Air pressure is inaccurate (solved with CL system)

Height above ground level is set to 1 m

That’s really low for the Sky to be mounted.

1 Like

Is your sky mounted that hight ?

What are the value of accuracy and repeatability of the different sensors ?

It’s the Air which is 1 m above ground level. As far as I know is Air measuring air pressure and temperature. Standard should be 2 m above grassy ground. Sky is on the roof.

1 Like

Duh. That is correct. I have no idea why I thought Sky

1 Like

Sensor values seem to be OK; no jumps etc. The difference between WeatherFlow and Davis is more or less constant.

260 m for stationheight above sea-level and air height of 1 m seems perfectly acceptable values for your location. And doesn’t explain the difference you see with other nearby stations (but note that you can only compare the sea level values ( asuming they are setup correctly). 4 hPa is about 40 m difference. So even if your station height is a few meter off, it wouldn’t explain the difference

1 Like

Make sure you understand that the Davis Vantage Pro’s barometer sensor is located inside the console unit and not inside the hygrometer on the station. Due to elevation difference of your AIR and Vantage Pro console, the pressure observations may differ. Your Vantage Pro’s barometer sensor should be calibrated to your nearest METAR station (local airport station) as recommended by NWS. It is also a requirement for being a MADIS station as part of CWOP.

I know. The difference in height between the Vantage console and the Air ist about 2 m.

It looks like your pressure sensor is a bit out of calibration. We are close to launching the pressure component of our continuous learning system, which will automatically apply a correction to units that need it. Please be sure your elevation and height above ground are set correctly and give us a bit of time to roll this component out. Thanks!

4 Likes

Hello, I have the same problem. Mine is consistently close to 3mb too high. I was told that the unit would self adjust using some AI calibration but this hasn’t happened. I have a government survey marker nearby so the programmed elevation is reasonably accurate, well within 1mb accuracy. Initially I adjusted the program elevation to adjust the atmospheric pressure but I was told to set it to the correct elevation and await calibration.

If you read the messages from WeatherFliw and David, you will learn that it is being worked and has not been turned for all stations.

Thanks Gary, that would have been a good response a couple of months ago rather than telling me to wait and it would shortly sort itself out. I keep a close eye on the QNH as I’m a pilot and fly from my property. The weather station is to provide me with conditions at my runway so I’m looking forward to being able to rely on it and not have to add 2.7mb. I don’t usually frequent the forum. Only check it when the summary email comes through.

WeatherFlow is constantly checking stations and gathering data. This is a huge effort to analyze and build the capabilities to calibrate each station.

Without very precise and expensive equipment at every site it’s still going to be a best effort. I have been discussing accuracy with techs at KAPA and it’s a constant battle to keep the equipment calibrated and accurate. And that’s at the high end. This consumer product is a bigger.

Thanks Gary, well it’s obviously consistently wrong, maybe a way for the owner to make an adjustment without having to change the elevation would be a simple solution. I study all the aviation weather stations within 70km and the sea level pressure is within 1mb so it would be very easy over a short period to reach acceptable calibration.

Like kids at Christmas we are anxious for accurate data. I’m waiting for humidity to be adjusted so I can better control the temperature inside my house. Air pressure is important to me but with so little air here I don’t notice when it’s a little off. But I’m learning. Now to figure out what to do with several million data records.

1 Like

Hi @mountwilson . Thanks for your notes and thanks for participating in the forum. As you may surmise from the title of this thread – the Continuous Learning system has not yet been deployed system-wide for pressure. Once deployed, the CL system will inspect multiple data sources in your immediate area every day and apply calibration corrections to your specific sensor if warranted. This allows us to automatically QC your data every day and keep your station in accurate calibration on an on-going basis.

Until we deploy the CL system for pressure, you are welcome to self-calibrate / fudge pressure values by changing your elevation (which amends the values in the formula used to calculate sea level pressure). Once we deploy the CL system for pressure, it won’t really matter if your elevation is fudged…the system can accommodate. However, in the interest of good data, we would encourage you to change the elevation back to your actual station elevation value. Then the CL system will automatically evaluate and calibrate the pressure within 24hrs.

Hope that helps understanding, and thanks for your patience. The CL system is new cutting-edge technology which takes time to develop and test. The goal is accurate sensor calibration based on solid science that happens automatically and consistently over time with no manual user intervention required – as it should be!

5 Likes

Is anywhere a definition of the station level and how the hight over ground for the sensors is relative to that value ?

This, for shure, only applies to those hubs with internet connection to the WFservers ?
What about those without ? May they be “updated” / sequentially calibrated as they connect to an app ?
What process is applied for those locations where totally no internet is available? Is there an option to download and store the data via Bluetooth to the app for a certain period and then upload them later with internet connection of the mobile device to the WFServers and afterwards download the calibration and apply those to the station via Bluetooth ?
And additional, but related, I was thinking about for quite a little time : is a process in development for manual firmware updates (theme like above) ?

There are two pieces of “metadata” used to determine total height above sea level: the elevation (the height of the ground relative to sea level) and the “height above ground” (the vertical distance from the ground to where your AIR is installed). More info about how pressure is displayed is here: https://help.weatherflow.com/hc/en-us/articles/115005398308-FAQs

Correct. If your station is not connected to the internet you will not benefit from the continuous learning system. We will likely add a self-calibration feature at some point in the future.

3 Likes

Pressure calibration for all stations should now be solved. See: CL System for Pressure calibration explained