Author Topic: WeatherFlow/Tempest vs Davis Vantage Pro 2  (Read 916 times)

0 Members and 1 Guest are viewing this topic.

Offline parkernathan

  • Forecaster
  • *****
  • Posts: 507
    • WeatherTogether
WeatherFlow/Tempest vs Davis Vantage Pro 2
« on: December 07, 2021, 02:36:07 PM »
I currently have a Davis Vantage Pro 2. I'm debating whether to eventually purchase the necessary hardware to publish its data to the Internet (such as a WeatherLink Live, which I can pick up around $175 from Scientific Sales), or if I should invest in a WeatherFlow/Tempest station instead. I've heard some good reviews on the WeatherFlow/Tempest, and I believe Baron is using them to build out their Critical Weather Institute network. The fact that they include an on-site Lightning Sensor also interests me.

If someone has compared the Davis Vantage Pro 2 with the WeatherFlow/Tempest, which one do you prefer?

Thanks!
Nathan Parker

Visit my blog at http://weathertogether.net

Offline Adrian23

  • Senior Contributor
  • ****
  • Posts: 294
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #1 on: December 07, 2021, 04:55:59 PM »
Well i just purchased my Tempest and i am absolutely loving it especially those 3 sec wind updates. I have a ambient meteobridge which allows me to share my data with CWOP, Mesowest etc.. Its all about location as always with weather stations as i am sure you know.

So far so good!

Link to my station.

https://tempestwx.com/station/63157

Online CW2274

  • Forecaster
  • *****
  • Posts: 6200
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #2 on: December 07, 2021, 05:05:01 PM »
 From the Tempest website....

"While weather apps abound, the Tempest app leverages real-time data, advanced forecast modeling and the oversight of in-house meteorologists. Data from each Tempest is processed with advanced machine learning which checks the information for accuracy, allowing us to apply daily calibrations and ultimately improve your forecast over time."

Complete deal breaker for me. My data is just that, mine, not others blended in as well.  :roll:

Offline parkernathan

  • Forecaster
  • *****
  • Posts: 507
    • WeatherTogether
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #3 on: December 07, 2021, 10:03:19 PM »
Thanks everyone for the info! I'll check out the Tempest page. My Davis Vantage Pro 2 is currently connected to an Earth Networks Hub, so I'm stuck at sharing data with others even on it (so that wouldn't be a huge deal breaker for the Tempest). I do have a spare Vue console that can accommodate a data logger or could pick up a WeatherLink Live so I have a little more control over some of the data I share.
Nathan Parker

Visit my blog at http://weathertogether.net

Online CW2274

  • Forecaster
  • *****
  • Posts: 6200
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #4 on: December 07, 2021, 10:30:59 PM »
Thanks everyone for the info! I'll check out the Tempest page. My Davis Vantage Pro 2 is currently connected to an Earth Networks Hub, so I'm stuck at sharing data with others even on it (so that wouldn't be a huge deal breaker for the Tempest).
You're misunderstanding "sharing". We all share our data, that's basically why we post. If I understand correctly, what the Tempest does is it extrapolates from other sources to manipulate your raw data. So if the app or whatever feels your data is incorrect, it "fixes it". No bueno, especially for the dynamics of wind and rain.

Offline parkernathan

  • Forecaster
  • *****
  • Posts: 507
    • WeatherTogether
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #5 on: December 08, 2021, 01:09:51 AM »
Thanks for the followup. I believe (someone correct me if I'm wrong) this doesn't apply to viewing your own current/historical data, but how it feeds the data into their hourly forecast engine. It feeds your station data in with other's station data, plus other sources (such as weather models) to help them fine-tune their hourly forecast engine. That's how ENcast works (which powers WeatherBug and some other apps that use it). My station (since it's connected to the EN weather network) feeds data in with other stations from around town to fine-tune the ENcast hourly forecast engine. My own current/historical data never changes, but the more stations that are in a given area, the more accurate my hourly forecast from ENcast is supposed to be.

Which with at least four stations in the county, it's still pretty wildly inaccurate at times, and I've learned a human is better than a computer-generated forecast.

I could be totally off base here (WeatherFlow/Tempest may do something totally different), but that's how it works for ENcast, and it sounds as if WeatherFlow/Tempest is doing something similar.
Nathan Parker

Visit my blog at http://weathertogether.net

Offline vinceskahan

  • Senior Member
  • **
  • Posts: 85
    • home site
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #6 on: December 13, 2021, 07:28:04 PM »
I currently have a Davis Vantage Pro 2. I'm debating whether to eventually purchase the necessary hardware to publish its data to the Internet (such as a WeatherLink Live, which I can pick up around $175 from Scientific Sales), or if I should invest in a WeatherFlow/Tempest station instead.

Strongly suggest buying a Davis USB datalogger and sticking with the VP2 unless you really need solar/lightning sensors too.

The WF gear 'fixes' your rain readings the next day based on some magic math from surrounding sites, but all it did was make the inaccurate readings worse here for me.  It totally manufactures synthetic lightning readings based on surrounding sources of data they don't reveal to us.  It also tries to correct for case heating/cooling based on detected airflow and more magic math, but my experience is that it's slow to warm up and slow to cool down regardless.  And their 'continuous learning' stuff will quietly attempt to retune your sensors.  Rain is very questionable and can read massively high or low.  Most importantly, the Tempest has big issues in many locations trying to even keep the unit charged via its solar.  You need constant Internet connectivity as well for it to work.

That said, they're nice people to deal with and obviously working hard but it is a bit of a slow grind it seems.  Personally I'd wait til they come out with a 3rd-generation station (plus a year) someday hopefully.

In contrast, the Davis gear is more old-school 'just install it and it works' without any need for anybody Internet connectivity etc, and if you go with the Serial/USB loggers you can run a variety of software such as (my choice) weewx and upload to just about anywhere you can think of.   Davis gear also has a long proven track record for hardware quality and area equally good people to deal with in my experience.

Given you're already invested in Davis, $175-200 for a logger and raspberry pi is far cheaper than $350 for a Weatherflow unit as well.
WeeWX sites:
  Davis VP2+DFARS to a Seagate Dockstar (pogoplug)
  EcoWitt GW1000, WH32 outdoor T+H, multiple WH31 indoor T+H, WH51 soilMoisture (docker)
Home site:        https://www.skahan.net/
Wunderground: KWAFEDER15
PWS:                KWFEDER15
CWOP:             CW6881

Online CW2274

  • Forecaster
  • *****
  • Posts: 6200
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #7 on: December 13, 2021, 07:44:26 PM »
I currently have a Davis Vantage Pro 2. I'm debating whether to eventually purchase the necessary hardware to publish its data to the Internet (such as a WeatherLink Live, which I can pick up around $175 from Scientific Sales), or if I should invest in a WeatherFlow/Tempest station instead.
The WF gear 'fixes' your rain readings the next day based on some magic math from surrounding sites, but all it did was make the inaccurate readings worse here for me.  It totally manufactures synthetic lightning readings based on surrounding sources of data they don't reveal to us.  It also tries to correct for case heating/cooling based on detected airflow and more magic math, but my experience is that it's slow to warm up and slow to cool down regardless.  And their 'continuous learning' stuff will quietly attempt to retune your sensors.  Rain is very questionable and can read massively high or low.
AFAIC, they can manipulate whatever data they want...til it ends up online, which is probably 90+% of the time. This is soooo unacceptable. If they do it....wonder who else does. Can't imagine.  :evil:

Offline vinceskahan

  • Senior Member
  • **
  • Posts: 85
    • home site
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #8 on: December 13, 2021, 07:53:14 PM »
I come from an old-school oil refinery instrumentation background where sensors work right out of the box with no 'continuous learning' required, so I guess I'm enough of a fossil that I'm not their target audience.  Darn kids !!!!

That said, I find nothing even slightly objectionable with how WF does business, FWIW, and I certainly don't even think about what a downstream site I feed does with my data.  If I did, I wouldn't feed them anything.

(someday - looking at you Weather Underground....yup you....)
WeeWX sites:
  Davis VP2+DFARS to a Seagate Dockstar (pogoplug)
  EcoWitt GW1000, WH32 outdoor T+H, multiple WH31 indoor T+H, WH51 soilMoisture (docker)
Home site:        https://www.skahan.net/
Wunderground: KWAFEDER15
PWS:                KWFEDER15
CWOP:             CW6881

Online CW2274

  • Forecaster
  • *****
  • Posts: 6200
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #9 on: December 13, 2021, 07:59:59 PM »
That said, I find nothing even slightly objectionable with how WF does business, FWIW, and I certainly don't even think about what a downstream site I feed does with my data.  If I did, I wouldn't feed them anything.
You obviously have no issues with this. I absolutely do. Manipulation is just that, a falsehood. Completely in line with today's "climate" of thinking.

Offline vinceskahan

  • Senior Member
  • **
  • Posts: 85
    • home site
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #10 on: December 13, 2021, 08:11:25 PM »
So shop elsewhere and vote with your wallet.  I did.  I know of others who did so as well.
 
Regardless - I'm not interested in political discussions here please.  Really.

FWIW -  I didn't agree with their 'technical' approach so I sold/gifted all my gear to others who were more ok with how the gear works under the hood.   I had no problems with them as a company or individually in the least.  I simply wanted an old-school no-internet-required just-gimme-the-blasted-sensor-readings station.



WeeWX sites:
  Davis VP2+DFARS to a Seagate Dockstar (pogoplug)
  EcoWitt GW1000, WH32 outdoor T+H, multiple WH31 indoor T+H, WH51 soilMoisture (docker)
Home site:        https://www.skahan.net/
Wunderground: KWAFEDER15
PWS:                KWFEDER15
CWOP:             CW6881

Online CW2274

  • Forecaster
  • *****
  • Posts: 6200
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #11 on: December 13, 2021, 08:27:13 PM »
So shop elsewhere and vote with your wallet.
 
I'm not in the market. I've owned my VP2 and it's data just shy of 15 years.

Political or not, I'm doing nothing more than stating facts.

Offline Bashy

  • brecklandweather.com/pws2021
  • Forecaster
  • *****
  • Posts: 1369
  • brecklandweather.com/index.php
    • Breckland Weather
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #12 on: December 13, 2021, 11:11:24 PM »
I was thinking of the WF as a second station alongside my VP2 and just for the lightning, but after reading all i could find, there is no way i would use 1, not even as a second station, As for the rainfall, there's 3 other davis stations around me, they're farms, all within a few miles of me, we pretty much agree on temps and humidity etc but the rain can differ quite a bit at times, there is no way i would want my rain data manipulated or any other for that matter, each site is different, from the lay of the sensors to the surrounding buildings, trees etc, so how do you even know if your data is gonna be correct after its been fiddled with, whats to say the other stations are correct after fiddling, makes no sense to me, none of it, if WF are basing it on the local METAR's etc then that would only work if there is one in your back yard. I don't know, your data is unique to your yard and should be an average between x amount of stations, heck, what radius do they even use, 1, 5, 20 miles?
Kind regards
Bashy

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 6223
    • LiVE WEATHER
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #13 on: December 14, 2021, 10:04:17 AM »
Nothing wrong with having one, I wouldn't replace the Davis FARS with one. The biggest issue I see is being an all-in-one station wind, precip and temperature should be measured at 3 different elevations but for a backyard station for the average Joe, it offers a lot for the price including UV, solar, lightning detection and uses the latest SHT-31 with the dust film protector so never needs a filter clean.  One of the things the AI does is act like a FARS without the moving parts using a known internal temperature, solar energy, and wind speed. In my limited testing, it seems to do great within a couple of tenths most of the time but not as good as a real deal (FARS) running warmer under no or low wind conditions. Very similar to other passive radiation shields I've tested. This could be caused by the time of year where our solar energy is only reaching 400 W/m2 on a good day so no AI adjustment is being done. That's a mistake however even under low solar conditions a passive shield will heat up when calm or just a little breeze.  It fails the snow reflection test miserably running 9-10 warm when the sun angle reaches certain points. This is what I've experienced in limited use over the past month.
Randy

Offline vinceskahan

  • Senior Member
  • **
  • Posts: 85
    • home site
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #14 on: December 14, 2021, 02:01:03 PM »
One of the things the AI does is act like a FARS without the moving parts using a known internal temperature, solar energy, and wind speed.

If you can trust all three things are accurate to begin with, sure.

The problem there is you are using unknown (questionable) quality sensors like solar+wind to tune the output for a natively much more trustable sensor (temperature).   So you're using less-trusted to alter the reading of the more-trusted.

I ran long comparisons of the old Air+Sky and Tempest gear versus my VP2, a home-built ds18b20 sensor rig, and some AcuRite and Ecowitt sensors and found that in the dark the temperatures and humidity readings lined up very well so that tells me the native sensor they're all using is likely the same (or close enough).

But in sunlight the Tempest was slow to warm up (morning) and slow to cool down (after sunset).  That predated their tweaking based on wind+rain but my hazy recollection there is their max alteration is 2 degF or so.   That's one of the reasons I liked the old Air+Sky setup better - keep the Air in shade and you don't see the case heating/cooling effect.

I couldn't compare wind readings due to siting differences. 

The snow reflection thing you're seeing is the first I've heard of it.  Wonder if you could MacGuyver some kludge setup to only expose the solar panels to direct sun.   I don't recall if I've ever seen where the temperature sensor is located in the Tempest but I bet somebody in the WF forums on their site might have.

WeeWX sites:
  Davis VP2+DFARS to a Seagate Dockstar (pogoplug)
  EcoWitt GW1000, WH32 outdoor T+H, multiple WH31 indoor T+H, WH51 soilMoisture (docker)
Home site:        https://www.skahan.net/
Wunderground: KWAFEDER15
PWS:                KWFEDER15
CWOP:             CW6881

Offline vinceskahan

  • Senior Member
  • **
  • Posts: 85
    • home site
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #15 on: December 14, 2021, 02:11:51 PM »
there is no way i would want my rain data manipulated or any other for that matter, each site is different, from the lay of the sensors to the surrounding buildings, trees etc, so how do you even know if your data is gonna be correct after its been fiddled with.

Yes - many of us asked for a "turn all that fancy stuff off" switch but it was never provided, but I didn't worry it too much since I ran weewx listening to the local UDP broadcasts so I got it before their server messed with it.  What I had in weewx was the unaltered (bad) rain data so their next day RainCheck didn't matter to me one way or the other.

After a couple years I had pretty much given up on the rain being accurate, which mattered a lot to me (I compare to my CoCoRAHS manual gauge) and when I heard about the temperature tweaking based on uncertain-accuracy wind+solar, I kinda decided they weren't a match for what I was looking and were continuing to move in the wrong direction architecturally for what I was looking for.

That said, it was great fun writing a threaded UDP listener utility and Hub broadcast simulator and fiddling with their nice APIs.  I certainly got my money's worth there by far, but at this point if the VP2 ever breaks to the point where I can't replace pieces affordably I'd probably just buy a Vue or VP2 sensor suite in as a replacement.
WeeWX sites:
  Davis VP2+DFARS to a Seagate Dockstar (pogoplug)
  EcoWitt GW1000, WH32 outdoor T+H, multiple WH31 indoor T+H, WH51 soilMoisture (docker)
Home site:        https://www.skahan.net/
Wunderground: KWAFEDER15
PWS:                KWFEDER15
CWOP:             CW6881

Offline Tln7559

  • Slangenbeek
  • Member
  • *
  • Posts: 16
    • Weather&PV Hengelo_Slangenbeek
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #16 on: December 16, 2021, 03:51:49 AM »
After a year of operation with Tempest have similar impression as Vince.

Intending that Tempest would become newer replacement, run the Tempest for evaluation parallel to a 20 year old LaCrosse_WS7000 and a slightly younger TFA_Nexus.
It took Weatherflow 6 months to find the autocalibration-correction for pressure, and rain is still not aligned.
Other flaw of Tempest is the weak solar battery system for darker regions of the earth:
in the Benelux and more north in Europe it means frequent & long drop-out in winter.
The promised power booster is a long overdue extension.
Conclusion/Result:
- The 'old guys' do the primary job, delivering data 24/7 and will continue as long as they live.
For WS7000 modern replacementsensors implemented (with SHT31) for temperature and humidity, raising quality for the WS7000 and extending life.
- Tempest is/remains add-on for the sensors missing in the old PWSes, and backup for the 'old guys' during sufficiently sunny conditions.
Local read-out of Tempest at UDP-level with WeeWX, enabling direct adjustments
=> autonomy ;-) as usual with the older PWSes.
- Domoticz acting as auxiliary software, support & gapfiller for all sensors and uploads.
« Last Edit: December 18, 2021, 04:43:58 AM by Tln7559 »
Sensors: TFA_Nexus + LaCrosse_WS7000 + Tempest + DIY
Software: WsWin + WeeWX + Domoticz

Offline wxperson

  • Senior Member
  • **
  • Posts: 50
Re: WeatherFlow/Tempest vs Davis Vantage Pro 2
« Reply #17 on: January 05, 2022, 03:12:07 PM »
I have both stations and have similar observations as others.

Davis VP2 has lasted for 10 years with no real issues.   I did replace the tipping bucket with the new tipping spoon and that helped rainfall measurements quite a bit.

The Tempest has several  issues.

1. Battery issues.   I live in Atlanta, GA and in the fall the unit will not charge because of a line of deciduous trees to my south.  It will fall to levels where the system will no longer function properly.

2. Wind values can go nuts and erroneously reports high winds (40+ mph) if water droplets are on the unit.  They have fixed some of this via software but even so... the measurement of gust is inferior to the Davis.

3.  The biggy.   The haptic rainfall device does an awful job of estimating rainfall sometimes off by 100%.   They tried to fix this by allowing you to use "neighborhood" real reported values (COCORAHS) but that has limitations.

I do like the lightning detector on the Tempest which the Davis does not have.

George

 

anything