WF performance in the 'polar vortex' weather

I was curious if WF is getting any different behavior in the super-cold weather hitting the Minnesota and surroundings areas. I was poking around the map a bit today and it seems Winnipeg CA is about the coldest area currently according to the station reports both here (WF) and on the weewx station map (lots of station types there).

A family member back east texted me earlier saying he had a customer in Iowa who had four flat tires that seem to have cracked in the frigid cold there. Amazing.

1 Like

Im observing a WF station getting down towards -40F.
Thats 10F below the spec of -30F. The dewpoint did jump to 0
when it got below -40F, but still reporting RH.
Edit, Temp also went to 0 when below -40F.

Lots of very cold temps recorded across the network on 30Jan. Our systems have a sanity check in place to ‘not display’ values below -40F/-40C due to sensor manufacturer limitations. However, the raw data from a couple stations actually went past the -40F mark last night.

Here are some of the coldest obs:



You guys need to find a way to comp the ‘under minus 40 club’ something fun.

It’s pretty darn cold when Celsius and Fahrenheit meet.


Speaking about cold weather –

Over the last two days I have been observing the feels like and wind chill readings from my WF. These readings have been the same as my temperature reading. All 3 are reading 2.5 degrees F now.

I checked a few forecasts and they are showing feels like about 14 degrees lower. Is this a known problem?

@djcrupi This might be of interest:

1 Like

We are just being brushed by the polar vortex here, but I find it interesting that my WF, which is usually in lock step with my Rainwise MkIII for temperature, is reading 3 degrees colder than the RW at -9F. Maybe it’s because the WF is on the eave of the unheated shed closer to the ground. The RW is on the eave of the unheated porch.

1 Like

The humidity in my Airs is reading right on with the SHT-31 in the Davis VP2. Temperature-wise, they all agree more-or-less…that it’s COLD! :rofl: One of the Airs hit -9.9 F, the other -8.7 F, and the VP2 reported -8.8 F…


My humidity is always at least 10-15% off from my VP2 on both my Airs. My VP2 is only about 6-7 months old.

FWIW, my VP2 is less than a year old…and newer than one of the two Airs…

1 Like

I’ve always noticed that humidity is extremely slow to respond to actual changes. Often a value seems to just sit there forever despite actual changes that are taking place. I mentioned the age of my VP2 only because there seems to be concern that the SHT-31 drifts over time as many of these humidity sensors tend to do.

1 Like

Before I posted my question I checked both NOAA and the NWS and there calculations were indicated to be based on 3 mph wind speed. Also, according to NOAA’s site, it appears that while the wind is measured at 33 feet, it is then calculated at 5.5 feet to simulate the face of anaverage height person.

Maybe I’m misunderstanding something.

Hi @djcrupi, I was quoting from this post by a member of WF staff: Frozen Sky Devices?

As they said there isn’t just one definition of windchill, but I know that WF use a cutoff of 5 mph. So if all your temperature and feels like reading are the same, then the 1-minute avg wind speed must be below 5 mph. If it isn’t then it is probably worth getting in touch with WF support so they can help troubleshoot.

Wind chill on the WF system should be calculated when the temp is 50F or below and the 1min avg wind speed is above 3mph. Exact formulas for derived metrics (when multiple sensed parameters are used in calculation) can be seen here:

On a related note, Heat Index is calculated only when temp is 80F or above AND RH is 40% or greater. Again, refer to the derived formulas on the page noted above.

The “Feels Like” value = wind chill or heat index when relevant…otherwise it’s the temp.

We’ll take a look to ensure the APIs are calculating correctly. Thanks for the heads up everyone! Stay warm.

1 Like

Apologies! I was confusing miles per hour and kilometres per hour. 3 mph is approximately 5 kph.

Thank you, that matches what I have read.

05 Feb '19 Update: Thanks to you guys/gals, we found the code for our API was calculating Wind Chill beginning at 5mph instead of 3mph. This has been corrected to begin calculating wind chill when avg wind meets or exceeds 3mph. Thanks!


I wasn’t going crazy after all! Glad you managed to find the bug, now I just need to update the PiConsole as it is also calculating wind chill currently with a cutoff at 5 mph.

That’s great, thank you.