Difference in rainfall data between WeatherFlor PiConsole and official Tempest app

Hi! I was lucky enough to get a nice rainy day here on the East Coast a couple of days after getting my PiConsole running. But I noticed today that the console and the official Tempest app are giving two very different measures of the daily rainfall. Just curious if this has been a reported issue, or (more likely) if it’s user error on my part somehow. Screenshots attached!

Screenshot RPi4 2023-04-28 at 14.29.08

Surprised nobody responded yet. This is due to Nearcast rain totals vs. what is reprted from the haptic sensor. You can tell this from the icon next to the 0.59" total in the first screen. If you turn off Nearcast rain, it should match the PiConsole. You can toggle Nearcast rain off and on … both totals exist on the server … only one at a time is displayed in the app. See https://help.weatherflow.com/hc/en-us/articles/360024436634-NC-Rain-Nearcast-Rain-#:~:text=Go%20to%20settings%20>%20stations%20>%20choose,on%20Tempest%20System%20Rain%20Data. .


Thank you, and sorry for the noob question! I’m curious to dig in more and cross reference the two readings against nearby NOAA measurements. But I’m also wondering, in the community’s experience: which of the two tends to be more accurate?

It tends to vary back and forth. It seems that for my stations where I have submitted a significant number of co-located rain readings from my CoCoRaHS gauge, the non-NC rain is often better. For the station where I have not submitted data, it switches back and forth.

It’s not really a matter of which one is more accurate, it’s a matter of which type of data you want to see. With Near Cast off, it’ll be accurate to your specific mounting location. With Near Cast on, it’ll be accurate to an area around your house.

Think of putting 10 Tempest devices in your yard. When it rains, each will record a measurement for that specific spot, but each will be a bit different, despite them all being accurate. With Near Cast on, you essentially average them all.

Both Near Cast on and off provide accurate measurements, they just provide them for different areas.

Are you sure about that? Have you tried a setup like this?

I can guarantee you that the Tempest in this location does not agree with the CoCoRaHS gauge all the time. If the calibration is off on the Tempest, then it doesn’t give an accurate measurement. Yes, this has been tested. This is the location of a field test station.

How much do you typically find them to be off by?

That varies depending on conditions. It isn’t easy comparing to a CoCoRaHS read at 0700 when the Tempest resets at midnight. I’ve attached a spreadsheet of the data I collected between January 3, 2022 and June 19, 2022. I removed the data from days with snow as the Tempest/SKY does not measure snow fall. The devices labeled with BY in the name are the ones in the Back Yard on the same post as the CoCoRaHS gauge. Remember that comparing days doesn’t work due to the 7 hour delay in the CoCoRaHS reading compared to the Tempest and SKY units. Tempest-Field was installed on 1/3/2022 and had no calibrations applied due to the submitting of my data to WeatherFlow. The other stations were calibrated and possibly had calibration adjustments due to this data being submitted for them prior to and during this data recording time.

Here is a screen capture of the first few rows, the whole file is attached below:

Rain measurements all stations 2022 - subset.zip (20.8 KB)