[Weather] dual drains -- was Re: conversion help

[Mark J]

MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii


In message <20061121052735.40084.qmail at web31711.mail.mud.yahoo.com>
          "Herbert E. Plett" <cachureos at yahoo.com> wrote:


> --- claude felizardo <cafelizardo at usa.net> wrote:

> > What about two sets of drains similar to the condensate drains in an air
> > conditioning unit?  A relatively small drain into a small pair of
> > buckets for counting light rain, then an "overflow" drain into a larger
> > bucket for counting down pours.  Does anyone make anything like that?


> at first it sounds reasonable, but...

> how do you know when the 'large' bucket is almost full because of a pour and
> then rain rate falls back to 'light'?
> in these configurations big and small buckets steal rate mutually and the end
> result is anything but accurate.
> at least if you build the funnel that doesn't overload one small bucket, the
> worst case scenario will be a delay in measure of pour rate (rate clipping),
> but no error in total water fall.

> a rain gauge for both, very heavy rain and very light rain at the same spot, is
> a real challenge and most probably not solvable with a tipping bucket type.


Yes. And I for one am trying to record some idea of what's actually
happening, and the tipping bucket doesn't do it at all well except for
steady conditions within its design limitations. In that previous
thread I suggested an upper limit of 300mm/hr (laCrosse), and I'm
happy that at that rate, over one hour, I will record close to
300mm/hr and will get a cumulative total of near 300mm. But rain's
never like that...

This morning was typical: I knew it was raining because my
conservatory roof was shouting it to me in the dark, but the raingauge
count didn't move, and it hasn't moved since; officially, it hadn't
rained, but tell that to the postman...

This type of design has two weaknesses: the funnel, and the bucket :-)

The funnel needs to be very large to record light rain, and very small
to record heavy; it needs to deliver every tiny drop to its tip
without beads of water hanging all over its surface, and none should
splash out - and it needs to deliver it without delay, lest rainfall
rate be distorted, but not so fast as to impart KE, lest rainfall rate
be distorted...

The bucket's a design disaster. We know that. It's insensitive to
light rainfall, and gets overwhelmed with heavy; and rainfall rate
measurement is in conflict with that of total rainfall. The last is
easy: use a measuring cylinder. The former is not so easy, as it
involves mass per units of time or time per units of mass; counts per
minute or minutes per count. Instantaneous measurement is not
possible, but is inferred; mm/hr, but for how long? Over one hour,
very good; but the rain might have stopped long since, and the peak
was missed...

It needs fresh thought, and no; I haven't any sensible suggestion
either :-)

[snip]

-- 
E.Mark Jolliffe
www.poppy-land.co.uk