ASOS and the psychro-dyne reference thermometer to be +/-1.8F but that the ASOS needs to only be +/-5F. Seems like +/-3.6F would make more sense. The ASOS users guide does list the max error of the ASOS temperature sensor at +/-1.8F. I guess because its just a validation procedure?

If you look at the dewpoint validation, its even more lax.
« Last Edit: December 19, 2023, 03:12:48 PM by CW7491 »

Offline Buick

  • Senior Member
  • **
  • Posts: 99
    • KWABELLI466
Re: ASOS/AWOS Accuracy
« Reply #41 on: December 19, 2023, 04:30:35 PM »
It does sound a bit like if your local ASOS reports data that is actually accurate and not just "within tolerance" you're kinda lucky...
VP2 (6163) w/ 7345.425 & 24hr FARS
6410 @ 26' AGL, 12' above rooftop
6415 @ 12' AGL
7714
Rainwise 111
Stratus
WiFiLogger2
WeatherLink Live

Offline DoctorKnow

  • Forecaster
  • *****
  • Posts: 1984
Re: ASOS/AWOS Accuracy
« Reply #42 on: December 19, 2023, 04:48:18 PM »
KRAH (RDU) is supposed to be relocated because it is so far out, it's ludicrous. They have the jet planes blowing hot exhaust on the thermometer it is so close to the runway. Last summer it was the hottest thermometer that you could find in the county, way off.

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 6364
    • Valentine Nebraska's Real-Time Weather
Re: ASOS/AWOS Accuracy
« Reply #43 on: January 19, 2024, 12:11:09 PM »
My experience was the local ASOS ran high (+1.8f) and this continued for almost 6 years before they finally changed the temperature sensor out. I have a psycho-dyne reference thermometer and stood right by the ASOS and confirmed the error was there multiple times. Wish I had known who else to complain to at the time, so just blogged about it on my website.  One of the local North Platte meteorologist read the blog (I'll leave name off) so he convinced the tech to change the temp sensor because the max RH for KVTN over the past 6 years had never exceeded 93% so once the sensor got changed, BAM both issues fixed. Having the air temp read +1.8 prevented the RH from ever exceeding 93%.
Randy

Offline ASOSWX

  • Member
  • *
  • Posts: 11
Re: ASOS/AWOS Accuracy
« Reply #44 on: January 19, 2024, 12:41:17 PM »
That 1088 probe is checked against a reference on the transmit logic card and that temp is very accurate. When going to the 1088 upgrade years ago the 1088 was kept in place for an a/t backup sensor.  That muffin fan would build up dirt and fail.  On a side note, I did have beers at Peg Leg Brewery in N Platte.  Great Place.

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 6364
    • Valentine Nebraska's Real-Time Weather
Re: ASOS/AWOS Accuracy
« Reply #45 on: January 19, 2024, 01:13:39 PM »
I don't know all the specifics of why it wasn't caught going on 6 years but do know the FO contacted me soon after posting this back in June 2022 about it getting fixed and personally told me the ASOS was fixed. I can confirm as of today it works great. 
So anyone out there who suspects your local ASOS isn't running up to snuff check the RH, if it can't exceed 93-94% even in soupy fog there is an issue.

 [ You are not allowed to view attachments ]
Randy

Offline CW7491

  • Senior Contributor
  • ****
  • Posts: 205
Re: ASOS/AWOS Accuracy
« Reply #46 on: January 19, 2024, 11:22:37 PM »
The MEM ASOS temp/dp is inoperative. Looks like an aspirator fan issue. The local WFO says the temp/DP is being manually augmented. ASOSWX, do you know how this is done and what instrumentation they use? [ You are not allowed to view attachments ]

Offline TheBushPilot

  • Senior Contributor
  • ****
  • Posts: 151
  • Calibration Technician
Re: ASOS/AWOS Accuracy
« Reply #47 on: January 20, 2024, 01:01:41 AM »
Don't they use the OID interface?

https://www.weather.gov/media/asos/appen.pdf
"There is always more than one thing influencing anything we are trying to measure." ~ Sherman Fredrickson
"Do it right or don't do it at all."

Met Instruments Project
Central Chasers

Offline ASOSWX

  • Member
  • *
  • Posts: 11
Re: ASOS/AWOS Accuracy
« Reply #48 on: January 20, 2024, 10:07:07 AM »
CW7491,
The FAA will use data from their SAWS or AWS system.  I don't know "brand" of sensors. at a small airport or have a Contract Weather Observer (CWO)  they may use a portable device.  @ bush pilot. The OID is simply a dumb terminal where they see the ASOS data and augment (signed in) if needed hence no "auto" in the ob.  The old CRT's are long gone and thin client now.

Offline ASOSWX

  • Member
  • *
  • Posts: 11
Re: ASOS/AWOS Accuracy
« Reply #49 on: January 20, 2024, 10:17:03 AM »
They are on a PMS like all FAA equipment, so I'm sure it's checked at least once a year. If it is off, this could potentially be critical with pilots now that icing conditions are becoming more prevalent.

There are quarterly, Semi-annual, and Annuals on the System.  The DTS1 is sent in every 18 months for calibration.

 

anything