[wxqc] wxqc Summary Page

[bfgiii at adelphia.net]

---- Philip Gladstone <philip at gladstonefamily.net> wrote: 

=============
Ben

The Error Standard Deviation is a measure of how much variability there 
is in the error. In this case, if the mean error is (say) 10 degrees, 
and the standard deviation is (say) 5 degrees, then most readings will 
have an 'error' of between 5 and 15 degrees.

The purpose of displaying this (and using it) is that a small SD 
indicates that the sensor is tracking the analysis well. The mean error 
may indicate a calibration problem.

For example, with pressure readings, it was the case that many stations 
did not report the altimeter pressure reading (roughly sea level 
pressure, but subtly different). Their readings would track the analysis 
reasonably well, but with a noticeable offset. The use of the SD 
distinguished these sites from those with a pressure sensor that was 
broken, or a site with incorrect lat/long information (so the analysis 
pressure was from a different part of the country!).

Does this help?

Philip

Hello Philip,
Thanks for your consideration.  I see now that my SD is not out of line at all, especially when I compare it to the airport ASOS 3.4 miles from my station and two other good stations within a reasonable distance from here.

http://weather.gladstonefamily.net/site/K3B2
http://weather.gladstonefamily.net/site/C0467

I'm guess that I'm as close in my mean error and SD as I could hope to be and I don't want to try and fix something that isn't broken.  Do you agree?

Re: the BP auto scaling thread, we get some considerable movement in pressure this time of year when cold fronts approach and move off the New England coast.  We've had many instances of intense T-storms falling completely apart as the hit the marine air layer about 5 to 10 miles in from the shore only to reform again over Massachusetts Bay.  The BP can jump all over as these cells degrade, dissipate and reform again as they march out to sea.  Some wild runs in temp/humidity, too.

Thanks again, Philip.

Ben Goodrich
CW0467