Cool - my udp listener syslogged the unexpected messages and their content…
Sep 19 12:29:42 zero listen.py: [wf-udp-listener]: starting to process messages
Sep 21 03:00:49 zero listen.py: [wf-udp-listener]: unknown data type in {"serial_number": "HB-00010412", "type": "calibration", "value": "Air Calibration received"}
Sep 22 01:00:55 zero listen.py: [wf-udp-listener]: unknown data type in {"serial_number": "HB-00010412", "type": "calibration", "value": "Air Calibration received"}
Sep 23 01:00:49 zero listen.py: [wf-udp-listener]: unknown data type in {"serial_number": "HB-00010412", "type": "calibration", "value": "Air Calibration received"}
Still seeing nightly events like this, just after midnight local time.
Can somebody from @WFmarketing explain what Air calibrations you’re sending nightly ?
Or are these just dry-runs exercising your system to get ready for whenever you do the Sky calibrations for real ?
Hi @vinceskahan Every day our Continuous Learning (CL) system is comparing your station data against more than a dozen other observations within a couple kms of your location. Right now the CL system is enabled for temp and humidity. If the comparison values meet an R-squared QC threshold, a nightly calibration is applied to ensure that your sensor data is always spot on. And yes, this is just the tip of the iceberg.
FWIW, my WF temp+humidity are basically exactly on the money vs. my VP2 that sits 20 feet from it, as long as the VP2 isn’t in bright sun. The VP2 seems to read a couple degF high when in the sun, perhaps the shield is losing its effectiveness after almost 10 years up.
Still thinking a bit re: whether I should remount the Air to a location next to the VP2 to aid in comparison.
I wouldn’t expect the wind to match between the two, as the WF is 8’ lower and about 50’ from the VP2. The VP2 should be far more accurate for our prevailing wind patterns from the S/SW/W given its elevation, (that is, once my replacement 6410 sensor arrives from Davis in a day or two and I get the chance to put it up).
Given your problem with vibrations causing false readings, I decided to try to go for the most stable mount for the WF rather than trying to get higher elevation to get truer wind readings. Unfortunately I don’t have a good way to colocate everything and get a good elevation for wind readings that wouldn’t cause a ‘bird perch’ opportunity that would ‘really’ honk things up.
Interesting … as my Air is in my Living Room reading 20C when outside air temp is 3C, is that going to start causing you (or I) problems with totally bogus readings?? I was gonna move the Air outside after I set it up, but never got around to it, and right now its more interesting monitoring my indoor temp!
Hi @dave . Good question. As @vreihen mentioned, the ‘indoor’ designation flags the data and disqualifies it as valid outdoor data. In addition, our CL system will compare the data from your AIR against other outdoor observations in your immediate area. Once it recognizes that there is no correlation and the R-squared values are way below the QC threshold, the automated daily calibrations are not applied. Putting the ‘smart’ in weather.
