The rainfall recording on my ‘station on a hilltop’ (Ecowitt Wittboy) seems to have gone crazy! It seemed to be fine until mid-March, but in the 2-3 weeks since then it’s recorded 3500mm of rain. Now I know it’s been wet in the UK, but the area it’s in usually only averages 1000-1300mm of rain per year so 3500mm in just a few weeks seems rather implausible.
I’ve seen comments about the accuracy of piezo rain ‘gauges’ but I’d always assumed they were about recording 3mm when there was only 2mm or vice versa, not massive over-reading of rainfall. I’m somewhat puzzled that this has only just started happening despite the station having been installed for about 12 months. It also doesn’t seem to have happened after a severe weather event, e.g. very strong winds, etc.
I’m a 4 hour round trip from the station with limited access time due to the venue having paying occupants much of the time so I probably won’t be able to access it until our annual trip to the area at the end of July. Given that, does anyone have any suggestions of why this might be happening and/or what could possibly be done to fix it?
I have a WeatherFlow Sky and lately a WittBoy.
I can’t say that I am impressed with this new technology measuring Rain.
I also purchased a WH40 and I am trying to calibrate the WittBoy as good as possible.
Unfortunately the WH40 has different readings in comparison to my VantageVue also.
I am also comparing to a normal analog rainmeter.
Would be better if WD could use the calculated rain from WxFlow. They use other stations data to “correct” your station data to more accurately measure rainfall. I’m positive they know the haptic is not good at measuring rain, to many variables involved. Sucks but it is what it is…
There has been a whole load of discussion on the WXForum about this issue which has caused some controversy! I’m guessing this is not about calibration though… I’d suggest contacting Ecowitt and see what they say.
I’d say this is unlikely. Some of the excessive rainfall amounts occur at night when it’s unlikely there will be any birds flying around. I don’t think there are bats up there, at least not in winter.
I’ve considered this, but the property isn’t really suitable. One issue is that it’s really exposed with nowhere to really put a rain gauge in the places where it should be sited. Second is that it’s a holiday cottage so if I put it anywhere near the ground then it could be poked at or have a football kicked at it.
It’s an Ecowitt so doesn’t use WxFlow and it’s also not WD based. There’s no computer at the location to process the data so it’s fed directly into a PWSWD web site. So I only have access to ‘raw’ data.
Unfortunately not.
No, it gross overreading, unless there have been biblical level floods up there. I’ve no ‘real’ data though, I’m just assuming that 3.5m of rain in a few weeks is unlikely to be real. I’ll have to see if I can get an answer from Ecowitt.
Ecowitt Wittboy looks very similar to the Tempest but sorry about mixing them up.
Yes, the haptic sensor is terrible at determining rainfall amount. In my experience there is a sweet spot that when the rain drops are a perfect size the readings measure very close to my VP2. But of course that rarely happens. Most of the time the Tempest will over-read. And if the rain drops are small then it will also not detect any rain.
In my opinion the haptic rain sensor has no place in any station. It does such a terrible job of measuring rain amounts that Wxflow has to correct the data before publishing on their website.
To get enough dashboard testdata, I use parallel a Davis-VUE, a WeatherFlow and the Ecowitt-WS90 with a < 30 cm distance between the devices.
On top of the apartment building.
They have different rain values but on average they are very similar.
Very small raindrops and smaller amounts are better measured by the electronic devices.
But the WeatherFlow and the Ecowitt do not like spiders and other small crawling insects.
That gives strange results albeit not that high as Chris is measuring.
But cobwebs can cause rain values even when it is not raining.
Bigger problem: water leakage:
The first WS90 was “not good enough” sealed and the water leaked into the device giving very strange and high values before being scrapped.
It sounds negative, but the Davis-VUE and another “mechenical” Ecowitt also have had their share of measurement-problems.
We pride ourselves on standing behind our products.
We proudly uphold our commitment to quality technology. We warrant that all products we manufacture are of good material and workmanship, and free of defects when properly installed and operated.
Limited 1 Year Warranty
We warrant that all products we manufacture under this warranty are of good material and workmanship and, when properly installed and operated, will be free of defects for a period of one year from the date of purchase.
Most of the time my Tempest measures X2 to X3 the rainfall measured by my VP2. But if I go to the Wxflow site to view my data they have “converted” my measurements to almost match the VP2. So they know their sensor can’t reliably measure rainfall without applying corrections. I would bet that these corrections are from “non-haptic” stations, ie. tipping rain buckets, from “near” my location.
Yes, both the VP2 and Tempest stations have pro’s and con’s, but a tipping rain gauge if far more accurate than these haptic sensors could ever hope to be. I understand how disappointed you must be with the performance of your remote station and with no way to make corrections. Even if you were running WD there is no correction factor that could be employed to offset the discrepancy with any degree of accuracy.
Without question, this is not what my experience has been. Readings almost never match up. The only time my readings will be similar is on those very rare occasions where the rain drops are the ‘perfect’ size to be analyzed correctly by the haptic sensor. And in almost every case, very small rain drops are not even detected by the haptic sensor.
Hi
I do agree with the notes on wittboy haptic sensors, I use one, and sometimes when it is just a drizzle of rain it might not register any, so that’s why I have a separate rain collector for this reason, so that I can adjust any “Slight Rain” that comes along.
I suspect the problem with mine is more than likely an electrical or similar fault. The station has been installed for 9-10 months and whilst I’ve no way to verify the accuracy of it the massive rain rates have only occured within the last 2-3 weeks. Since the problems started they are happening consistently so the behaviour of the haptic sensor changed dramatically recently.
Waterleaking.
When my wittboy arrived it looked like this.
I didn’t know there was an elastic rubber band around the top.
I found out when I contacted Ecowitt support.
They told me to try to pull up the band so it would fit nice along the upper edge.
But I couldn’t.
It was perfectly aligned at the bottom edge.
It was probably a miscut gone wrong at production.
Ecowitt sent me a new one for free.
When you looked from the top that didn’t look good, water probably could have leaked there.
I’ll have to see if my brother can see anything wrong when he’s next there with a ladder. The station is mounted on the chimney so not very easy to access. I doubt he’d be able to see the rubber band or top surface from the ground.
