[wxqc] Rain gauge cal
gladstonefamily45.net at tedlum.com
Tue Oct 11 22:32:28 CDT 2011
But in all seriousness, some things don't carry over. In the calibration
lab its important to account for all of those little items, whether it
be the accuracy, tolerance, resolution, uncertainty, etc. Becasue in the
calibration lab you need a standard that is many times more accurate,
sensitive, resolute, etc. in order to measure that 5% accuracy of the
sensor under test. That's how you know its +/- 5% and not +/- 6% or +/-2%.
So, when you are talking calibrations and standards measurements, yea,
all those little fractions are important. In the field is a completely
different situation, you don't expect the same specs. as you had in the
lab. But too often people change context not taking that into account.
The most glaring this here, though, is that we all - myself most
centrally included - have been running around calling the Davis gauge an
8" literally for years. Leave it to someone to actually measure the damn
thing... what a concept. Not sure where 8" got stuck in my head, but it
did, and I've been real wrong for years. That in and of itself renders
the whole argument moot, because that's a pretty big error before you
even get started.
But, regardless of the specifics the point that I have been trying to
get across is this. Think of a VHS tape... remember those? Every time
you made a copy of a copy you would loose 1/3 or the resolution. Same
idea with references and standards. Every time you use a standard to
judge another instrument expect to loose something... You can't use a 5%
instrument to judge the accuracy of another instrument to withing 5%.
And for each generation you lose more and more. But this is for
calibrations and standards, it does not carry over into other contexts.
On 10/11/2011 6:20 PM, Paul Grace wrote:
> Thanks- It's humorous to see people attempting to compensate for 0.1%
> errors due to the variable density of rain, against the basic 5%
> accuracy of the sensor.
> *From:*wxqc-bounces at lists.gladstonefamily.net
> [mailto:wxqc-bounces at lists.gladstonefamily.net] *On Behalf Of
> *jdr1 at metdata.com
> *Sent:* Tuesday, October 11, 2011 1:35 PM
> *To:* wxqc at lists.gladstonefamily.net
> *Subject:* [wxqc] Rain gauge cal
> I noticed there are several people that are trying to calibrate their
> Davis gauges to a 4"gauge. Victor referred to the Davis gauge as an
> 8" gauge, it's not an 8", it is a 6 ½ inch gauge I have two of the
> Davis tippers on the same post mounted at 5' above the ground on two
> different Davis Stations. They rarely have had the same rainfall
> amount over the three years they have been mounted together. I had a
> 8" NWS standard non-recording gauge only a few feet away several years
> ago and had the same problem of differences in the amount the tippers
> recorded and the 8" recorded.
> I have been a Davis dealer for 18 years and have talked to their
> engineers on several occasions about these differences. To check the
> calibration on the Davis tipper, 5.44 ml equals one tip. The accuracy
> of the Davis Tipper is +/- 4%, +/- 1 rainfall count between .01"and
> 2.00" per Hour. +/- 5%, +/-1 rainfall count between 2.00" and 4.00"
> per hour. If you want to be absolutely sure that the Davis gauge is
> accurate, you can send it back to Davis and have it NIST certified, or
> if you are purchasing a new station then request a NIST certified rain
> gauge. Davis Instruments web site has a page on rain gauge
> calibration, www.davisnet.com <http://www.davisnet.com/>.
> Rainfall is extremely variable and can vary in amount over a few feet
> or even inches, this depends on the wind or like of. The NWS people
> even admit that the tipper is sometimes not accurate in high rainfall
> rates or high wind. If a tipper is mounted on a post that can be
> shaken by the wind there will false tips and inaccurate rainfall
> readings. Mounting and siteing of a rain gauge is very important,
> NOAA has several publications on the proper siting or weather
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the wxqc