Yesterday was the first full work day for my new Tempest. We had a storm that started with blowing dust/sand (we live in the AZ desert) and the unit logged 0.13" of rain before the actual rain arrived later. Does the AI have the ability to compare humidity and rain sensor data and determine that the reading is not rain? Rain check is ON.
There really is no relative humidity number at the surface or near the surface that will indicate that the stuff hitting the rain sensor is rain vs. something else. Since most precip forms aloft and falls when it becomes heavy enough, it often falls through unsaturated lower levels (including the surface) and does not neessarily completely evaporate before hitting the surface. The RH of those levels could be less than (and potentially much less than) 100% . So, humidity would not be a good thing to use.
I donāt know of any reliable algorithmusing humidity that could be used to determine that the stuff hitting the Tempestās haptic sensor (or, causing the sensor to vibrate in the case of false wind-generated readings if the mount is not solid enough) is indeed rain.
1 Like
Thank you for the explanation. Obviously not reliable to depend on humidity for determining what hits the haptic sensor. Could you put an optical rain sensor on top of the Tempest, similar to what they use for rain sensing windshield wipers?
The original rain amount (1.10") was adjusted (to 0.20") by the time I checked this morning, thanks to Rain Check. Iām looking forward to watching and learning. Thanks again!
Actually I think it is doing exactly thatā¦ suppressing rain alerts when humidity is low. But I think that is tuned for things like a bird landing on top. Iām not sure but it might be that when it gets a lot of āfalse rainā detections, it finally decides that apparently it is raining, even though humidity is low. If that is the case your sand storm will be detected as rain. Or perhaps your humidity wasnāt low enough.
Weatherflow calls this āenhanced rainā, it means that when the forecasted probability for rain is less than 1% or when the humidity is less then 50%, no rain should be reported.
Make sure that the wind is causing vibrations by having a good mount.
I would like to know the meteorology behind choosing 50% as a floor. It seems arbitrary to me. Precip falling through a dry layer may or may not evaporate, and I am not aware of any studies that show that precip cannot occur below 50% RH (but, I have not done an exhausitve search for one either 
). There are a number of reports (although, I canāt verify these are reliable) of measurable precip occurring with very low RH values, e.g., https://www.city-data.com/forum/weather/2005381-rain-low-humidity.html and 115Ā°F World Record Hot Rain Hits Needles CA & Storm Starts Fire .
I donāt know, but my guess it is that the 50% isnāt based on meteorology but on statistics. They have the data to find out how often it rains with such low humidity values (rarely I think, as in such conditions most rain would have evaporated before reaching the ground). That probably out weights the advantage of having less false rain reports.