Wind, vibration and wrong rain readings

Hi all,

My Tempest is currently driving me crazy :rage: I had it mounted on an aluminum pole (~1.6m high) on my flat roof but on windy day, I got rain reports all the day whereas the sky was blue. I went up on my roof and notice that the pole was indeed vibrating. So, I cut it by ~40cm, filled it with sand and put some kind of rubber like double sided tape around the pole, inside the parasol stand.

I was pretty sure this would stop those false rain readings caused by wind. Indeed it improved but I still see a correlation between wind gusts and rain rate. I see many people with Tempest on metal poles and I really wonder how people do not get wrong rain reports.

I wanted to have Weatherflow calibrate the rain sensor with my data (I have a rain gauge next to the Tempest as you can see on the picture) but until the false readings are solved, it doesn’t make sense.

I’ll try with a PVC pole filled with sand to see if it is better than metal. It’s crazy how this thing is sensitive to vibrations. And Weatherflow barely mention it in the documentation. I really doubt that the pole in concrete example shown in https://help.weatherflow.com/hc/en-us/articles/360047672153-Example-Installations would withstand more than 20km/h wind without reporting rain.

I know there is already many post regarding this issue but this issue makes the rain sensor totally unreliable for me (+ the fact that it does not detect light rain).

If anyone has a (simple and nice) solution that will work for sure, please, tell me. Or anything that could be 3D printed would work for me too.

Another question. Is there a way to test if the setup would report false rain? Like tapping the pole, shaking it, etc. Waiting for wind gust of over 30 km/h to see if things improved is a bit painful :roll_eyes:

Stiffer isn’t necessarily better. I have my units mounted on vinyl wrapped wooden closet rod poles that extend at least 2m above the top mounting point and I only get false rain when a bird lands on them. They do wobble in the wind but at too low of a frequency to be detected as rain.

Had the same thing happen yesterday on mine, when winds were gusting in the mid 20’s. Reported rain, but there was no rain.

I just got a PVC tube. I will give it a try this week-end. Probably will fill it with sand. There is not much wind currently but some gusts around 20-30 km/h should be back on Sunday.

I just don’t understand why not everyone is complaining about this. I see many pictures of Tempest on high metal poles and I just can’t imagine they do not generate false rain readings on windy days.

I have a galvanized steel pole of 3 meter in length.
I put a flap of rubber between the steel pole and the PVC pipe.
The PVC pipe is filled with sand. It’s wobbling in high winds. But not in that frequency that it detects false rain.

1 Like

I have my Tempest mounted on the highest eave of the house, with a six foot vinyl wrapped wooden pole (part of which is braced against the eave). I normally have no issues with false rain reports, but yesterday we had wind gusts all day of up to 38 mph with clear blue skies. It wound up reporting 75 minutes of “rain” accumulating 0.09 inches, which was all imaginary. Aside from a clear windy day with high wind gusts, it’s been pretty accurate. It’d be nice if I could just tell it that no really it’s just windy (as it can already detect) and I promise there’s no rain.

1 Like

That is definitely something many of us have wanted to be able to do. Here is a feature request for it [API] Delete Selected Data. Go vote for it if you haven’t already.

One thing I have wondered about is if wrapping a wire or something similar around the length of the pole, the same way antennas on cars are wrapped or otherwise have a spiral on them which reduces the vibration and wind noise they would otherwise create, would also reduce false rain readings.

I am getting false rain reading too. Here’s my set up. 15’ off the ground. Metal pole is definitely oscillating in 15*mph winds.

This weekend I’m going to run two guy lines to the tops of my other fence posts to stabilize the pole.

This is the first time I’ve had decent wind since installing the unit, and the first time for false rain readings.

Hopefully stabilizing the pole will solve the bad data. (I also think my wind speed readings are being effected too)

Screenshot of rain readings just before the station photo was taken

There is also a good possibility that the fencing is creating vibrations. I found that my chain link fence was a source of false rain. The same pole not mounted to the fence eliminated the false rain.

If the mount between the pole and the tempest has no wobble then I would expect that the vibrations can be heard by placing a screwdriver onto the pole at the tempest and your ear against the handle of the screw driver.
If you hold your pole and give it rapid jiggling movement for a couple of minutes see what it records. Then lift the mount up or the pole out of the mount and do the same.
Holding the pole free of its mount and giving it rapid shaking in my experiments would never cause false rain, so if yours does then it might be the connection between tempest and the pole. So remove the tempest from the mount and shake it. If it records rain with you shaking the tempest i would contact support.
I also put silicone tape around the join between mount and tempest to reduce possible movement in extreme wind.
I have also held my sky on a pole out the top of a car travelling and it was strange how it could have no rain but once it begins detecting rain I suspect that the algorithym detects rain more easily. I did once report to weatherflow how comparing the detection of rain compared to actual rain once it begins reporting rain it becomes more sensitive and continues reporting after the rain stops. But they probably fixed that. I suspect the same might happen with false rain due to vibrations. Once it commences it might require a rest from any vibrations to stop thinking that it is raining.
I suspect the vibrations are coming from the pebbles on the roof. In this situation I would try placing your stand into a sack of sand to remove any contact with the pebbles. But I would also sure up the way the pole is held into the stand by filling it with expandable foam or similar idea to remove any metal to pole contact.
You need to try to stop sound travelling to your pole. Remember as a child perhaps you made a string telephone with two tin cans. Now imagine how you can stop that sound travelling along the string…by including shock absorbers or dampeners which can also include heavy objects like lead or timber or cement or bags of sand.
Cheers Ian :slight_smile:

1 Like

mine is on an acrylic pole. It wobbles, hardly any false rain from wind, but all this wobbling probably does make the wind measurements less accurate. On average it should be ok, but the wind gust and lul values are possibly effected.

Thank you all for your reports/suggestions/comments :smile: I just replace my aluminum pole with a PVC one, filled with sand. Put some layers of tape at the top so that the Tempest has no chance of moving around the pole. It is going to wobble with wind for sure but let’s see if this is better than vibrations/false rain detection.

I also thought about that. Something like a “door seal” glued along the pole in spiral that could “break” the vortex generated by the wind around the tube. I was also wondering if a square pole would be any better than a round one!?

I noticed that too. Like when my Tempest was installed on a 160 cm aluminum pole without sand, it started recording false rain in the morning when the first gust came and didn’t stop for hours even if the wind was no longer as strong as before. Could be that the pole was still vibrating but I really had the impression that the detection was more sensitive once it started reporting (false) rain.

I removed the pebbles before installing the parasol sand. I just check again before and it is very stable, I don’t expect any issue from there. And I doubt the pebbles are getting so much wind that they can induce vibration into the pole. But I’ll try to remove the ones around the stand if I still get false rain detection.

1 Like

I’m sure that is the case as rain isn’t “detected” immediately, it has to continue for some time, 5 seconds maybe, before it is determined not to be false. Maybe @dsj can weigh in on this.

If you want to conform to ‘official’ weather recording standards there are lots of specifications. For example they say the standard for the height of the thermometers is between 1.25 m (4 ft 1 in) and 2 m (6 ft 7 in) above the ground.
Sometimes I think I should get a 10m pole to measure wind strength for a small turbine :wink:

Thanks definitely a possibility.

I am trying to figure out a stand-alone location where I can get it high enough and also set up some guy lines so the pole doesn’t wobble.

It’s a problem I’m having a hard time thinking through.

I wish I could annotate the graphs to indicate erroneous data readings for historical and Weatherflow data hygiene.

Hi @Hungrysquirrel,
I have had several pole with guy wire setups over the years. This current design uses a U bracket with a bolt through the base of the pole so that it can tilt over by releasing some of the supports and guy wires. And for the guy wires I use turnbuckles with a hook on one end so that I can adjust the tension and then unhook to release the wire easily to tilt the pole. And in this pic you can see I have the U bracket base down the slope of the roof which allows the pole to tip towards the apex of the roof so that I can work on things at an easy height when the pole is tipped over for maintenance.
The windsock creates extra vibration which can cause false rain in extreme winds but I can prevent that from happening using my vibration resistant mount seen in another thread. However the Tempest is not suffering as much vibration induced rain as the Sky does because the Tempest has better internal vibration resistance.
The top of this pole gets a large swing on it due to the wind sock but handles over 100kmh fine.

And in the maintenance position showing the old vibration resistant mount on the left side. When the old wind sock was lower down the pole.


cheers Ian :slight_smile:

2 Likes

A rather complex problem. The pole will shed vortices. These cause it to vibrate side to side normal to (across) the wind flow. This vortex shedding frequency can be calculated using the Strouhal Number. The Natural Frequency of the pole needs to be outside if this number. Then vortex shedding does not cause the vibrations that lead to false rain readings. This part is rather simple. The next part not nearly as simple. This is the influence that near field objects have on the wind flow, you need to look at the 360° picture for this. That is upstream, to each side and downstream of the pole.

Hence ideal pole is 10m into clear air, then if fully clear air, (very difficult in suburban area due to deeper boundary layer), then only the wind caused vortex shedding to consider.

Alternative, place a “spider” along the length of the pole. Three/four guy wires fixed top and bottom with turnbuckles. You may be able to use these to “tune out” the vibrations, as changes in rigidity can be bought about in such fashion, by varying the tension in the guy wires.

Good luck with these.

3 Likes

Wind is back and with it… tada… false rain detection :face_with_symbols_over_mouth: The PVC pole is clearly oscillating in the wind compared to the aluminum mount I had before. It doesn’t seem to be any better or worst I think. Just as bad as before.

I’m pissed, I’ll keep things like this for a time, not trusting that haptic rain sensor and hope @WFstaff eventually manage to improve the situation in software real soon (though I doubt since the issue seems to be known for years). I had higher expectations for that price range.