Haptic rain sensor not detecting rain

In Southern California, much of our annual rain total is comprised of light rain, below the sensitivity of the Tempest haptic sensor. I’ve already reported this to customer service, but I’m wondering how many others are affected by this problem. When I look at other local Tempest installations, they have ALL missed today’s rain total of approximately 0.12” (verified by a graduated cylinder and my WS-1400-IP), and if left uncorrected, the Tempest will probably miss half the annual rain total in SoCal. The only way I can see to correct this is to increase the sensitivity of the sensor, at the risk of getting false positives from vibrations, etc. As it stands, I have to run two separate weather stations to keep accurate rain records, and by the way, the “rain check” feature hasn’t helped a bit.

It’s not an easy problem to solve. If the haptic sensor is too sensitive, then even the slightest breeze going across the sensor will record some activity. However, if the sensitivity is turned down, you miss rainfall. Even the tipping buckets are not accurate, a quick drizzle won’t be enough to tip the bucket. The only sure method is a graduated cylinder.

I have an Acurite Atlas and a Tempest. It’s not unusual to see the total rainfall recorded to differ by 0.50+.

Does anyone know why when I turn on rain check for a particular day or event, it effects all my other past readings in history? Why doesn’t it let those older past readings alone?

It doesn’t sound like an enabled feature for raincheck at this time. It’s an all or nothing.

I should add that it missed a rain accumulation of rate of 0.21”/hour, which should hav3 been detectable IMO.

My Tempest didn’t detect the light rain we had in San Diego this morning. It would be nice if the sensitivity was configurable. Perhaps it needs an add-on moisture sensor to work in conjunction with the haptic sensors.

1 Like

Having an add-on moisture sensor is an interesting idea. How would it prevent a false light drizzle/rain reading if there is heavy dew or fog?

I was thinking that the moisture sensor would dynamically adjust the sensitivity of the haptic sensor.

if (moisture) 
    set sensitivity high
else 
    set sensitivity low

Or the moisture sensor is a confirmation for the haptic sensor. Maybe the sensitivity is always high and for a rain event to trigger both haptic and moisture sensors need to be active.

2 Likes

It may also benefit to do the same if/else statement with the barometer to turn up the sensitive.
If barometer pressure is dropping set sensitivity high; else set sensitivity to low.

3 Likes

I’m returning my Tempest within the 60-day return period.

I have the same problem with precip here in the Seattle area. Lately, most our rain has been very light but enough to produce puddles on my driveway and even soaking the Tempest. Yet, Tempest happily reports no precip all day!

I’ve brought it up several times to tech support (complete with screenshots of radar showing the area covered in rain, screen shot of the Tempest app showing no rain and the Tempest itself being absolutely soaked. All they say “yes, we know, very light rain is not detected”. With they’d been upfront about that earlier.

I really like the rest of the hardware but the rain detection is a total miss as far as I’m concerned.
Add to that the fact that the app doe snot even keep track of monthly let alone annual rain totals, I’m pretty disappointed.
When I spent $250 I had high hopes this would be the best weather I’d ever owned. It was not to be.
Back it goes. :frowning:

And in Seattle you get LOTS of long periods of light rain. What are you going to replace the Tempest with?

I have two Sky units a couple years old. Not sure how long the learning process goes on but today I had a thundershower with the following results.

Sky # 1 .89"
Sky # 2 .94"
Cocorahs “91”

While not always perfect I have noticed a fair amount of events where this sensor is close to what actually came down.

Most likely WS-2000 from Ambient Weather.

Are they accurate in light rain?

My VP2 is right on the money here about 1-2 miles from Puget Sound between Seattle+Tacoma almost in typical PacificNorthWet rain from ‘the marine layer never lifted’ to ‘monsoon for here, just a ten minute annoyance for near Philly :-)’

(lightly related - if anybody near Seattle wants to pick another Tempest station cheap let me know via PM please.)

Just installed my Tempest 2 days ago and it properly detected rain last night. However, just a few minutes ago, we had a light rain and I was watching the app, waiting for the rain notification to come and… nothing :frowning: No rain detected :thinking: And again, light rain right now and nothing detected. That’s a bit disappointing. Any news about this issue from Weatherflow?

I have an official rain detector in my town (2.5km away from my Tempest). I will check if they report something during that time frame and also check the cumulative values.

If or until WF develops a wireless “tipping bucket” to integrate into the System, experience has shown that the Haptic Sensor is not reliable for accurate local measurements. We “old timers” have been nagging for this for a long time now… The Haptic Sensor would be great for early detection of precip. (in which a Tipping Bucket lags), but until an integrated system using the best of both worlds is available, I’m afraid we’re out of luck there.

1 Like

Based on direct comms with Weatherflow, it’s pretty apparent that the haptic sensor has insufficient signal-to-noise. Whether that’s a consequence of inadequate sensitivity, perhaps due to the reduction in sensitivity due to the outer shell, a high background noise level in the system, or the inability to deal with other noise inputs not related to rain (wind, vibration, etc.), I think the overall solution should include better signal conditioning or filtering.

It sounds to me like the rain sensor is useless. Not only that light rain is not detected, but also the rain is missing in the rain amount.

Drizzle can be found all over the world. I have the same problem with the insensitive sensor (Berlin, Germany).

We had a few light rain events yesterday and the official rain detector from my town didn’t notice anything either. Earlier that day, we had stronger rains and my Tempest recorded 2.9 mm whereas the official records says 5.4 mm. Of course, both are 2.5 km apart so take this with caution. I’ll monitor a bit both results and see how it goes.

Weird is that Weatherflow says that the rain sensor of the Tempest is accurate between 90-100% compared to a rain gauge. I will see if I can get a rain gauge and do some comparison also.

1 Like