Re: do you have a simple way of roughly estimate the max frequency? how about this interpolation scheme?
From: Cyrus Yunker (cyrus.yunker_at_gmail.com)
Date: 12/29/04
- Next message: George Bush: "Standard deviation of angular measurements"
- Previous message: Chunhui Mei: "help to solve a equation"
- In reply to: Ronald H. Nicholson Jr.: "Re: do you have a simple way of roughly estimate the max frequency? how about this interpolation scheme?"
- Next in thread: Tim Wescott: "Re: do you have a simple way of roughly estimate the max frequency? how about this interpolation scheme?"
- Messages sorted by: [ date ] [ thread ]
Date: 29 Dec 2004 09:16:02 -0800
Ronald H. Nicholson Jr. wrote:
> In article <3384ukF3rpa1gU1@individual.net>, Jerry Avins
<jya@ieee.org> wrote:
> >You can't sample rainfall at all because at any instant there's ...
> >drop or there isn't, and sampling is about what happens at insta...
> >When a drop hits, the derivative is, at the very least, very lar...
> >between drops, it is zero. So sampling the derivative isn't much...
>
> On any of the visual rain gauge I've used, a single rain drop wou...
> register above measurement error (noise). Fortunately, a rain ga...
> an integrator, so one can just wait long enough to see about how ...
> it takes for the change in measurement to become perceptible, and...
> use the reciprocal to get the derivative within your measurement ...
> bounds.
>
> The derivative is quite important in terms of choice of clothing ...
> rain gear. An inch of downpour in 10 minutes will have quite a d...
> effect on ones comfort outdoors compared an inch of sprinkles spr...
> roughly evenly over 24 hours (other environmental conditions bein...
> equal). Due to evaporation effects and the possibility of damage...
> blossoms, this rate makes a difference to farmers as well.
A look into the methods of measuring rainfall I think would be
appropriate here. Often what's used, in a situation where rainfall is
to be monitored automatically, is what's called a "tip bucket rain
guage." This device outputs a pulse train or triggers a contact
closure every time a fixed, yet small, amount of rainfall occurs. A
small vessel is split in two and is fixed atop a hinge of sorts. When
enough rain enters one half to cause it to tip over, a switch closure
event happens and that side is emptied as the opposite side begins to
fill. Repeating events can be logged (with time, date) or multiple
events can be counted and the total logged on a periodic basis
(sampling) either with or without a reset. Rainfall rate can be
determined from the count reached between samples and rainfall total
over the course of a day, week, or month can be easily determined from
a running total. No need for instantaneous samples or trying to
measure "visual" rainfall guages (or individual drops!).
The limiting factor is how small of an event you'd like to measure. If
your bucket doesn't tip, the water either evaporates and is not
recorded or is carried forward to the next event, skewing results
there. (You could attempt to model or correct for this, adding a half
bucket to each event where little rain and hot weather follows (or low
RH%), and or where cool weather and little time separated two events.)
The rate resolution you'd like determines your sampling period. Better
may be to record each closure and "sample" (time divide) however you'd
like later. This would reduce data set size and allow for flexibility
later, however some data preparation may be required. Any significant
rainfalls though may actually increase data set size beyond periodic
sampling of counts. Your approach should take into account a decent
appraisal of expected rain events, the capabilities of your data
storage device, and how often you can retrieve collected data.
My recommendation would be to setup your datalogger to timestamp and
record non-zero counts every two minutes, resetting the counter each
time. Adjust to taste. If you require a large dynamic range perhaps
log a large and a small set of buckets - small for sensitivity and
large to reduce the chance of saturation and missing drops due to
splashing during a heavy event. Discount the large bucket (don't add
it!) if the rate wasn't especially high. Factor it in if the large and
small bucket data doesn't match. Quantized, pseudo-integrated rain
data? Hope this helps.
-- @ Ag. Engineering Help? @ Random Advice? @ questions __ 80d ___ org
- Next message: George Bush: "Standard deviation of angular measurements"
- Previous message: Chunhui Mei: "help to solve a equation"
- In reply to: Ronald H. Nicholson Jr.: "Re: do you have a simple way of roughly estimate the max frequency? how about this interpolation scheme?"
- Next in thread: Tim Wescott: "Re: do you have a simple way of roughly estimate the max frequency? how about this interpolation scheme?"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
