Cumulus Support

Hi Steve!
Would it be possible to implement another rain rate calculation algorithm for Fine Offset weather stations, as the current one is not accurate enough?
I read in another topic that you tried implementing one, but you gave up since there was a problem with long log intervals...
Is there a way to change it myself? And if not, would you mind providing me with an experimental version which has this Davis-like rain rate calculation algorithm?
Of course it would be better if this change could be included in a new beta version...

I do intend to revisit the algorithm at some point.

That would be awesome, Steve!
Keep up the good work!

steve wrote:I do intend to revisit the algorithm at some point.

Hi,

Just some food for thought:

As is often being pointed out on another forum (concerned with microcontroller software), to measure a "rate" (e.g. mph in a cycle computer) it is often best to measure the interval between successive "events" (wheel revolutions, rain tips, etc.), rather than to try to calculate from a very limited number of "counts" in a given period. Of course with the FO (and probably most rain gauges), it takes quite a long time to get even two tips (i.e. up to 0.6mm of rain) before you could make the calculation. But does a single tip have any significance? It could have been just one raindrop which "tipped the scales".

Incidentally, I've experimentally mounted a 14cm funnel above one of my FO rain gauges (nominally 2.5x the collecting area, so that I can set the Console to "inches", but read the displayed LS digit as "0.1mm"). This does indeed give a much "cleaner" (non-staircase) rainfall curve in Cumulus, but disappointingly doesn't actually give any earlier indication of the onset of rain (because the water collects as droplets on the surface the funnel). Someday I'll experiment with "Rain-X", but to give an "early warning" is where the Hydreon RG11 really should score well.

Cheers, Alan.

AllyCat wrote:As is often being pointed out on another forum (concerned with microcontroller software), to measure a "rate" (e.g. mph in a cycle computer) it is often best to measure the interval between successive "events" (wheel revolutions, rain tips, etc.), rather than to try to calculate from a very limited number of "counts" in a given period. Of course with the FO (and probably most rain gauges), it takes quite a long time to get even two tips (i.e. up to 0.6mm of rain) before you could make the calculation. But does a single tip have any significance? It could have been just one raindrop which "tipped the scales".

That's how the Davis works, and that's what I've implemented in Cumulus (but it's currently unused). The rain rate is zero until the second tip within a certain period (I forget how long it is on the Davis), and the rate depends on the time between successive tips.