Tempest Calibration

So I’ve mounted the Tempest in close proximity to the Sky and am getting significantly lower wind speeds. We’re currently having some gusty winds and the Tempest is consistently lower by at least 25%. I’ve even tried switching the 2 locations to rule out location bias and the results are precisely the same. I’m not sure if this is something a calibration can take care of, but it is a concern. My Sky was always in good agreement with my Davis VP2 for moderate to high wind speeds, but the Tempest is not in agreement with either the VP2 or the Sky.

The easier issue is a barometric pressure that’s high by about .04". I thought that perhaps there would be a calibration on that last night, but there wasn’t. Temp & RH seem good and the passive radiation shield seems to be doing a decent job with temps in the sun about 1 degree higher than a shaded location.

It will take a few days to adjust itself.

1 Like

Gary, I’m sure calibration is the case with barometric pressure, but is that also true with initially low wind speeds relative to the Sky and VP2?

There was quote from quote along time back in this topic from WeatherFlow about the sonic sensors requiring a little time to “settle in” once a SKY module has be installed. I wonder if the same is true for Tempest?

Once installed at a specific location (“in-situ”), the AIR and SKY devices will go through a range of calibration routines. The routines for the sonic anemometer for example actively evaluate the sine / cosine curves and frequencies to ensure the sensors are capturing the right sonic wavelengths and nothing has been jarred during shipping. If amiss, your SKY will recalibrate and capture corrections in the onboard firmware (the sonics are so sensitive that a change in gap height or angle the size of a human hair can throw things off).

3 Likes

Peter, thanks much, this hopefully will be the procedure to bring the wind speeds up. I’m less concerned about the barometric pressure. I’m confident that will calibrate over the next few days.

Well I was wrong about the barometric pressure calibrating over a few days. I placed the Tempest in service on Monday and the pressure is still high by .05” relative to my Air, VP2 and other WF devices within my area. I have no evidence that any calibration has occurred, including my lower wind speeds relative to my Sky & VP2. Nearly a week in I would have thought that would have happened.

The calibration process regarding pressure ran twice for your Tempest but was rejected as not good enough. That I can see … why is another story that only tech folks can tell. So yes it is running but for now the results don’ pass QC. Might be next run … and no, I have no clue when …
On the other hand humidity seems to have adjusted a tat.

Regarding wind … the calibration is in Tempest and not server side for as much I know and shouldn’t take that long. It needs wind from all sides at least once to auto calibrate the 4 sensors …

Here are your 2 devices and remember the scale isn’t perfectly the same but overall I find them close. Even a few feet can influence … mine are about 1 meter apart and nearby trees influence enough to have different readings (sometimes wind direction as it creates turbulence). There are difference but overall those look pretty nice to me.


1 Like

Eric, thanks so much for that! :slight_smile:

I have noticed the winds are definitely in closer agreement now. I’ll have to wait for winds with a northerly component because that’s where I had the issue. The S/SE winds seem to be fine and as you say, it needs to see winds from all quadrants.

Sunny, in my case I’m only inquiring in the event Tempest wind speeds continue to be lower than my Sky. At that point I’d see if recalibration would yield a different result.

that’s interesting. Are you sure it isn’t just the different location of each unit? Give it a few more days to do some self calibration, if it still is a problem, try swapping them.

Yeah, interestingly I did just that, swapped locations with the same results. No matter where I put the Sky relative to the Tempest, it records higher wind gusts, often by 25%. It appears that the higher the gust the greater the disparity. All other readings look excellent, especially since my barometric pressure finally calibrated yesterday.

I’ll give it a few more days and if there are no changes I’ll try the old on-off switch. :slight_smile:

1 Like

higher wind gusts is something different than higher winds. Check the sampling rate. It might be that the tempest is sampling much faster, and therefore is capable of getting higher gust.

It’s the other way around. The higher gusts are on the Sky, not the Tempest. Did they increase the sampling rate with the Tempest? I haven’t noticed that. Both seemed equally fast in updating.

A lower sampling rate probably explains the lower gust you see. The tempest apparently scales down the sampling rate depending on battery power. That might explain it. with Voltage <= 2.415 or >= 2.41 the sampling is reduced to 6 seconds. Lower than that and it switches to a 1 minute interval. Lower than 2.3 and it goes to 5 minutes.

Sunny, I’m seeing the very same quick sampling rates on both. As I watch both updating live, they’re both updating at least at 2 second intervals. Both are very fast. I see no difference between them at all. Voltage on my Tempest is 2.55v. So really no sampling rate issues.

then I don’t know. (fastest sampling an sky is 3 seconds. I don’t know about the tempest)
you can find out about the sampling rate using the “</> API” button on top of the community web page (this page). Perhaps there are better ways but, go there select rest. click authorize. Get the stations id of the tempest, and use that id to get the observations from that device. (there are examles but just click the try it button)

Sky And Tempest have the same sampling speeds, except Tempest has automatic sampling reduction with voltage compared to the toggle for Sky

Again 2 charts of both your stations and again, I don’t see such great differences. What you can see, wind directions are all over the place indicating or bouncing or swirling due to nearby objects (walls, trees …). Also sampling between 2 stations isn’t on the exact same moment, you can have 1 or 2 seconds between them and that can be largely enough to have other readings and bearings.


2 Likes

Eric, it seems average wind speeds are very similar but the consistency is the Sky usually records higher peak gusts than the Tempest. A 9mph gust on the Tempest usually shows up as something like a 11-12mph gust on the Sky. I’m not sure how this will translate to much higher wind gusts during more windy conditions. It’s probably more by anal. personality than anything.

BTW, I do think our winds are truly light and variable this morning.

I’ll see what happens in truly windy conditions.

1 Like