WXforum.net
Weather Station Hardware => Other Weather Station Hardware => Topic started by: Brian J on October 26, 2018, 09:37:15 AM
-
Hi all, I'm seeking opinions on the accuracy of and experiences regarding the WeatherFlow stations. I know there may be issues with rainfall - of which I don't much care. Temp, dewpoint, pressure, wind are my concerns. How is the sonic anemometer correction doing in rainy conditions? How high a wind have you all measured? Any thoughts are much appreciated. My preference is for no moving parts, which limits the list of options as far as personal stations are concerned.
Thanks!
Brian
-
Here's a live link to my custom weewx report, which plots the WeatherFlow station and my Davis VP2 on the same graphs:
http://ae3.homelinux.net:845/weewx/compare.html (http://ae3.homelinux.net:845/weewx/compare.html)
As we are expecting a nor'easter here with 35+ MPH wind and 1-3" of rain tomorrow, feel free to stay tuned to that link during the day and we'll both see how it performs.....
(https://images.hvtraffic.com/images/2018/03/18/20180318165938-647049e9.jpg)
-
Here's a live link to my custom weewx report, which plots the WeatherFlow station and my Davis VP2 on the same graphs:
This is awesome. It looks pretty good. The pressure differences are presumably just calibration. Does your VP2 correspond well to other stations/ASOS?
Thanks for sharing this!
-
Here's a live link to my custom weewx report, which plots the WeatherFlow station and my Davis VP2 on the same graphs:
The pressure differences are presumably just calibration. Does your VP2 correspond well to other stations/ASOS?
Beat me to it, same question.
-
There's an ASOS (KMGJ) across town to the west, but it is in an open area with crazy wind exposure and I'm lucky to see 29 MPH peak gusts. There's also KSWF about the same distance to the east.
WeatherFlow just added the barometer to their continuous learning (CL) algorithm in the past week or so, and I just peeked at the ASOS to see that I was reading 30.086 in the WF app when the ASOS was reading 30.07. (weewx was reading 30.06, because I had a pre-CL adjustment hard-coded into weewx. I have made no attempt to calibrate the Davis barometer.
https://smartweather.weatherflow.com/share/910/grid (https://smartweather.weatherflow.com/share/910/grid)
One can never have too many weather stations...and they all basically read the same thing most of the time:
(https://images.hvtraffic.com/images/2018/09/25/20180925175422-0b9fad1b.png)
(https://images.hvtraffic.com/images/2018/07/02/20180702123658-66151ea2.jpg)
I'm even more curious than you are about if the ultrasonic is going to drop out in 20+ MPH winds and moderate/heavy rain.....
-
WeatherFlow just added the barometer to their continuous learning (CL) algorithm
This question is not necessarily for you, but I've always wondered why some PWS have to "learn" the pressure. It's merely a measurement, like every other piece of data. Soon as power is supplied to a VP2 or Vue console the pressure reading is instantly displayed, there is no "learning".
-
I have been happy with my WeatherFlow. It seems to read speeds a little higher than my Davis Vantage Pro on windy days. Not sure which is correct. Temp matches well with the Davis and I have it mounted in a box like this converted to a temp shelter. A few weeks ago I saw these on Amazon for $25 .
https://www.gardeners.com/buy/wood-shutters-lantern-with-led-timer-candle/8592509.html
-
WeatherFlow just added the barometer to their continuous learning (CL) algorithm
This question is not necessarily for you, but I've always wondered why some PWS have to "learn" the pressure. It's merely a measurement, like every other piece of data. Soon as power is supplied to a VP2 or Vue console the pressure reading is instantly displayed, there is no "learning".
With Acurite (and I think some LaCrosse systems now) "learn" a sea-level adjusted pressure by averaging pressure readings over time (e.g. 2 to 4 weeks). This allows a reasonable setting to be achieved without the user having know or do anything. The big disadvantage is that it is very slow method.
Acurite only uses the "learning method" for the local console display value. Values transmitted to a computer or network are adjusted using more typical methods.
My understanding of the WeatherFlow CL system is that it uses your elevation (gained from the network or manual input) to adjust the pressure, then further "fine-tunes" using regional pressure data from the network. It's almost like they are re-calibrating the baro sensor daily instead of yearly like you might to account for an aging sensor's drift.
While that sounds good on paper, I'm unclear what practical advantage this has. Maybe it's because the WeatherFlow baro sensor is outside, exposed to wide temperature variations, instead of being situated in a more temperature-stable indoor environment?
-
WeatherFlow just added the barometer to their continuous learning (CL) algorithm
This question is not necessarily for you, but I've always wondered why some PWS have to "learn" the pressure. It's merely a measurement, like every other piece of data. Soon as power is supplied to a VP2 or Vue console the pressure reading is instantly displayed, there is no "learning".
This allows a reasonable setting to be achieved without the user having know or do anything. The big disadvantage is that it is very slow method.
Holy crap. The only thing I needed to know was my elevation MSL and select "Altimeter Setting" on the console (obviously my choice of scale). No waiting, instant, accurate pressure!
How can it be easier than that?
-
]Holy crap. The only thing I needed to know was my elevation MSL and select "Altimeter Setting" on the console (obviously my choice of scale). No waiting, instant, accurate pressure!
How can it be easier than that?
By just unpacking the device and letting it figure it out on its own.
Not all consumers understand how to set a barometer, or even know what their elevation is.
-
While that sounds good on paper, I'm unclear what practical advantage this has. Maybe it's because the WeatherFlow baro sensor is outside, exposed to wide temperature variations, instead of being situated in a more temperature-stable indoor environment?
WF is ultimately planning on having every sensor in their stations calibrated via CL, which of course requires finding/licensing global data source(s) for each reading to calibrate against. I'm guessing that barometer was a low-hanging fruit, and think I remember reading that they had a global satellite source for high noon UV measurements to get those trimmed in via CL.
The WF station's barometer will take elevation and height inputs out of the box and do the basic corrections for the Bosch BMP-280 sensor to get a ballpark figure without CL, so there isn't an initial learning period to get at least a rough value within the BMP-280's production variances.....
-
Interesting station. I would like rainfall comparison with manual Cocorahs gauge.
-
As we are expecting a nor'easter here with 35+ MPH wind and 1-3" of rain tomorrow, feel free to stay tuned to that link
Your Northeaster high risk has turned out to be marginal - not much rain or wind, based on your 27 weather stations there. No wind UFN but the GFS promises rain there Monday and Friday, so I'll keep watching...
-
Thank you all for your comments and especially for the VP2/WF comparison page which is fantastic. With your and other's comments, I went ahead and ordered a station from WeatherFlow. I'm a little nervous at all the sometimes long-term calibration which seems necessary but it seems like a very serious and promising hardware solution and one that I can well use (I'll take it mobile after calibration is done). I look forward to sharing my comparisons when I have them.
Brian
-
There's an ASOS (KMGJ) across town to the west, but ...
I'm even more curious than you are about if the ultrasonic is going to drop out in 20+ MPH winds and moderate/heavy rain.....
Man forget about the weather stations, I wanna know who is doing the cool sky writing in the background of you one pic, that looks like some serious aeronautics going on back there!! LoL :grin: =D> \:D/
(http://bethelcares.com/Temp_Img/weathersky.jpg)
-
There's an ASOS (KMGJ) across town to the west, but ...
I'm even more curious than you are about if the ultrasonic is going to drop out in 20+ MPH winds and moderate/heavy rain.....
Man forget about the weather stations, I wanna know who is doing the cool sky writing in the background of you one pic, that looks like some serious aeronautics going on back there!! LoL :grin: =D> \:D/
(http://bethelcares.com/Temp_Img/weathersky.jpg)
See this thread for answer with more pictures. :
https://www.wxforum.net/index.php?topic=35327.msg361984#msg361984
-
The continuity of the WxFlow and the VP2 seems very impressive aside from precipitation amount. Is this correct?
-
Well the rainfall accuracy on the VP2 is nothing to write home about. That's why I was asking if anyone compared to 4" diameter manual Cocorahs gauge.
-
My aerocone VP2 and 4" Stratus/CoCoRaHS are within one tip of each other on most rain events, and share a mounting J-pole.....
(https://images.hvtraffic.com/images/2018/04/08/20180408193650-3f6b251b.jpg)
-
I like most everything about it, except that "haptic" rain sensor (I just commented on another thread, been meaning to here).
IMO, it will never work. Its' precip measurements are entirely software synthesized, not actually measured. Derived solely from inherently indirect haptic (and audible?) inputs incapable of sensing all the permutations of precip.
Trying to do that for such a variably 'physical' thing as rain, or precipitation of any/all types, which varies in so many properties, will never be accurate across the spectrum of precipitation encountered in actual conditions.
-
I like most everything about it, except that "haptic" rain sensor (I just commented on another thread, been meaning to here).
IMO, it will never work. Its' precip measurements are entirely software synthesized, not actually measured. Derived solely from inherently indirect haptic (and audible?) inputs incapable of sensing all the permutations of precip.
Trying to do that for such a variably 'physical' thing as rain, or precipitation of any/all types, which varies in so many properties, will never be accurate across the spectrum of precipitation encountered in actual conditions.
The haptic sensor was one of big reasons I bought in (as well as the ultrasonics).
Maybe haptic rain measurement is a dead-end. Maybe not. We won't know until we've really given it a shot.
Personally, I'm optimistic for the tech.
-
I like most everything about it, except that "haptic" rain sensor (I just commented on another thread, been meaning to here).
IMO, it will never work. Its' precip measurements are entirely software synthesized, not actually measured. Derived solely from inherently indirect haptic (and audible?) inputs incapable of sensing all the permutations of precip.
Trying to do that for such a variably 'physical' thing as rain, or precipitation of any/all types, which varies in so many properties, will never be accurate across the spectrum of precipitation encountered in actual conditions.
The haptic sensor was one of big reasons I bought in (as well as the ultrasonics).
Maybe haptic rain measurement is a dead-end. Maybe not. We won't know until we've really given it a shot.
Personally, I'm optimistic for the tech.
That's the description I had rattling around my noggin:
"haptic rain measurement is a dead-end".
Trying to measure an infinitely variable tangible object intangibly. IMO, it just cannot be fully trusted.
Yes, I have read through much of the WeatherFlow forum pages in respect of the "rain" problems. (forgot to initially mention that)
I'm not willing to give it a $300 shot. Maybe they should offer an optional standard tipping type precip gauge. I would probably bite for that.
The ultrasonic anemometer however, I've been drooling over getting one of those for the boat for years.
The standard mast-top wind direction and speed devices, almost exactly like those of most PWS's, are prone to damage atop a 44 some-foot mast (sea gulls, salt air, etc). Corrective repairs aloft are a real PITA.
The top-end marine systems makers have offered nearly trouble-free ultrasonic true and apparent wind measurement units for mast-top mounting for a long time. Their target market is the high $pending yacht racing crowd of course, not so much the lowly cruisers.
I think that the actual devices are sourced from Airmar, then branded by B&G, Raymarine, etc., as part of their top-end navigation packages. Though you can buy just the anemometer and a GPS-driven display, still quite pricey (as is anything "marine").
http://www.airmartechnology.com/weatherstation-info.html?category=WX (http://www.airmartechnology.com/weatherstation-info.html?category=WX)
BTW, the key marine difference from stationary PWS's is the derivation of true and apparent wind conditions in relation to the heading and speed of the moving vessel, which requires a log, gyro, or GPS input and some simple vector calculations. See: your typical 'storm chaser' WX nut vehicle.
-
Weatherflow came out this morning with a firmware update to the hub and I am now tracking my Atlas within .05 hundreds of a inch of rain all day long and still going, DSJ over there told me that the CL part is still not released yet for the rain calibration just a good old factory cal /algorithm!
-
Weatherflow came out this morning with a firmware update to the hub and I am now tracking my Atlas within .05 hundreds of a inch of rain all day long and still going, DSJ over there told me that the CL part is still not released yet for the rain calibration just a good old factory cal /algorithm!:
I re-mounted my SKY this evening just before a brief shower. I'll have to check to see if I have the latest firmware.
Anyway, my Atlas and CoCoRaHS match at 0.09", but the SKY is over-reporting with 0.14. I thought the overage might be due to my re-mounting work, but that only accounts for 0.01" according to the time graph.
Here's what I currently have for firmware:
HUB 98, AIR 20, SKY 43
It looks like my HUB rebooted about 4 hours ago, so I'm assuming that might have been the update.
-
Yup your on the latest firmware,it seemed to miss out on the 5 hundreds of a in in a very light spurt of rain I am here in the Midwest and we are getting slammed yesterday and today.
-
My other issue is that the AIR seems to be a bit "deaf" with regards to lightning.
The Atlas picked up 25 strikes during today's storms. Zero for the AIR.
Even though it's going to put the AIR in the sun, I'm going to relocate it near the Atlas and see if that makes any difference in lightning detection.
I did notice a "Lighting Disturber" sensor status tonight for the AIR. I'm not sure what's causing that or why it isn't going away.
-
Mine seems to work reasonably well I have it out in a open area 4, above a grassy area on same pipe as sky and a
old 5-1
I have it mounted in a home made wooden Stevenson Shield
[ You are not allowed to view attachments ]
[ You are not allowed to view attachments ]
-
Nice shield I like... =D>
I'm thinking the Davis 7714 shield it would fit inside also. They have a nice platform inside to hold sensors and for passive shield works well.
-
Update on my WeatherFlow issues.
WeatherFlow customer service is great! The sent me a new AIR because of my lightning issue with a pre-paid mailer to send the old one back for analysis. The new one is now in service.
Also, I noted the UV reading on my SKY was way too high (e.g. almost 8 when forecast and Atlas readings were 3) compared to my Atlas and the EPA forecasts. WeatherFlow support checked it remotely and made an adjustment on their end. Now there is much agreement between them!
-
Any user updates on Weatherflow?
I've been looking at them all and it looks to fit the bill for what I am after.
I'm in the Desert and need humidity reading below 10% most of the year so that thins the heard a lot.
I also like that fact the Metrobridge can grab the data.
It would make a nice Xmas present for myself! ;)
Thanks!
-
I too am interested in any updates. Is there any way to connect this to some type of base station as you would with most PWS?
-
They do not offer a console, only the smartphone apps and web page. I have had a $49 Amazon tablet hanging on the wall for over a year now running the Android WF app, and look at it first even with an Acu-Rite console hanging right below it:
(https://images.hvtraffic.com/images/2019/01/07/20190107231813-c6cc5c92.png)
Let's see Davis or any other PWS vendor sell a color touchscreen console for $49!
There are also a few neat display options listed under the 3rd party software section in WF's support forum, one using tablets and two using Raspberry Pi computers to make cheap console substitutes with nicer graphics than the WF app has.....
-
Thanks for the pix, very nice. Can the display be set to a continuous 'on' state?
Also, as you have both the Davis (as I do) and the WF, what do you see as the pros & cons of each?
-
There are free Android apps to keep the screen on 24/7. I'm using one of them, but don't recall which off the top of my head.
I like that the VP2's wind sensors do not have any large horizontal surface for snow to pile up on! These pictures from last winter motivated me to buy the VP2:
(https://images.hvtraffic.com/images/2017/12/11/20171211145651-32343876.jpg)
(https://images.hvtraffic.com/images/2017/12/11/20171211145655-2350237b.jpg)
WF's haptic rain sensor is currently a work in progress. They are developing a continuous learning process to calibrate each individual unit, and there is a lot of promise in their idea of a rain gauge with no moving parts. Just keep in mind that it can't measure snow/melt. The one unique thing that it can do is detect and alert on the start of rain when it detects the first drops, which is useful in home automation and for sending smartphone alerts. Right now, I wouldn't use the WF station by itself without a good tipping bucket of some sort for measuring rainfall manually.
WF's strong point is making your station data available via multiple vectors, from their cloud service all the way down to local UDP broadcasts of live 3-second (wind) observations on your local network segment. As the person who developed/shared the weewx station driver for weatherflow-UDP, I'll say that it is child's play to grab the data and do whatever you want with it. The person who came up with this idea at WF deserves a Nobel Prize, and every other vendor should make it so easy!
In terms of accuracy, the WF station is tracking nicely against my VP2 and Acu-Rite 5-in-1 under most conditions. Siting the WF Air in a spot without sun may be a challenge for some, but I'm having great results mounted to the north wall of my detached/unheated garage. The WF's wind readings will sometimes get wonky in the rain, as the collected moisture change the ultrasonic sounds. (I would imagine that every other ultrasonic with a covered "slot" for wind has the same problem.)
I like the WF hardware a lot, but at the end of the day my VP2 has earned its spot as the "official" PWS (out of the three in my yard) for reporting to various services.....
-
Vreihen, thanks for the info! I really do like the elegant nature of the WF sensors and the fact that it's devoid of any moving parts. As you say, there are some issues, but it looks like most of them will get ironed out with time.
I'm happy enough with my VP2, although frustrated with the dated nature of the hardware (particularly the base units), that I'll sit tight and see how things progress at WF.
-
I have been considering a WF, but recently the Lufft WS10 was brought to my attention. Doppler precipitation measurement.
Radiation shielding is my remaining question before beginning to convince my wife of the dire need for an upgrade to the PWS.
-
I have been considering a WF, but recently the Lufft WS10 was brought to my attention. Doppler precipitation measurement.
Radiation shielding is my remaining question before beginning to convince my wife of the dire need for an upgrade to the PWS.
Nice... but on sale for $1199 USD?
-
Well, you can get it for a bit less.
If I am looking for better quality, etc. why not add the bonus of Doppler?
I’m holding out until I know it will be accurate in sunlight. I can’t see any way that it is going to be in a radiation shield.
Lufft does good work though, so I’m hopeful.
-
Well, you can get it for a bit less.
If I am looking for better quality, etc. why not add the bonus of Doppler?
I’m holding out until I know it will be accurate in sunlight. I can’t see any way that it is going to be in a radiation shield.
Lufft does good work though, so I’m hopeful.
Looks very nice! All of the pictures don't show radiation shield so you might be good.
I would be interested in how you grab the data, be nice if a Meteobridge could parse thru it. if that was the case, I would be interested in the Lufft as well.
I picked up a Weatherflow, nice unit, I had to put bird spikes on it to keep it from measuring precipitation when Birds landed on it :roll: and I have a new rubberized mount I need to install to reduce false rain readings in high winds, so we will see how it goes!
-
If I am looking for better quality, etc. why not add the bonus of Doppler?
I’m holding out until I know it will be accurate in sunlight. I can’t see any way that it is going to be in a radiation shield.
https://www.lufft.com/products/compact-weather-sensors-293/lufft-ws10-smart-weather-sensor-2421/ (https://www.lufft.com/products/compact-weather-sensors-293/lufft-ws10-smart-weather-sensor-2421/)
Have you looked at the Lufft specs closely? +/- 2 C in calm air with the heater off (what if the heater is on?), precipitation accuracy is 20% under lab conditions (ie: synthetic/perfect rain drops), and +/- 10 degrees on the wind direction. Probably useful for controlling HVAC or irrigation in a low-maintenance automation setup, but not really a scientific instrument.....
-
If I am looking for better quality, etc. why not add the bonus of Doppler?
I’m holding out until I know it will be accurate in sunlight. I can’t see any way that it is going to be in a radiation shield.
https://www.lufft.com/products/compact-weather-sensors-293/lufft-ws10-smart-weather-sensor-2421/ (https://www.lufft.com/products/compact-weather-sensors-293/lufft-ws10-smart-weather-sensor-2421/)
Have you looked at the Lufft specs closely? +/- 2 C in calm air with the heater off (what if the heater is on?), precipitation accuracy is 20% under lab conditions (ie: synthetic/perfect rain drops), and +/- 10 degrees on the wind direction. Probably useful for controlling HVAC or irrigation in a low-maintenance automation setup, but not really a scientific instrument.....
Yeah, I too kinda gave it a :roll: when looking at it last night. And at $1000+ a pop, I think you're paying for all that stuff to be crammed into one space. I also noticed how they mention it's a "good price performance ratio", not sure what standard they use, I guess the "everything crammed into one place" standard.
-
If I am looking for better quality, etc. why not add the bonus of Doppler?
I’m holding out until I know it will be accurate in sunlight. I can’t see any way that it is going to be in a radiation shield.
https://www.lufft.com/products/compact-weather-sensors-293/lufft-ws10-smart-weather-sensor-2421/ (https://www.lufft.com/products/compact-weather-sensors-293/lufft-ws10-smart-weather-sensor-2421/)
Have you looked at the Lufft specs closely? +/- 2 C in calm air with the heater off (what if the heater is on?), precipitation accuracy is 20% under lab conditions (ie: synthetic/perfect rain drops), and +/- 10 degrees on the wind direction. Probably useful for controlling HVAC or irrigation in a low-maintenance automation setup, but not really a scientific instrument.....
Thanks for that. I have no idea what specs I looked at, but confirmed that it isn't very good at all for a PWS.
The quest continues...
-
I have a WS10 and while it's not a bad unit, I'd probably opt for the WS700 instead. Unfortunately, the lowest I've seen the WS700 listed for is $3,828.00. If anyone comes across better pricing do give me a shout.
The one application where I think the WS10 really does shine is automation.
-
look out for the new improvements recent kit arrived as some improvements over the original .
solar power option no more batteries
improved cushion mount to suppress vibrations triggering false rainfall measurements
superhydrophobic coating to reduce wind error readings during rainfall
many underneath the hood firmware tweaks to improve overall performance .(ongoing)
still the slickest looking newcomer to the market and it can only get better and really doesnt getting any simpler in terms of installation..
living in hope that the AIR temperature model gets a dedicated optional optimally designed radiation shield it simply dont like sun direct and residing in a location that
cant achieve 100% shade all year round .
overall opinion
being that it is a campaign funded product changes dont always come overnight or in a timeframe everyone may expect or want but comparing to last years models the recent additions are beneficial.this is my sixth package and this the first that has had no anomalies out of the box so its worth noting the guys at weatherflow do listen and take constructive realistic ideas away and put them to practice .
(https://res.cloudinary.com/brian-underdown/image/upload/v1554484013/wf_j6ubzs.jpg)
-
I'm glad to see a solar panel option. I do wish they would also make a wired adapter. If a wired option (for power) was available they could perhaps add a heater to the dome.
-
weather34, I was about to pull the plug on one. So should I wait for this new version?
Regards,
-
-
look out for the new improvements recent kit arrived as some improvements over the original .
solar power option no more batteries
So the solar panel module has a battery pack built in?
-
look out for the new improvements recent kit arrived as some improvements over the original .
solar power option no more batteries
So the solar panel module has a battery pack built in?
hello
without pulling it apart im guessing it is lithium flat type cell in there probably phosphate type its pitch black dark as i type and i only see a minor drop from daylight hours of 3.4c to dark at 3.3v .I should have mentioned must remove aa batteries no requirement anymore
txweather drop them an email see if they readily available
heater more additional costs could always make your own using the plated contacts..
probably more useful info on there forums ...
-
in a couple a weeks ill do a full review write up on it im particularly curious of rainfall performance in terms of measurements...
-
look out for the new improvements recent kit arrived as some improvements over the original .
solar power option no more batteries
So the solar panel module has a battery pack built in?
batteries no requirement anymore
What happens when there's insufficient sun? Super-cap? If only that, how long will it last?
-
look out for the new improvements recent kit arrived as some improvements over the original .
solar power option no more batteries
So the solar panel module has a battery pack built in?
batteries no requirement anymore
What happens when there's insufficient sun? Super-cap? If only that, how long will it last?
i cant answer that we are past our winter we are in that moderate to high uv period but im sure that has been addressed from what i recall they've been testing since last august on prototypes . the sky generally only starts to fail when voltage drops to around 2.5v that was quite apparent on aa lithium batteries after 6 or. 7 months .
-
I'm curious about the accuracy of their ultrasonic anemometer. Does anyone have real life experience or info on that?
-
I'm curious about the accuracy of their ultrasonic anemometer. Does anyone have real life experience or info on that?
Live WeeWX report of Davis VP2 and WeatherFlow Sky on same mast:
http://ae3.homelinux.net:845/weewx/compare.html (http://ae3.homelinux.net:845/weewx/compare.html)
(https://images.hvtraffic.com/images/2018/09/25/20180925175422-0b9fad1b.png)
(https://images.hvtraffic.com/images/2018/03/18/20180318165938-647049e9.jpg)
The best comparison is the vector average winds at the bottom left, since it smooths out the ultrasonic's ability to measure a draft that isn't strong enough to push the spinny cups. That number is usually within a few degrees and 1 MPH on any day, but this morning is way off because the overnight winds were 0-1 MPH and only the ultrasonic could measure them.....
-
I'm curious about the accuracy of their ultrasonic anemometer. Does anyone have real life experience or info on that?
this morning is way off because the overnight winds were 0-1 MPH and only the ultrasonic could measure them.....
Not to hijack the thread, but tell ya what, I just replaced the speed cartridge on my anny and it consistently reads 1 mph. Don't think it ever did this before and that very well might be because this cartridge is different than what it replaced. Considering it's only $20 to the door and literally one screw to replace it, I think it's well worth it.
-
Not to hijack the thread, but tell ya what, I just replaced the speed cartridge on my anny and it consistently reads 1 mph. Don't think it ever did this before and that very well might be because this cartridge is different than what it replaced. Considering it's only $20 to the door and literally one screw to replace it, I think it's well worth it.
I'd have to see that, maybe they promised, but a 1 MPH 'wind' isn't much to overcome the inertia of the cups and associated bearings.
-
Not to hijack the thread, but tell ya what, I just replaced the speed cartridge on my anny and it consistently reads 1 mph. Don't think it ever did this before and that very well might be because this cartridge is different than what it replaced. Considering it's only $20 to the door and literally one screw to replace it, I think it's well worth it.
I'd have to see that, maybe they promised, but a 1 MPH 'wind' isn't much to overcome the inertia of the cups and associated bearings.
Look for yourself. Cup inertia doesn't last for 15+ minutes...
https://www.wrh.noaa.gov/mesowest/getobext.php?wfo=twc&sid=C2274&num=48
-
Look for yourself. Cup inertia doesn't last for 15+ minutes...
https://www.wrh.noaa.gov/mesowest/getobext.php?wfo=twc&sid=C2274&num=48
Impressive.
-
@vreihen: thanks. Though I can't really answer my question from the images, all I see is the curves of the fancy new devices are off from the Davis, sometimes they look almost random, looking at wind and precip... I'm mainly interested in a no-moving-parts anemometer that doesn't cost in the €1000+ range. Their method for measuring rain and rain rate is interesting, but reading other threads, it seems their rain gauge is not exactly reliable or accurate (yet?).
-
FYI, Friday and yesterday were not the best for comparing precipitation. We had sleet on Friday afternoon->night, which currently over-reports on the WeatherFlow due to the sound it makes, and some of it sat in the Davis cone until it warmed up yesterday before being measured. The WeatherFlow logged some false rain yesterday as well, since it is attached to my deck railing and we were doing some construction work with sawing/hammering on the deck. (There's also gaps in the WF data, as a result of my having to shut down the WeeWX computer and move it away from the construction zone.)
WeatherFlow has data scientists working on continuous learning algorithms to make their stations smarter, but still have not rolled out the rainfall code yet that I know of. If rainfall is important to you, I'd recommend using a decent tipping bucket or 4" manual Stratus rain bucket with the WF station.
WF as a company really stands behind their products, and has competent tech/customer support folks. They also publish open API's for multiple ways to access YOUR data for free, unlike the walled gardens set up by other PWS vendors in this price range such as AcuRite and Ambient who will only let their customers share with WU. There is a thriving community of third-party developers with support for the WF station's data, including the WeeWX UDP driver that I developed and freely share.
Although there are some pitfalls to the non-mechanical precipitation measuring technologies used by WF and the optical Hydreon RG-11, they do have the advantage of being able to send a start-of-rain alert upon detecting the first drop. This is useful for home automation folks, as well as people who leave their car windows/roofs open in the driveway. By the time a tipping bucket logs its first spoonful, those leather seats are pretty well soaked.....
-
Yeah, I understand the benefits and I'm sure they will present us with a polished, modern and accurate sensor cluster in the future, but it's apparent that they're not quite there yet. Their methods with all solid state sensors are the future IMO and it's also clear that someone has to take up pioneering, but now it's but a bet. In a year we'll see where they'll have arrived. On a side note, I was eyeing the Hydreon device for precip alerts, exactly for the purposes mentioned. It's also a very inventive device, but also lacks accuracy, so it's mainly for signalling rain and approximate rain rate.
-
Yeah, I understand the benefits and I'm sure they will present us with a polished, modern and accurate sensor cluster in the future, but it's apparent that they're not quite there yet. Their methods with all solid state sensors are the future IMO and it's also clear that someone has to take up pioneering, but now it's but a bet. In a year we'll see where they'll have arrived. On a side note, I was eyeing the Hydreon device for precip alerts, exactly for the purposes mentioned. It's also a very inventive device, but also lacks accuracy, so it's mainly for signalling rain and approximate rain rate.
Ive always been on the sidelines about rainfall and the technology used Ive not been shy of voicing that view to wf and I dont do it with an attitude . i had a lengthy skype with an engineer from a different company (wants to remain anonymous) he basically had a 2 year development budget using similar technology and concluded impossible to achieve any consistency however he did highlight more time was needed and a rethink on capture area.
I have pulled apart 5 of these sky units and pretty understand whats under the hood and the vulnerable areas , i tend to look at this technology with a thinking of microphones . I remain positive for the future about this product simply due the constant minor visual changes that are actually quite significant in performance values.
as above my hope and would love to see it made priority but you cant have it always is a dedicated temperature radiation shield designed smartly with efficient results.
100% shade idea is just not a good way to achieve sales or as a recommendation you end with a microclimate temperature readings and this doesnt always reflect surrounding area in general . i,e walk one side of my property it constantly remains damp in the short winter months little airflow or sunlight walk the other side of property or even a few metres out of the shaded area where it gets 8-10hrs of sunlight and good airflow totally different feel and readings altogether and you dont need a instrument to see the difference.
personally think this educated forum community should get behind this company as vreihen mentions they are listeners and realistic in there responses i was a bit dismayed early last year where a few field testers had little background or previous ownership of weather equipment ..however the field test group has rapidly grown and seeing some really good observations and realistic ideas quite frequently discussed ..
as for solar panel most drop ive seen under darkness is a drop 3.21v to high of 3.51 during peak sunlight period.
-
What's the real world accuracy of the rain sensor at the moment? 70%, 80%, etc.
-
from what I see based on observations is some inconsistency due to intensity of rainfall , however I can only speak for my own observations and comparisons to my hardware I have (vp2 calibrated for metric at 0.2mm per tip) and homemade measuring tube..for example this evening SKY recorded 0.7mm and Vp2 was 1.0mm light and variable max of 12kmh winds rain lasted around an 40 minutes . so under those conditions the difference is close enough to say which one is right ?? and create that confusion. What i may get chance is to see how pans out under really intense rainfall in the coming months as we approach stupid flash flood season then dry an hour later as if nothing happened. so time will tell , i recall last year the inconsistency was quite noticeable and very
susceptible to false rain measurements when winds were at that constant 50kmh or above for long periods . in recent months false rain under strong wind gusts has not been so apparent.
-
That falls in line with what I've read before. ~70% accuracy.
Are there any solid state precipitation sensors that are 100% accurate? All of the radar-based sensors that I've come across are rated at around 90% accurate. That's better than 70% but still not 100%.
At the moment I'm testing a Lufft R2S-UMB. It's rated for >90% accuracy and from my testing it's closer to 95% in sub 30MPH winds. Testing against a CoCoRaHS rain gauge.
-
If you read the CWOP siting guidelines, it seems that no rain gauge achieves 100% accuracy when there is any wind. They have graphs from studies showing a very dramatic falloff in rain capture under crosswind conditions.
Rich K.
-
achieving accuracy is always going to be a challenge in all in one hardware types , do you position optimally for wind or rain , so in regards to this hardware that is going to be difficult in fact most all in ones is going face the challenge so you you simply have to find a reasonable balance between the two .
wind ultra sonic technology very sensitive to nearby reflections and backdraft so i guess optimum is an open space aprox 8-10 feet of the ground however is that possible in an urban area ? so always a trade off when reading recommendations if your property,location is very different.
i live in a sea view apartment no option of using it near ground level its 80ft agl high winds and rainfall is always a challenge no matter what hardware i use ..
in regards to the hardware its as good as any all in one solution once again siting the temperature component in a dedicated radiation shield it will certainly compare well to competitors offerings . 18 months to 2 years if continuous development is made you will have a reasonably good high end consumer based product . personally based on recent delivery of hardware i would put more trust in durability over those fine offset offerings ..
-
Installed my Sky solar-panel today. Installation was very easy, and the unit powered up right away.
Delivery was really quick. The 40% coupon expires May 31, so hurry up if you haven't ordered yet.
https://shop.weatherflow.com/products/solar-power-accessory?utm_source=SWS+account+owners&utm_campaign=99c116ea0a-SOLAR_EMAIL_ANNOUNCEMENT_MAY2019&utm_medium=email&utm_term=0_e3814546b6-99c116ea0a-382252029&mc_cid=99c116ea0a&mc_eid=5447dd91b8
Use the discount link or code below to get 40% OFF when ordered before May 31, 2019. Buy Now 40% OFF. (use discount code at checkout step: "SOLAR40FF" )
[ You are not allowed to view attachments ]
[ You are not allowed to view attachments ]
-
yes, this is a great addition, the solar charge pack :)
mine has been working great for the last couple of months
-
WeatherFlow just added the barometer to their continuous learning (CL) algorithm
This question is not necessarily for you, but I've always wondered why some PWS have to "learn" the pressure. It's merely a measurement, like every other piece of data. Soon as power is supplied to a VP2 or Vue console the pressure reading is instantly displayed, there is no "learning".
With Acurite (and I think some LaCrosse systems now) "learn" a sea-level adjusted pressure by averaging pressure readings over time (e.g. 2 to 4 weeks). This allows a reasonable setting to be achieved without the user having know or do anything. The big disadvantage is that it is very slow method.
Acurite only uses the "learning method" for the local console display value. Values transmitted to a computer or network are adjusted using more typical methods.
My understanding of the WeatherFlow CL system is that it uses your elevation (gained from the network or manual input) to adjust the pressure, then further "fine-tunes" using regional pressure data from the network. It's almost like they are re-calibrating the baro sensor daily instead of yearly like you might to account for an aging sensor's drift.
While that sounds good on paper, I'm unclear what practical advantage this has. Maybe it's because the WeatherFlow baro sensor is outside, exposed to wide temperature variations, instead of being situated in a more temperature-stable indoor environment?
With no friction and resistance I would say the weather flow is more accurate for wind. It takes a certian ammount of wind to turn the cups.
-
My other issue is that the AIR seems to be a bit "deaf" with regards to lightning.
The Atlas picked up 25 strikes during today's storms. Zero for the AIR.
Even though it's going to put the AIR in the sun, I'm going to relocate it near the Atlas and see if that makes any difference in lightning detection.
I did notice a "Lighting Disturber" sensor status tonight for the AIR. I'm not sure what's causing that or why it isn't going away.
The acurite lighting detectors give false lighting sticks so the air maybe more correct, or unless you have experienced lighting in your area you may not know.
Would not trust the acurite lighting sensor either
-
My other issue is that the AIR seems to be a bit "deaf" with regards to lightning.
The Atlas picked up 25 strikes during today's storms. Zero for the AIR.
Even though it's going to put the AIR in the sun, I'm going to relocate it near the Atlas and see if that makes any difference in lightning detection.
I did notice a "Lighting Disturber" sensor status tonight for the AIR. I'm not sure what's causing that or why it isn't going away.
The acurite lighting detectors give false lighting sticks so the air maybe more correct, or unless you have experienced lighting in your area you may not know.
Would not trust the acurite lighting sensor either
No. The Air is not more correct. It's pretty clear that it under-reports strikes, at least with the two Airs I have worked with.
The Acurite is far more reasonable in its count. It is also consistent with the lightning strike alerts sent out by my local TV station.
-
I'm assuming it's the same, but looks like you Weatherflow folks may be getting some love from the NHC.
https://www.nhc.noaa.gov/text/refresh/MIATCUAT5+shtml/060351.shtml
-
I'm assuming it's the same, but looks like you Weatherflow folks may be getting some love from the NHC.
https://www.nhc.noaa.gov/text/refresh/MIATCUAT5+shtml/060351.shtml
might give you insight into there network which has been around prior to the consumer releases https://weatherflow.com/professional-services/weather-networks/custom-designed-mesonets/the-weatherflow-hurricane-network/
-
Acurite dont have accurite lighting stricks either. Lots of false stricks. They where worse when they first came out but still would not totally rely they are correct.
When the altlas came out lots off issues, but weather flow have a better and more dedicated team and have professional meteorologist, and team, whereas acurite dont.
I have heard a lot of false lighting stricks on the altas
-
.............
-
I'm assuming it's the same, but looks like you Weatherflow folks may be getting some love from the NHC.
https://www.nhc.noaa.gov/text/refresh/MIATCUAT5+shtml/060351.shtml
might give you insight into there network which has been around prior to the consumer releases https://weatherflow.com/professional-services/weather-networks/custom-designed-mesonets/the-weatherflow-hurricane-network/
Ah.....
-
Ah.....
The experience from operating a commercial weather observation network as their "day job" in remote coastal locations has trickled down into the design decisions for their consumer stations, such as not using spinny cups and tipping buckets that need occasional attention. I'd be willing to bet that they also have more meteorologists and data scientists on staff than any other PWS manufacturer.....
-
The "Ah" was not meant as a slight, as you apparently took it, but as an acknowledgement of WF's involvement as I knew nothing of it. Remember, I'm the one who brought this to attention in the first place.
-
Ah.....
The experience from operating a commercial weather observation network as their "day job" in remote coastal locations has trickled down into the design decisions for their consumer stations, such as not using spinny cups and tipping buckets that need occasional attention. I'd be willing to bet that they also have more meteorologists and data scientists on staff than any other PWS manufacturer.....
problem is meteorologist and scientist wouldnt or cant squeeze all that knowledge into a $300 consumer product otherwise they would have no professional based industry to sell or provide too .
.. ps im not rubbishing before mr rogers misinterprets but hey we can rubbish accurite thats ok .. but hey lets not talk or discuss about the rain because im still so unconvinced on its consistency and recent methods almost two years on now its still most common issue raised on there forums. [ You are not allowed to view attachments ]
as for lightning plop it near or in a vicinity of a gas combi boiler or when a neighbors lawn mower is running or someone welding nearby , i can trigger both of mine with my zippo lighter and many other methods . best one is a put a sky solar panel on it . but i get it not everyone has the same outcome but thats why i go on about consistency from one report to another from one day to another. best position i found was horizontally mounted i use two of them purposely for lightning not temperature.i have reasonable results of approaching storms long before i can hear them.
anyway show me the consistency in rainfall measurements across a wide user base of weatherflow consumer hardware and if you got about 10 hours read this public thread and look for any consistency from report to another ..there are reports from same users one day it is all good best thing since the invention of the wheel the same users next day not so good .. has we say in turkish tutarlılık yok .... :-k :-k https://community.weatherflow.com/t/rain-update-june-2019/3699/172
-
I have seen all the weather stations around me all have slight differences and a weather station is only as good as the best location as no one in a city can really have.
Many are placed on a roof. The wind maybe correct, rain will be off, dew point and temp.
Have it lower down the temps, dew points, rain will be more correct, so unless you got the land to set things up properly your not going to get consentant results.
I know all the 5 in 1 stations its recommended setting it to 10 feet as this gives the best all round balanced data.
I found with drizzle weather flow wont pick it up.
Normal rainfall my weather flow works well and pretty much spot on.
With even the bucket tipping are off at times, strong wind and heavy rain will give incorrect results.
There is no perfect system and the nearest is the old fashioned manual rain gauge that you got to manually empty.
My temps with weatherflow are inline with all the stations around me and within 1 degree of offical forcasts.
That is since I put my air inside an ambient radiation solar shield mounted on the south side of the house with ventilation all around it
I know weather flow intend to come out with a tipping bucket so it will catch drizzle and are committed to the sensor they use also.
If and when they come out with a tipping bucket the best of both worlds as weather flow indicates as soon as the first few drops start and something a tipping bucket cannot do and by the time it shows rain the ground allready has a good soaking of rain.
Also unless you buy a Davis, there is no other station offers quick wind updates. A 3 second wind update is impressive in the weather flow price range and Miles better than the Altlas's 10 second wind update which is only really giving an average, whereas weather flow gives more wind data and love when it does rain tells you how many mins it's been raining.
Love my weather flow; sure it's not perfect as no station is and certainly in a city cannot have the perfect spot and weather varies from 1 spot to another.
At work which is 6km away from home gets more rain than at home or can be the other way around so every station will vary. And looking at weather underground even every weathet station has some variations so who is to say your station is right or wrong.
If there are massive variations then yes it will be your station but minor ones and pretty much in line then it's safe to say for your area it's most likely pretty correct.
I know my weather flow is pretty much in line with the stations around me expect when its drizzling and then rely on my echowitt to indicate correct rainfall which is pretty accurate according to my manual rain gauge.
Having 2 stations fills the missing gaps like rain the weather flow wont pick up. But at times buckets dobt register rain properly either and why a manual gauge is the best backup
-
Acurite dont have accurite lighting stricks either. Lots of false stricks. They where worse when they first came out but still would not totally rely they are correct.
When the altlas came out lots off issues, but weather flow have a better and more dedicated team and have professional meteorologist, and team, whereas acurite dont.
I have heard a lot of false lighting stricks on the altas
Yes, we know there were some problems with lightning on some of the first production model. That was tracked down to a misrouted motor wire caused by a particular assembler at the factory. The problem was corrected, and Acurite committed to fixing any that were found in the future even after the warranty period. It was not a problem encountered by everyone. I have two production models in operation and have had no lightning strike problems at all.
I have one mounted near my weatherflow Air and the Air is basically "deaf" in comparison. It's interesting since they both use the AS3935 chip. My Atlas and the local TV weather station alerting me through my smartphone have both been accurate reporting lightning strikes in my vicinity. Not so with the Weatherflow Air for some reason.
Also, Keri Strenfel of Acurite is a professional meteorologist on staff at Acurite.
-
Also unless you buy a Davis, there is no other station offers quick wind updates. A 3 second wind update is impressive in the weather flow price range and Miles better than the Altlas's 10 second wind update which is only really giving an average, whereas weather flow gives more wind data and love when it does rain tells you how many mins it's been raining.
How are you getting 3-second wind updates from the Sky?
I'm running a stock Sky, Air, and Hub, but I'm only getting 1-minute reporting from the app, and 5 minute reporting from wunderground.
-
Easy dont have power saver turned on. Gives me 3 sec wind onscreen updates average wind and gives gusting.
Much better than acurite which I found there stations not as reliable as ambient or ecowitt which is better than ambient.
A wind speed of ten seconds is a joke I find but the 3cseconds of weather flow is miles better than the altlas.
Through the app I am getting a 3 sec wind update. The specs says you get a 3 sec update. If not there is something wrong with yours.
It will give an average wind speed updates every minute but it you look at your app the wind is showing every 3 secs plus direction.
For what your paying for the atlas a 10 sec wind update is far from good enough it's a joke. That put me off atlas right there and none other with fast wind updates bar davis
-
yes, its easy to get fast wind updating from weather flow :)
works great that :)
-
Easy dont have power saver turned on. Gives me 3 sec wind onscreen updates average wind and gives gusting.
Much better than acurite which I found there stations not as reliable as ambient or ecowitt which is better than ambient.
A wind speed of ten seconds is a joke I find but the 3cseconds of weather flow is miles better than the altlas.
Through the app I am getting a 3 sec wind update. The specs says you get a 3 sec update. If not there is something wrong with yours.
It will give an average wind speed updates every minute but it you look at your app the wind is showing every 3 secs plus direction.
For what your paying for the atlas a 10 sec wind update is far from good enough it's a joke. That put me off atlas right there and none other with fast wind updates bar davis
I don't have the power saver on.
Guess something is messed up somewhere.
My weatherflow has also been falling off the network often of late. Sometimes I have to completely reset the WiFi network information.... other times it just comes back on it's own.
I wouldn't call it "a joke", but weatherflow certainly hasn't been as accurate or reliable as I'd like from a weather station for general purposes. I like the technology direction though.
-
I think it does a great job, especially if you use a second windspeed/rain sensor set at ground level for rain
(they ideally need to have a separate rain sensor)
and the best thing is they are open to suggestions (weatherflow) and are wanting to improve
but you are not going to please everyone
but just because someone does not agree or does not like does not mean someone else will feel or think the same
-
One thing that I'm not real comfortable with are their "calibrations".
I can understand it from a research and development aspect to improve the equipment, but at a consumer level I'm left wondering if I'm really seeing "my" data or something that's more of a "group effort".
One example: "RainCheck". I can have it "On" or "Off", of course, but is there a way to compare my reading to the radar-derived reading?
-
but there can be a lot of variation in rain readings between rain gauges close to each other
and so an average over those is probably better
which is what is happening when they calibrate against other known sources and rain radar
at least they are trying to get the system improved and auto learning etc etc
kudos to them I say for doing all that
but there will be people who don't have the same opinion etc but that does not mean everyone will and should have that same opinion
-
I don't have the power saver on.
WF sends four different observation types via UDP and the web API. The first one is obs_air and obs_sky observation packets, and they are sent every minute for things like temperature and wind averages. The second one is rapid_wind, which is sent every 15 seconds (power save) or 3 seconds (power hungry). They call it rapid because it is sent rapidly, not that it is alerting to a wind gust or something. The third and fourth packets are for the start of rain and a detected lightning strike, and those are sent immediately whenever one of those events is observed. There are other packets sent for hub status and debugging.
If you are only seeing wind every minute, you are probably looking at obs_sky and not rapid_wind.....
-
but there can be a lot of variation in rain readings between rain gauges close to each other
and so an average over those is probably better
which is what is happening when they calibrate against other known sources and rain radar
at least they are trying to get the system improved and auto learning etc etc
kudos to them I say for doing all that
but there will be people who don't have the same opinion etc but that does not mean everyone will and should have that same opinion
This kind of blurs the line between honest, independent data and just taking a third-party "best estimate" using radar.
It's sort of like what wunderground used to be doing with pressure data... ie, they think your pressure reporting isn't accurate, so they started adjusting it in a "black-box" fashion. That did not go down well with folks who were taking great pains to be accurate with their pressure data.
I can understand it being used to tweak the accuracy of the haptic sensors during the development phase, but the way they are presenting "Rain Check", it sounds like this is a "permanent" solution for poorly sited or calibrated sensors.
-
I don't have the power saver on.
WF sends four different observation types via UDP and the web API. The first one is obs_air and obs_sky observation packets, and they are sent every minute for things like temperature and wind averages. The second one is rapid_wind, which is sent every 15 seconds (power save) or 3 seconds (power hungry). They call it rapid because it is sent rapidly, not that it is alerting to a wind gust or something. The third and fourth packets are for the start of rain and a detected lightning strike, and those are sent immediately whenever one of those events is observed. There are other packets sent for hub status and debugging.
If you are only seeing wind every minute, you are probably looking at obs_sky and not rapid_wind.....
I've seen the 3-second data on the network. I'm just not seeing it in the app or wunderground.
I've been running on the assumption that they were throttling due to server load.
Something is amiss. I just took a look at my records, and it seems since May I have had many more Weatherflow outages than I realized. A few times I've had to re-enter my Wifi info by hand to get it running again, but apparently it's dropping more frequently and coming back on its own.
Oddly, the outages seem to start shortly after I added the solar panel to the Sky.
-
I think it does a great job, especially if you use a second windspeed/rain sensor set at ground level for rain
(they ideally need to have a separate rain sensor)
and the best thing is they are open to suggestions (weatherflow) and are wanting to improve
but you are not going to please everyone
but just because someone does not agree or does not like does not mean someone else will feel or think the same
I second that mine is right up there with my ecowitt and Davis vue and find mine accurate expect for drizzle which it wont pick up
Drizzle for us is a very rare event. Even compared to every other station in my area my weather flow is as good as any of them.
Remember lots of people mount theirs on the roof so every condition expect wind will be incorrect and rainfall will be well out. My ecowitt ( ambient 2029a) clone does a good job of rainfall compared to my manual rain gauge but wind updates are poor like the altlas, and even mounting the altlas on your roof is only giving average wind speed not actual wind speed.
That's the beauty of the wind flow a 3 sec actual wind speed.
Everything I read about 5 in 1 most are mounting them wrong anyway which will account to variations.
They always recommend 10 feet and this will give the most balanced readings.
No 5 in 1 should be on the roof, at least ambient have fixed this idea where you can extend the cups and wind direction miles higher than the station and why altlas members dont use this feature beats me.
And has been said you can get a sperate sky unit and place it at ground level to give more accurate rainfall.
-
The Atlas sampling window is also about 3 seconds. (The 5n1 window was 4).
The highest wind value recorded during the 10-second reporting interval is then sent.
It's not an average over the reporting interval.
-
Yeah but does not update to your screen every 3 secs like weather flow. Only updateams wind speeds at 10sec interviles. Still not good enough in my opinon
-
I've seen the 3-second data on the network. I'm just not seeing it in the app or wunderground.
The app and web site both use the 3-second rapid_wind info to update the icon on the left side of the wind card. I'd grab a screen capture of mine, but ironically there's a hurricane spinning off the coast and we're not seeing as much as a draft here right now.....
-
Just joined the Weatherflow club. If I scanned this forum more carefully, I would have known about these sooner, as I see lots of geeks did.
The tech is pretty cool, have to admit. No doubt lots of questions about accuracy.
I have a Davis Pro II, way up on steep high roof where, apart from the winds, it shouldn't be. I have long been dissatisfied with its siting.
So, the Weatherflow "Sky" sits down on a fence top between 2 houses and trees nearby. The winds shouldn't be right on the money there whether the tech works or not. The haptic rain sensor? Holy cow, I have no idea about that, but I love the idea of it. I have a Stratus manual rain gauge right next to the Sky, which has in the past reported more rain than the roof Davis, which it should down at the 6 foot height (except for surrounding houses, etc). So I will get a pretty good comparison for rain amounts.
The Weatherflow looks like fun to me. If nothing else, it's great to see someone trying out new weather tech on us.
-
now that's the spirit!
-
Had my first experience with "Rain Check" last night.
Went to bed with my WF showing about .26 inch rain. Got up this morning, with no more rain, and it was adjusted up to .42.
My roof Davis, which under-reports due to being over 25 feet up, showed about .46 inch of actual measured rain.
My Stratus manual tube, on the fence 2 feet from the WF, showed .32.
Gosh, I don't know what exactly to believe. Or maybe there's just not anything to actually believe.
If my WF is not able to actually measure the rain, then why do I need it? I can probably just get a number from the same source software that WF uses, right? https://metstat.com/real-time/qpe/
-
Had my first experience with "Rain Check" last night.
Went to bed with my WF showing about .26 inch rain. Got up this morning, with no more rain, and it was adjusted up to .42.
My roof Davis, which under-reports due to being over 25 feet up, showed about .46 inch of actual measured rain.
My Stratus manual tube, on the fence 2 feet from the WF, showed .32.
Gosh, I don't know what exactly to believe. Or maybe there's just not anything to actually believe.
If my WF is not able to actually measure the rain, then why do I need it? I can probably just get a number from the same source software that WF uses, right? https://metstat.com/real-time/qpe/
Prescinding the siting issues, it's hard to beat an old fashioned rain gauge for truth.
You've hit exactly my concern with "rain check". I'm not sure if it's a temporary fix or final solution.
I'm hoping they are still trying to make the rain sensor work instead of throwing in the towel.
Maybe it's just not a great idea to mix the rain and wind sensing in one unit?
-
For the past year I've been testing out different precipitation sensor technologies. This includes radar (Lufft R2S, WS700, WS10) and piezoelectric/haptic (Vaisala WXT536,WeatherFlow).
TLDR: Get a bucket sensor
Long version:
None of the radar-based devices provided even remotely accurate data. I just gave up on them. The Vaisala WXT line and the WeatherFlow devices appear to use very similar if not the same technologies. Both rely on interpreting the impact of rain drops and overall these do perform better than radar-based sensors. Vaisala claims their sensor is 95% accurate but this is BS and they know it. The WXT536 even allows you to connect a rain bucket. That being said, the Vaisala sensor *is* more accurate than the WeatherFlow sensor (but not by much). My guess is that the slightly improved accuracy isn't due to a better algorithm but rather due to better construction. The Vaisala unit has a very sturdy stainless steel rain sensor surface.
So the bottom line is that from what I've seen with hands on experience, none of these new technologies can beat a bucket in terms of accuracy. WeatherFlow should consider selling a wireless rain bucket.
-
I find mine with steady rain mine has been within .2mm compared to my manual rain gauge and we dont even have raincheck here. So most of the time pretty accurate
-
said many times but unfortunately its not all too apparent straight out the box its when you go through seasonal changes and occurrences of different rainfall events the issues start to arise.overall haptic sensor technology truly is not proven for consistency in rainfall measurements , i tested 6 weatherflow units over a 2 year period the biggest failing was rainfall measurement in every device all it had going for it was rain detection start after that it was in no mans land . piezo component sensor technology is also somewhat questionable when used for wind measurements again rain is a factor and so is sound ... and the future dont look bright in part 2 on the rain front but there is always hope for a surprise around the corner . ;-) ;-)..
the 20% -+ specification in the weatherflow hardware is way too high and the trained eye will pick up on that
its a bit like having your electric meter readings done with a device that has a 20% -+ tolerance you could either pay upto 20% more or 20% less or an hifi amplifier that has the right channel running -+20% you end up with an unbalanced sound ..-+20% is a considerable margin...
the keyword is consistency and sadly in the case of weatherflow rain performance it isnt quite there yet but i dont think their ready to throw in the towel just yet and has for rain check said the joker to the thief there’s too much confusion.....
-
Unfortunately nothing beats a manual rain gauge. Even buckets cannot keep up with the tipping in heavy rain.
They need a system based on a a manual rain gauge that measures rainfall and empty after 12am or a time you select.
-
Well, the Met Stat thing is interesting. Those people have a professional, young staff of engineers and meteorologists that put out a complex measuring formula.
I would like to just have their raw estimate of how much rain I got as a baseline to check my other gauges. With Rain-Check, I guess that is basically what I am getting, so I can look at the WF at least as a paid subscription to their analysis.
The WF UV and brightness are a nice addition to my weather data, and there is not much reason to suspect the "Air" part of the system. My Air sits in permanent shade on my back deck. It is nice to know that it's not as cold at night or as hot in the daytime as my aspirated, shielded Pro II sitting up on the roof at over 25 feet.
My "Sky" is too low and between houses has little chance of giving me accurate winds, so I can't judge it. But it should get ballpark rain totals, better than a 25 feet up tipping bucket. But I guess Rain Check is the only way it can do that. I should keep track of Sky's raw rain total compared to it's RC adjustment bumps.
-
If you want to see a closer side-by-side, here's my custom WeeWX report comparing the VP2 with the WF Air/Sky:
http://ae3.homelinux.net:845/weewx/compare.html (http://ae3.homelinux.net:845/weewx/compare.html)
(https://images.hvtraffic.com/images/2018/03/18/20180318165938-647049e9.jpg)
(https://images.hvtraffic.com/images/2018/03/18/20180318170938-cf049f79.jpg)
-
That looks better sited than my WF. I knew I wouldn't be accurate on wind with the WF. But I would like to see really good comparison between my Stratus manual gauge and WF, they are 2 feet apart on the same fence. I might try to get them even closer.
The Rain-Check does get the rain total in the ballpark, so I guess it's good WF went to it.
Siting is the biggest challenge for the suburban weather enthusiast. I have no great options.
-
Loved those 2 stations on the same chart comparisons. I would like to do that, but WeeWx appears a bit over my head. I have never used Linux and know nothing about it.
-
as i said the other day part 2 or round 2 its called tempest and is another crowdfunding kickstarter project..
quote from wf
Tempest features a handful of hardware refinements, as well as the introduction of AI-powered forecast technology which will dramatically improve forecast accuracy at your house. All WeatherFlow station owners will benefit from the new AI-powered forecasting which will be rolled out as an app update in Q’1 2020.
[ You are not allowed to view attachments ]
-
I'm very curious about the hardware changes. Hopefully they'll list the sensors used in the new model so I don't have to buy one just to tear it apart.
That said, it appears they're sticking with the same wind and rain tech. I was hoping for a different direction for precipitation.
-
I'm very curious about the hardware changes. Hopefully they'll list the sensors used in the new model so I don't have to buy one just to tear it apart.
im sure in time if the adventurous testers are brave enough all will come to light , but note its an all in one solution so expect compromise in some specific readings . as always all in one solutions generally have some trade off unless you have a perfect urban environment. on first impression its a step forward but the seasonal changes will reveal just how good it is over the original.
i still hope for a optional rain solution not for false rain events issues but for relatively true accumulation values especially during long prolonged rain periods . if rain check thingy is still a part of the product then ill reserve its rain validation as suspicious..
note i see no means or mention of upgrading existing product other than signing up through the kickstarter campaign, perhaps weatherflow will offer some solution in the future i.e price for upgrading for existing users or may be not ..im sure some will feel hard done by in just over a year of original version being readily available and now something completely different , thats why you got to just love Davis hardware approach of upgrading and introducing improvements..
-
I understand people's reservations about Rain Check. But the literature on it does point out that the haptic system will continue to calibrate, "learn" with the help of rain check, each time a rain event occurs.
To the extent that this is really happening, I like it. I can't really calibrate my Davis bucket tipper unless I want to apply a constant offset, which I doubt many people go for.
I am confused about this new "Sky" pictured. This is a new model (heck I have had mine a week and they didn't send one of those) or is this in the testing phase?
-
I understand people's reservations about Rain Check. But the literature on it does point out that the haptic system will continue to calibrate, "learn" with the help of rain check, each time a rain event occurs.
To the extent that this is really happening, I like it. I can't really calibrate my Davis bucket tipper unless I want to apply a constant offset, which I doubt many people go for.
I am confused about this new "Sky" pictured. This is a new model (heck I have had mine a week and they didn't send one of those) or is this in the testing phase?
first and foremost rain check is not available to anyone outside USA and calibration long term is unproven I've seen good reports and bad reports,ive seen requests for it to be disabled and so on .
yes a brand new model but yes I guess in testing phase and no means to upgrade existing product unless they will offer an exchange program of some sort best liaise with them but the lack of upgradeability always concerned me going forward and not being in the USA the exchange or replacements, timeframe can be a costly exercise has they have no control import and local regulations, taxes and so on been there 6 six times at around $50 a time just for a replacement or test unit sent.
positives works with existing hub :-) for current owners
-
I'm glad this was posted, and now I see the word about the Tempest on their site. It's frustrating that a new "Sky" is announced just over a week after I purchased the system.
I hear you, regarding the lack of Rain Check overseas, that has to be remedied.
I haven't had much rain so far to evaluate, but the one time there has been some rain, Rain Check brought the total from .26 up to .46. I will be watching each time the total is adjusted after midnight to see how much is changed.
-
I'm glad this was posted, and now I see the word about the Tempest on their site.
Can someone provide a link to that?
Thanks!
-
I'm glad this was posted, and now I see the word about the Tempest on their site.
Can someone provide a link to that?
Thanks!
Found the announcement on their community forums: https://community.weatherflow.com/t/new-tempest-weather-system/4444
Sent using Tapatalk
-
I would not go with kickstarter as your beta testing the hardware and software and have the most problems. When it's out of kickstarter and in production the hardware and software usually have refinements
-
They are saying they will give existing WF owners a discount on the Tempest when it comes out. Don't know if that means you keep the old Sky or exchange it.
-
Would be tricky for those who live outside the usa? Would proberly cost more to send back than the discount is worth?
-
I noticed this bit from the WF staff:
"the functional difference would be nominal"
I have a feeling there's not going to be much progress in the hardware rain sensor accuracy. This is my guess on what the changes are/will be:
1) 360 degree solar panels
2) Integrated pressure, temp, hum, lightning sensors
3) Newer versions of existing sensors. For example, I believe the combo uv/light sensor they used in the sky is now EOL.
So yeah, some solid improvements but I think this is more about streamlining production and getting costs down (if the Air can indeed be deprecated).
Honestly, if they would just offer an optional wireless rain bucket unit then this would be a fairly solid consumer level weather station. I don't need/want AI to tell me what's going on in my backyard. I just need decent sensors.
-
Honestly, if they would just offer an optional wireless rain bucket unit
I realize this is not what you're after, but it is a quality option to at least view rain from inside the house.
https://rainwise.com/products/detail.php?ID=6692&Category=Rain_Gauges:Wireless&pageNum_cart=/products/index.php
-
1) 360 degree solar panels
Those seem odd to me. I guess it means they will work at any latitude, but they must be super-efficient given their size.
-
Think integrated temp sensor is a step backward. Having sky and air separate makes WF much more versitile
-
Those seem odd to me. I guess it means they will work at any latitude, but they must be super-efficient given their size.
A few owners purchased spare solar panels for their Air units (which are supposed to be in the shade), and discovered that they charged adequately from even indirect sunlight. Given this observation and also that they sell to folks south of the equator and have to make provisions for solar panels to face north for those locations, the 360 degree solar panels sorta make sense.....
-
So the Air just becomes surplus equipment I guess.
Apparently, they are combining all functions into the Tempest, along with all their excitement about AI forecasting. I'm guessing, judging from comments guys have made, they are not as worried about the forecast as they are about the accurate measuring of wind and rain.
-
A hugh set backward having all units built in one system. At 30 feet your not going to get accurate rain and temps. This will have the same issues as the 5 in 1 stations. Only get offical wind readings but everything else is wrong.
-
WF says they can help you calibrate your haptic rain sensor if you are able to place a manual gauge next to it for comparisons.
-
WF says they can help you calibrate your haptic rain sensor if you are able to place a manual gauge next to it for comparisons.
Are they not calibrated at the factory?
-
in reference to rain nope whole rain thing is a mess , is it a total lack of understanding? also uv is questionable and humidty can be questionable if you look at vreihen post a few back the humidity flat lines or maxes out at 90% whilst his vp2 fluctuates between 90 and 93% on that particular day . personally all this AI talk or Continuous learning is over my head sounds more like silicon valley tech talk. so called continuous CL has never worked on any of the units ive had , especially if you have a reboot for whatever reason earlier this month I saw uv8 on a non cloudy quite a way off for this time of year 5-6 is possible beginning of october when directly overhead. so i couldn’t attribute it to cloud reflection..
so lets see how this new pretty looking thing pans out who knows may be it will put us skeptics to rest.. :grin:
-
I think the Tempest has few changes if any to the rain sensor or the wind sensor. It's mostly the inclusion of the "Air" stuff into the "Sky" unit. Then they will have some AI forecasting software, but that doesn't affect any rain sensing.
I'm not in the know about how the haptic sensing is supposed to work. I simply noted on their community forum that for a person complaining that his rain totals were wrong, they said they could adjust, "calibrate" it further and especially if he could put a manual gauge next to it so they have a better idea about where the readings need to be.
I have a manual gauge just a foot from my Sky, and I may end up taking them up on the manual calibration. But so far, the Rain Check did subtract off the r
"rain" that was recorded when I moved the Sky closer by 2 feet to the manual gauge. We haven't had much rain at all here in north central Texas for 6 weeks for me to make any other rain comparisons.
I just ignore the low wind readings on my sky because it's only at 6 or 7 feet above the ground and between houses. I can just let it go because I have an annemometer (Davis Pro II) up on the roof at 30 feet, and it gets LOTS of wind.
-
WF says they can help you calibrate your haptic rain sensor if you are able to place a manual gauge next to it for comparisons.
Are they not calibrated at the factory?
They're supposedly calibrated by CL, "continuous learning", and they claim that the haptics will "learn" from each new rainfall, with the input from "Rain Check" to get better and more accurate readings. I'm merely repeating what I read from their website.
-
as i said its gone over my head , a manual gauge so put that into perspective how does the sky get its continuous data or is it just calibrated after an event then you pass it on .
haptic sensor look up piezoelectric because thats the sensor type it measures vibrations or sound ive yet to fathem which one . i have used an identical component I used on a acoustic guitar diy pickup.. took me a while to realize why the sensor looked so familiar when i first took a sky unit apart. i guess its what goes on after sensor picks up a vibration and attempts converts it to a measure but im wondering or guessing if there is some what if scenarios in the formula.
-
We got a big rain event in the DFW last night and I was ready with my rooftop mounted Davis Pro bucket tipper (up at 30 feet) and a Stratus manual gauge down at 6 feet with the Sky 1 foot from it.
The rain was very heavy for some time, and as long as that was the case, the Sky was doing a great job, right in line with both the Stratus and the roof gauge. In fact, Sky got out to a .10 in lead over the too high Davis, as one might expect.
But as the rain intensity settled to a slower, steady precip, the Sky started to lag. I watched it give back it’s .10 and eventually more. This morning, a check of the manual gauge confirmed what I saw.
Roof, 2.47 inches Stratus manual gauge, 2.60 inches Sky, 1.98.
Just now, I see that Rain Check raised the Sky to 2.16 in. Better, but still significantly below the 2.60 inches of the Stratus 1 foot from it.
-
The rain was very heavy for some time, and as long as that was the case, the Sky was doing a great job
...
But as the rain intensity settled to a slower, steady precip, the Sky started to lag.
My Sky unit exhibited the same behavior. It was fairly accurate during heavy downpours but anything lighter and the rain total was always lower than what it should have been.
Interestingly, I have a Vaisala WXT536 which uses a very similar (if not identical) rain sensor technology. It behaves the same. Works okay during heavy rain but under-reports during lighter to moderate rain.
WeatherFlow has pointed to unwanted vibration as a reason for inaccurate rain readings but the WXT536 is built much more firmly (with a steel sensor plate) and it still exhibits the same behavior.
After trialing both of these units (Sky and WXT536) I honestly think the haptic technology just isn't suited for rain detection. This is what has me scratching my head a bit with the Tempest. Either they've vastly improved the haptic technology or they're a bit crazy for sticking with the same flawed rain sensor tech. Time will tell but I'm not going to be at the front of the line to try it out this time around.
-
Just now, I see that Rain Check raised the Sky to 2.16 in. Better, but still significantly below the 2.60 inches of the Stratus 1 foot from it.
Here in Round Rock my Sky read about 0.2" too high. Rain Check brought it down to 0.1" too high. (It was 1.57" according to my Atlas)
I wish Rain Check would note how much it corrected to judge the Sky's performance. Is there anywhere you can see that?
-
There's an ASOS (KMGJ) across town to the west, but it is in an open area with crazy wind exposure and I'm lucky to see 29 MPH peak gusts. There's also KSWF about the same distance to the east.
WeatherFlow just added the barometer to their continuous learning (CL) algorithm in the past week or so, and I just peeked at the ASOS to see that I was reading 30.086 in the WF app when the ASOS was reading 30.07. (weewx was reading 30.06, because I had a pre-CL adjustment hard-coded into weewx. I have made no attempt to calibrate the Davis barometer.
https://smartweather.weatherflow.com/share/910/grid (https://smartweather.weatherflow.com/share/910/grid)
One can never have too many weather stations...and they all basically read the same thing most of the time:
(https://images.hvtraffic.com/images/2018/09/25/20180925175422-0b9fad1b.png)
(https://images.hvtraffic.com/images/2018/07/02/20180702123658-66151ea2.jpg)
I'm even more curious than you are about if the ultrasonic is going to drop out in 20+ MPH winds and moderate/heavy rain.....
I've been watching several of these stations with in 20 mile radius of my house and seem to be with in a degree for temp., hum. with in 3%, barometric Pres. usually a little lower .01-.02 inhg. rain seemed to be with in .02 and wind direction pretty close. Anyone have any idea on what the new Tempst stations will sell for when released?
Thanks,
twetherman
-
Anyone have any idea on what the new Tempst stations will sell for when released?
From the Kickstarter campaign "Save $X off retail price" discounts, I calculate:
Base kit w/WiFi Hub: $329
Storm/Fire kit w/cellular Hub and 7-day Hub backup battery: $399
Remote/Autonomous kit w/cellular Hub, weatherproof solar enclosure: $449
If you want in for less, it appears that there were four cancellations on the $229 deal on Kickstarter that are again up for grabs.
FWIW, I am in on the Kickstarter full Remote/Autonomous kit, which I plan to install on some wooded mountaintop property that my family owns because the place has interesting weather.....
-
A new WeatherFlow review was posted today...from the review, this looks like a promising weather station:
https://www.techhive.com/article/3563348/weatherflow-tempest-review.html
-
A new WeatherFlow review was posted today...from the review, this looks like a promising weather station:
https://www.techhive.com/article/3563348/weatherflow-tempest-review.html
This quote from that review.
"Not only that, WeatherFlow’s haptic rain sensor can also measure the intensity of rainfall. I can only wonder if it could be upgraded down the road to detect precipitation types as well. That, however, involves costly and complicated technology. Whatever the case, I found rainfall measurements generally accurate (further testing is needed here to make a full determination), so WeatherFlow’s new method of rainfall measurement gets a thumbs up from me at the moment."
The fact that his control data is a nearby NWS station is crazy. Specifically when it comes to rain data. His quote "I couldn’t spot any inconsistencies in the data from the control data I used, which is a nearby National Weather Service observing station", is hogwash. How do you in good faith write a review with control data like that.
Listen I'm not saying this station is bad. I know he said it needs more testing but to write a review without side by side testing is ridiculous and to give it an initial thumbs up is crazy.
-
I agree not only is it hogwash but one of the ways that WeatherFlow was able to make the haptic rain sensor more accurate was to implement cloud AI into the equation meaning the rain data is being adjusted based on other crowd sourced data. So if the rain matches NWS data...well then ding ding ding! There is a clue. :roll:
There is no console with this station. You can't see the data directly from the sensors. The data comes from the cloud after it has been manipulated.
I could be wrong. But color me unimpressed. I really haven't looked that much into it to really know for sure. Please correct me if I'm wrong.
-
Just curious what are the specs. on this unit? I couldn’t find any information on this from their website.
Thanks,
tweatherman
-
TechHive article link posted by MacGarage says that final specs have not been released. But I haven't seen any specs.
Pay no attention to the man behind the curtain.
-
Just curious how a weather station of that magnitude doesn’t have specs. published yet? Will not even consider buying until specs. are published.
Thanks,
tweatherman
-
TechHive article link posted by MacGarage says that final specs have not been released. But I haven't seen any specs.
That click-bait article was full of misinformation and lies...from the very first sentence:
I’ve been trying to get my hands on a WeatherFlow Tempest for years. When I first spotted the device on Indiegogo in late 2016 (that’s not a typo), it sounded like nothing else available at the time.
The Tempest wasn't even announced for pre-production backing until November, 2019!!!
Here's some more info that the alleged author apparently did not find:
https://www.kickstarter.com/projects/weatherflow/tempest-a-revolutionary-personal-weather-system (https://www.kickstarter.com/projects/weatherflow/tempest-a-revolutionary-personal-weather-system)
(https://ksr-ugc.imgix.net/assets/027/215/426/fbd7e60399b58fb3d980758b207f6c08_original.png?ixlib=rb-2.1.0&w=680&fit=max&v=1573751585&auto=format&frame=1&lossless=true&s=6e740c1b4193265b092666d2aa191642)
-
Not bad specs. but a little on the lagging side for update for temp. and humidity. I also wish the temp. would go down to at least -40F.
Thanks,
Tim
-
Rain accuracy specs (lower % is better):
Tempest: ±10%
Ecowitt WH40: ±5%
Davis VP2: ±4% (±5% when rate is +2in/hr)
Acurite Atlas: ±5%
-
Rain accuracy specs (lower % is better):
Tempest: ±10%
Ecowitt WH40: ±5%
Davis VP2: ±4% (±5% when rate is +2in/hr)
Acurite Atlas: ±5%
You forgot a couple, one tipper, one static...Rainwise 2%, Stratus 1%.
-
IMO rain measurement is the Achilles heal of the WF smart weather (original model) system. Between .25 and .75 inches, my collocated Acurite 5/1, Stratus gauge, and WF smart will often show almost identical readings for rain. In other scenarios, the WF is not dependable.
Rain measurement was also the prime complaint of posters on the WF forum IIRC. The Tempest model, when first described there by WF employees, had lots of enticing, very generous warranty and satisfaction guarantees (they don't mention these much in subsequent announcements or literature, AFAIK). I think there was a very long time frame for return for refund no questions asked. However my WF enthusiasm is sufficiently dampened that I'll wait to read lots of real life performance reports, here or in the WF forum, no matter what the warranty or purchase price discounts, before I would (ever) buy a Tempest.
I'm sure the WF folks are fine people. But it did bug me when they all but acknowledged that the haptic system is not accurate, and then kept everybody on alert for months, waiting for the promised magical software fix. That turned out to be the system that uses some kind of crowd source measures from nearby supposedly accurate sources, and that means that for accurate measuring rain at my location, the WF is of almost no use.
Weatherflow reps have even written in the forum that a mechanical rain measuring system is being entertained as a future option, which sounds to me like another way of saying that the issues with the haptic sensors are here to stay.
Here's what I do like: the AIR half of the system, for temp/RH and lightning works fine and was convenient to mount in a shady location. I was disappointed that WF integrated everything into one unit for the Tempest, since this means that the jury is still out on how accurate the Tempest temp reads will be over time. They claim to have built compensation into the software, so I have to wonder if this is not yet again some kind of crowd-sourced data, in which case I don't want one.
The WF mounting system is great for somebody like me who has very limited mounting locations. I'd wish that every manufacturer would make their station attachable to ready-made camera-mount hardware.
-
Plus or minus 10% on rainfall accumulation?
That's a joke. How can you deliver a product like that.
It's a great concept but one that needs a lot work.
-
Rain accuracy specs (lower % is better):
Tempest: ±10%
Ecowitt WH40: ±5%
Davis VP2: ±4% (±5% when rate is +2in/hr)
Acurite Atlas: ±5%
You forgot a couple, one tipper, one static...Rainwise 2%, Stratus 1%.
Which is for right now why I'll stick with Stratus.
-
Rain accuracy specs (lower % is better):
Tempest: ±10%
Ecowitt WH40: ±5%
Davis VP2: ±4% (±5% when rate is +2in/hr)
Acurite Atlas: ±5%
You forgot a couple, one tipper, one static...Rainwise 2%, Stratus 1%.
Which is for right now why I'll stick with Stratus.
Which is why I have both, about 7" apart. [tup]
-
Which is why I have both, about 7" apart. [tup]
Only two????? :lol:
(https://images.hvtraffic.com/images/2018/07/02/20180702123658-66151ea2.jpg)
.
-
Two's aplenty. You need a RW tipper so you can throw the white thingy and black thingy away. Keep the clear one. ;)
-
Doesn't Randy have about 5.
-
Two's aplenty. You need a RW tipper so you can throw the white thingy and black thingy away. Keep the clear one. ;)
That's a two year old picture. If I took a picture today, there is also my Blitzortung H-field antenna and a Rainwise 111 tipper (connected to an ESP32/Arduino with custom software) out there.....
-
Two's aplenty. You need a RW tipper so you can throw the white thingy and black thingy away. Keep the clear one. ;)
That's a two year old picture. If I took a picture today, there is also my Blitzortung H-field antenna and a Rainwise 111 tipper (connected to an ESP32/Arduino with custom software) out there.....
Interested in that custom software for the RW gauge, have you got any more info published anywhere on it?
-
Two's aplenty. You need a RW tipper so you can throw the white thingy and black thingy away. Keep the clear one. ;)
That's a two year old picture. If I took a picture today, there is also my Blitzortung H-field antenna and a Rainwise 111 tipper (connected to an ESP32/Arduino with custom software) out there.....
=D> Even after "Binging it" ( I don't do Goggle anymore) I'm not sure what you have. I'm clueless on this stuff. Does that mean you don't use your VP2 ISS as the transmitter for the RW, but instead you use something like a bluetooth to get the data indoors?
-
Two's aplenty. You need a RW tipper so you can throw the white thingy and black thingy away. Keep the clear one. ;)
That's a two year old picture. If I took a picture today, there is also my Blitzortung H-field antenna and a Rainwise 111 tipper (connected to an ESP32/Arduino with custom software) out there.....
=D> Even after "Binging it" ( I don't do Goggle anymore) I'm not sure what you have. I'm clueless on this stuff. Does that mean you don't use your VP2 ISS as the transmitter for the RW, but instead you use something like a bluetooth to get the data indoors?
Even better than Bluetooth, the ESP32s have WiFi
-
=D> Even after "Binging it" ( I don't do Goggle anymore) I'm not sure what you have. I'm clueless on this stuff. Does that mean you don't use your VP2 ISS as the transmitter for the RW, but instead you use something like a bluetooth to get the data indoors?
Even better than Bluetooth, the ESP32s have WiFi
Why indeed they do! :twisted:
Each Rainwise 111 bucket tip is 0.01", and counting tips with software debouncing is a basic Arduino learning project.
I have several WeeWX instances running. One has the Davis VP2 via a USB Envoy, one for the old Acu-Rite, two for WeatherFlow stuff, and one just for the Rainwise tipper using a butchered version of my WeatherFlow UDP station driver. It is interesting to compare all of the stations. (On a side note, the black aerocone thingy and the clear thingy track each other surprisingly well.)
The ESP32 that I bought also has a LoRaWAN 900Mhz radio on-board. While counting tips is a basic Arduino learning project, doing it running on a small battery requires putting the ESP32 into a deep sleep and only waking it up via interrupt whenever the bucket tips. What I found is that it takes 10-15 seconds to wake up, re-establish a wifi link, send the UDP packet, and go back to sleep. LoRaWAN is popular in farm field weather sensors, because it has long range and good battery life compared to wifi.
Long story short, the ESP32/wifi code is not not share-quality, and I'm not doing anything else with it. Once COVID-geddon is over and my workload returns to semi-normal, I'm going to focus on getting it to work between two ESP32's over LoRaWAN with a goal of at least one year of battery life at the tipper. THAT will be share-worthy..... ;)
-
Not to jump in here but has anyone here received the new Tempest station and had a chance to compare temp. and humidity readings vs. a Davis or other higher quality stations on the market. I have a hard time believing that the radiation shield on the new design will keep full hot afternoon temperature spikes down. Any input is greatly appreciated.
Thanks,
tweatherman
-
=D> Even after "Binging it" ( I don't do Goggle anymore) I'm not sure what you have. I'm clueless on this stuff. Does that mean you don't use your VP2 ISS as the transmitter for the RW, but instead you use something like a bluetooth to get the data indoors?
Even better than Bluetooth, the ESP32s have WiFi
Each Rainwise 111 bucket tip is 0.01", and counting tips with software debouncing is a basic Arduino learning project.
Wow. I'm soooo glad that all I had to do is plug my RW bucket into my VP2 ISS. #-o
GL with your project!
-
I have a hard time believing that the radiation shield on the new design will keep full hot afternoon temperature spikes down.
Yep.
-
a station ive been monitoring for some time outside of USA ... Singapore..
temperature differences vary higher 1-2c during daylight hours
rain variances always noticeable hard to draw any real conclusion on accuracy as tempest often stops recording rain (is this due to perhaps intensity ? )
wind readings very erratic to no readings (sensor failed ? or just stops recording due to power efficiency below threshold ? is there some feature to preserve power if no sufficient solar replenishment ? or the old common problem water on the sensors? )
1st observation I noticed last year solar panel orientation its common knowledge a vertical solar panel has less efficiency upto 40% less efficiency over lets say a 30 to 45 degree angled design and could be noticeable due to the low capture area .interesting to know if any reports of failed units due to lack of solar replenishment efficiency .
think temperatures look reasonably ok I never see large spikes but the other issues are often noted .
two quick screenshots taken with in seconds of each other
tempest
[ You are not allowed to view attachments ]
vantage pro2+
[ You are not allowed to view attachments ]
-
Not to jump in here but has anyone here received the new Tempest station and had a chance to compare temp. and humidity readings vs. a Davis or other higher quality stations on the market. I have a hard time believing that the radiation shield on the new design will keep full hot afternoon temperature spikes down.
The function still has not been enabled, so no comparisons are available yet.
With all of the data scientists working at WF on the company's "day job" (forecasting and modeling), this functionality seems attainable. They have the unit's mass, a solar radiation sensor, and a wind measurement to work with, and that should be enough to model reasonable temp/hum adjustment curves for most conditions.
If I was a betting man, I'd put money on WF's Tempest temp/hum AI being more accurate than the CWOP QC function when it's finished, and light-years ahead of WU's joke gold stars.....
-
I guess another big question is anyone biting the bullet and ordering one of these here from the forum?
Thanks,
tweatherman
-
I backed the Kickstarter last November, for the complete remote off-the-grid package with a solar-powered Hub and cellular interface. I'm hoping to put it on the mountain top land that my family owns in a nearby state, which is known to have some, umm, "interesting" weather. My first-generation WF hardware has more or less lived up to their promise of being almost maintenance-free (mostly battery changes), thanks to the elimination of moving parts. It should be interesting to see how often it needs to be visited...and how long it lasts when the first T-storm rolls through! #-o Anyway, they won't be shipping cellular units until probably late summer, so this will be a fall project at the earliest.....
-
Not to jump in here but has anyone here received the new Tempest station and had a chance to compare temp. and humidity readings vs. a Davis or other higher quality stations on the market. I have a hard time believing that the radiation shield on the new design will keep full hot afternoon temperature spikes down.
If I was a betting man, I'd put money on WF's Tempest temp/hum AI being more accurate than the CWOP QC function when it's finished, and light-years ahead of WU's joke gold stars.....
So does this mean that nearby (whatever that may be) PWS's temp/humidity will be gathered and displayed individually to the Tempest user as a generic temp just as is the rain?
-
So, just how are the sensor measurements displayed or transmitted?
What I mean is do you have the measurements in real time or are you getting data that has been processed away from your PWS?
I've looked briefly, but couldn't ever decide which.
Getting yesterday's accurate data today wouldn't make me happy if that's the case.
-
I think a user can check box QC data from the software part of things if a user chooses to do so.
Thanks,
tweatherman
-
I think a user can check box QC data from the software part of things if a user chooses to do so.
:???: Does that mean one can choose between using the "cloud" data vs ones own data?
-
From WeatherFlow,
“60 days from when you receive the package and at least make contact with us before the 60 day mark. We're rational people so if we're troubleshooting or you miss it but a few days no worries, we can still process a return.“ They also stated the they do the prepaid package back to them under the 60 day return policy.
tweatherman
-
having just spent twenty minutes perusing over the weatherflow forum there are multiple all too familiar start up issues . also digesting some of the finer points the use of cloud sourced manipulation for data is apparent .
so what you see displayed may not be your actual hardware readings , this relates to rainfall , lightning,uv index. so beware this latest product is rather unique and a very new different approach be it good or bad but if this approach suits you beware it has a strong dependency on a reliable active internet 24/7 .
common issues as seen in the past are rainfall and solar recharging,uv index...
take the time to peruse over the forum of end user findings .
https://community.weatherflow.com/t/tempest-not-charging/6522
https://community.weatherflow.com/t/lightning-reports-tempest/6262
https://community.weatherflow.com/
so my advice hang fire until they iron out the anomalies and remember there are no end user serviceable parts its generally replace whole unit when things go wrong but at least you have solid sound customer replacement support .
-
i didnt know whether to laugh or think are you for real ? but here is a weatherflow tempest installed on a mailbox !! yep a mailbox and due to the nature of the technology for the rain sensor i was absolutely amused .. :grin:
[ You are not allowed to view attachments ]
see for yourself and watch the delivery guy turn up .. https://youtu.be/r1allGBVj8g
[youtube] https://youtu.be/r1allGBVj8g[/youtube]
-
[-(
That is beyond ridiculous.
Surprised the HOA hasn't dropped the hammer.
-
I'm surprised that someone hasn't stolen it yet!!!!! :lol:
-
This one from the same house (before WeatherFlow install) is equally ridiculous. Must be something in the water over there.
https://www.youtube.com/watch?v=WuAuCE7h0LA
-
I'm surprised that someone hasn't stolen it yet!!!!! :lol:
Probably because it just looks like a used up Brita filter. A little larger but still.
[ You are not allowed to view attachments ]
-
Probably because it just looks like a used up Brita filter. A little larger but still.
I always thought that the design resembled a, ummm, muscle relaxer device formerly marketed by Hitachi, but to each his own..... :grin:
-
I can see that too now that you bring it up. It was a pretty magical device, so I've heard.
-
hmmmm its getting amusing now :grin:
[ You are not allowed to view attachments ]
-
Weather34,
How long have you had your new Tempest for?
Thanks,
tweatherman
-
i dont own one but i see identical issues that were all present in the previous product called sky , that being the big one rain related issues , i never understood or could fathom out why they continued with this approach regardless it was raised consistently to them over the last two years . i also really am alarmed that there forum for a new product called tempest is just full of all the common issues that were raised to them in the previous sky along with some new teething ones . seriously this low cost component haptic approach is terrible it just doesnt work consistently and ill stick my kneck out and say avoid this product until they change the approach they have had two or more years to resolve it and nothing has changed or improved . in my view consistent relative readings within a reasonable tolerance are the basic requirement of a weather station however i accept things can go a bit awol when some diverse weather turns up on your doorstep like a snowstorm,hurricane,tornadoes etc but on non diverse normal days i would expect something relative in return for what you see outside the window in the backyard etc..
only real advances ive seen from the tempest over the sky is the improvement on the temperature related sensor readings but the rain performance is just “ugly” .
-
So I guess a loud thunderstorm could also set it off?
-
Very interesting!
tweatherman
-
They apparently did try and make some improvements with the haptic rain sensor by isolating it more from the rest of the unit, I'm assuming with some elastic material. Not sure how it panned out.
I've got the original unit, and for the price, I'm pretty happy with it. It really does struggle in light rain though, generally to the point of not reporting it. Unfortunately I haven't got a manual or tipping gauge to compare against at the moment.
-
I just picked this up from a FB conversation:
I have to say, I spent $330 and bought the new fancy WeatherFlow Tempest weather station, no moving parts, used scientific AI to configure the data and such, looked really impressive... Has a lot of impressive features too. Well it’s rain “gauge” is flawed. Instead of collecting water it uses vibrations and it’s called a Haptic Rain Sensor, tells the exact time it started and stopped and the depth of rain thru pressure.
So we had a bad rain event here and it only showed .52” rain that night, my CoCoRaHS manual gauge had collected .70” in it, and then the next day the Tempest changed it rain amount to 1.55”!! So I eMailed them and said how can my Wx Station change in 8 hours from .52” to 1.55” when my manual true rain gauge only shows .70”?? I said I run a local Skywarn organization for the NWS in my area I need ACCURATE real-time readings not delayed flawed readings. That night I hooked my Ambient WS-2902 back up! I love my Ambient 2902!
That thing is a joke. It is cute ...I'll give it that. But it isn't a real instrument. If all you are going to get is AI rain data then why even bother having a weather station. You can get overall rain data averages for any area without a station. Yet many people buy this things....I just don't get it.
I don’t either. I mean it does have an impressive list of features it displays, but it’s flawed, an AI can’t give you weather results, and when I wrote them to say I’ll be returning it I found out that the reason my Tempest jumped from .52” to 1.55” in 8 hours is cuz they crowdsource the other Tempests in the area and the AI feeds the data based on what those other stations average. I said that’s total bullshit, you guys are meteorologists you should damn well know the rain at MY location won’t be the same at another location a mile or 2 away... That’s when I decided to dump them and I have to package it up to return it now. Unless I sell it.
-
Pretty damming.
-
they crowdsource the other Tempests in the area and the AI feeds the data based on what those other stations average. I said that’s total bullshit, you guys are meteorologists you should damn well know the rain at MY location won’t be the same at another location a mile or 2 away...
AFAIC this is a case of "pulling the wool" over unknowing eyes. They know that all this AI and such will be a carrot stick for those who like gadgets and most will be clueless on the "limitations" of said design. Frankly I personally don't care, unless they post their crap online...which is probably about 99.9% that do. ](*,)
-
Pretty damming.
Yeah, one review and time to dump it. LOL
-
I’m still sitting back and watching all the reviews. I was really liking what I was hearing from a lot of reviews.
Thanks,
tweatherman
-
Pretty damming.
Yeah, one review and time to dump it. LOL
Yeah, you betcha. It's an extremely flawed way to gather reliable data.
-
if you have about a month to spare go read there forum its riddled with issues from day one .
i can not fathom why weatherflow as a company with a good historic professional background fail to acknowledge the rain measurement via their haptic design is totally flawed . this has highlighted to them since early 2018 in the initial consumer product called SKY yet here they are with a second generation 2 years on with a product called tempest and its riddled with problems.
my own far afield observations from day one
a vertical solar panel has upto 40% less efficiency and to make that worse the capture area is relatively small ,so what you will see is seasonal replenishment problems under all scenarios be it cloud,tree cover,landscape elevation and so on. what makes it even worse is they have a threshold setting no user its fixed where if voltage is low wind sensor , rain sensor will shut down.
remember it is an all in one so there is some trade offs those being any of these wind,temperature,rain,solar replenishment .
rain measurements to sum it up its awful and unreliable to any enthusiast who cares about accuracy and keeping logs of data. time and time again they think rain start or onset of rain is a unique feature well i can think of nothing worse than seeing inaccurate realtime data and this rain check thing is a scam as far im concerned might as well just download an app from an app store .
uv sensor is problematic like all low end low cost uv solutions there is a so called calibration which basically connects to there servers and gets fudged .
lightning sensor in the tempest raised my eyebrows and think its safe to say lightning data you see is not always what is detected there is third party source inflating the figure. but having said that i can see its usefulness in awareness of probability of a storm.
the tempest product relies on being able to connect to there servers however for the advantageous out there disable or connect it to a local router without internet access as the product can simply be used over a local network and sit back for aweek or so and draw your own conclusions on what the tempest unit is actually measuring without any influence,corrections,manipulation from another source.
having had full access to two installations i drew a conclusion over a few months that there is some method a data manipulation coming into play and what is displayed is not always realtime hardware data.
as an hobbyist/enthusiast this would drive me nuts having become accustomed to having un manipulated realtime data that at times doesn't need an internet source to function.i.e power outages,internet outages everything i use aka Davis with a nanosd still fully functions under outages .
weatherflow are a innovative company that just aint got it right and when you see quotes of field testers “ they are super busy sending out replacements “ the word “replacements” rings alarm bells .. i still hold hope for them that they are to produce a reliable product with a separate rain gauge in the form of tipping or weight measurement but haptic method is abysmal.
fwiw im not a weatherflow basher just dismayed that soon after the tempest was released,dispatched to the backers of their campaigns the all too familiar issues that were highlighted way back in 2018 in their first generation product were basically present in the 2nd generation tempest along with a few new ones.
beware some will only say good things to keep their relationship with weatherflow in tact but there is no harm in critical if its valid .some will not hear of anything critical of the product and some will refuse to hear that their $300 innovative product is not living up or performing to expectations .
for me personally i refused to get involved in this second generation testing but i strongly suggested them to send a test unit to justin aka jgentry the trusted respected source to me anyway on here of testing the sht31 with the baking method but they didnt . if they had i would have mentioned a few other trusted knowledgeable sources on here to really help them iron out the finer points..
if your into home automation then this could be what your looking for...
-
Pretty damming.
Yeah, one review and time to dump it. LOL
Go to their forum and read about it. There are a lot of issues. Maybe you didn't see it that way.
-
if you have about a month to spare go read there forum its riddled with issues from day one .
i can not fathom why weatherflow as a company with a good historic professional background fail to acknowledge the rain measurement via their haptic design is totally flawed . this has highlighted to them since early 2018 in the initial consumer product called SKY yet here they are with a second generation 2 years on with a product called tempest and its riddled with problems.
my own far afield observations from day one
a vertical solar panel has upto 40% less efficiency and to make that worse the capture area is relatively small ,so what you will see is seasonal replenishment problems under all scenarios be it cloud,tree cover,landscape elevation and so on. what makes it even worse is they have a threshold setting no user its fixed where if voltage is low wind sensor , rain sensor will shut down.
remember it is an all in one so there is some trade offs those being any of these wind,temperature,rain,solar replenishment .
rain measurements to sum it up its awful and unreliable to any enthusiast who cares about accuracy and keeping logs of data. time and time again they think rain start or onset of rain is a unique feature well i can think of nothing worse than seeing inaccurate realtime data and this rain check thing is a scam as far im concerned might as well just download an app from an app store .
uv sensor is problematic like all low end low cost uv solutions there is a so called calibration which basically connects to there servers and gets fudged .
lightning sensor in the tempest raised my eyebrows and think its safe to say lightning data you see is not always what is detected there is third party source inflating the figure. but having said that i can see its usefulness in awareness of probability of a storm.
the tempest product relies on being able to connect to there servers however for the advantageous out there disable or connect it to a local router without internet access as the product can simply be used over a local network and sit back for aweek or so and draw your own conclusions on what the tempest unit is actually measuring without any influence,corrections,manipulation from another source.
having had full access to two installations i drew a conclusion over a few months that there is some method a data manipulation coming into play and what is displayed is not always realtime hardware data.
as an hobbyist/enthusiast this would drive me nuts having become accustomed to having un manipulated realtime data that at times doesn't need an internet source to function.i.e power outages,internet outages everything i use aka Davis with a nanosd still fully functions under outages .
weatherflow are a innovative company that just aint got it right and when you see quotes of field testers “ they are super busy sending out replacements “ the word “replacements” rings alarm bells .. i still hold hope for them that they are to produce a reliable product with a separate rain gauge in the form of tipping or weight measurement but haptic method is abysmal.
fwiw im not a weatherflow basher just dismayed that soon after the tempest was released,dispatched to the backers of their campaigns the all too familiar issues that were highlighted way back in 2018 in their first generation product were basically present in the 2nd generation tempest along with a few new ones.
beware some will only say good things to keep their relationship with weatherflow in tact but there is no harm in critical if its valid .some will not hear of anything critical of the product and some will refuse to hear that their $300 innovative product is not living up or performing to expectations .
for me personally i refused to get involved in this second generation testing but i strongly suggested them to send a test unit to justin aka jgentry the trusted respected source to me anyway on here of testing the sht31 with the baking method but they didnt . if they had i would have mentioned a few other trusted knowledgeable sources on here to really help them iron out the finer points..
if your into home automation then this could be what your looking for...
I agree with all your points. I've had the 1st generation since it was released. (Too) many issues to convince me that the station is a replacement for a Davis. Not even for the cheapest station I own which costs 1/3 of the WF price. Despite reporting obvious issues with rain and especially with wind readings continuously, no clear explanation was given by their staff. Replacements are not really a solution because eventually same problems pop-up. I received few Sky replacements and all show wind measurement issues. The Air is fine, no complaints.
I'm sure the 2nd generation is using same sensors but with more issues on power supply. Why would I compromise the readings if there is not enough sunlight to charge the batteries? I'm happy I didn't get involved into 2nd generation.
Data manipulation is another reason why I don't want this station anymore. I want to log the temperature, humidity, pressure, rain, lightning strikes and wind on my location. I'm not interested in "other sources". I can check the web for other sources myself.
-
New Year fireworks registered light rain at midnight here! :lol:
Sky was clear, nearly full moon in tow.
The sensor is like that, impulse noise is going to be recorded as raindrops.
-
It been 2 1/2 years since the last posting. Have the Tempest stations been fine tuned for better accuracy with rain by now? If it hails on them, is it recorded it as rain?
-
I have a couple of them now, one is mounted near the airport ASOS and the other at my place. The haptic rain sensor itself isn't accurate but WF has software that adjusts to surrounding stations so in the end doesn't look that bad. If you were able to watch prior to the adjustment say during the rain event you can see they overcount badly as much as triple sometimes. The rest of the features are pretty good like air temperature and if you don't mind the software adjustment on rainfall for the average Joe, not bad units.
-
Reliability is my biggest issue.
My first unit was purchased in December of 2020. In June of 2021 I noticed its firmware was quite old compared to others. I emailed support and they said they needed to replace it as they identified an issue. It was producing readings that were within its operating limits. A few days later and some climbing I swapped it out. This was peak of brood X and I even had a cicada land on my arm and start singing which freaked me out. But a quick cell popped up and I heard thunder so I got wrapped up and was done. Or so I thought!
Next AM I noticed the UV was 0 and we had bright sun. Status showed UV failed. I opened a ticket and they remote rebooted and then most of the sensors were no longer working. They said the lightning from the storm yesterday may have fried it but I know from in the past having a tree not 30 feet from the tower getting bark peeled to the ground from a strike had no effect on my prior unit.
So a few days later, up and down the tower and Tempest number 3 is flying high. That one worked well up until the end of last September. The battery voltage was dropping fast after a few days of no sun. It would recover to 2.8 but as soon as the sun set it was back to 2.5V. Next few days of overcast it was down to 2.4 and then the report rate gets dropped to save power and it finally went offline. They replaced that with number 4 which (again) was put up.
On April 22, during a quick gusty shower, the barometric pressure showed a sudden drop of 5 inches of mercury! We were nowhere near any tornadoes for sure! Of course it's borked again, in the ticket they indicated they would sending out a replacement. This will be number five.
It's getting old fast and I do all climbing/rigging myself but make no mistake about it, I don't enjoy it!
This unit constantly reports rain events during most wind events over 15 mph. Of course towers and hardware are quite "chatty" when it comes to acoustics, which the rain sensor is based on. (It's a $.80 piezoelectric transducer found in most annoying beeper type alarms)
TEMPEST number SIX was installed on a roof tripod mast early May. One June 24 it went AWOL. I was out of town. When I returned I power cycled the outdoor unit and it did connect but its battery was very weak and was in slow mode. It did recover after a few days of full sun. I opened a ticket with WF support and they said it looks OK now that it has recharged.
On July 7th we had a gully washer thunderstorm. I noticed the barometric pressure was off again as before. Nearly six inches lower! I opened a ticket with WF. They told me they have a new and improved unit that will be arriving by the end of the month if I wanted to wait or they could send a replacement right away. I opted to wait. They claim the "new and improved" unit features fewer connection points inside and an insect screen to keep pests out of what I'm assuming is the temperature shield area. It's supposed to be here by the end of the month so we'll see. I'm done with putting it up on the tower!
I suppose I have too high of an expectation for an all in one solution coming in around $300 USD.
The only good thing I can say is their support is outstanding. The unit is past its 2 year warranty period and they pay for shipping both ways and haven't asked for a dime. They also help out in a timely manner with things like calibration issues, et al.
If this is something you put on a post by your garden, then no problem.
On a roof, better be prepared (and comfortable) using a ladder frequently!
You can see in my sig links to both TEMPEST and Ambient Weather 2000 systems. The AW is about $25 cheaper than the TEMPEST and produces decent results and has had zero issues since I put it up last November. It's on the same tower but much lower.
My biggest concern is with (presumably) 100s of thousands of these (TEMPEST) units reporting from all over the globe, a HUGE amount of data, what good is it if it's flawed! It's not just the rain that has accuracy issues. Wind at times is just plain wrong. I've had the anemometer showing a constant 20 mph breeze in fog whereas the other real instruments are showing 0.0. Flying insects like sweat bees and even lady bugs traveling in the sensor gap will wreak havoc with the wind often showing gusts of 30-40 mph on days where the real wind never exceeds 10 mph. The anemometer also seems to be extremely sensitive to flow that's slightly off axis. Other ultrasonics, particularly professional units (eg. Vaisala) have never had such issues. I know I'm comparing a Rolls Royce to a Pinto but you get the idea. ;-)
-
Reliability is my biggest issue.
My biggest concern is with (presumably) 100s of thousands of these (TEMPEST) units reporting from all over the globe, a HUGE amount of data, what good is it if it's flawed! It's not just the rain that has accuracy issues. Wind at times is just plain wrong. I've had the anemometer showing a constant 20 mph breeze in fog whereas the other real instruments are showing 0.0. Flying insects like sweat bees and even lady bugs traveling in the sensor gap will wreak havoc with the wind often showing gusts of 30-40 mph on days where the real wind never exceeds 10 mph. The anemometer also seems to be extremely sensitive to flow that's slightly off axis. Other ultrasonics, particularly professional units (eg. Vaisala) have never had such issues. I know I'm comparing a Rolls Royce to a Pinto but you get the idea. ;-)
I don't blame you for your concern. My small town has 6 Tempest stations online with 5 more rural ones.
And then the OKC metro appears to have well over a 100 of them.