[wxqc] precision, accuracy and statistics (oh My)

Mark Kimble mjkits at rit.edu
Wed Apr 2 09:11:22 CDT 2008


HA.  I waited a long time to get a "weather station."  I'm not exactly
sure what the drive was, but it was always there.  I had to have one.
So finally I dropped down the bucks, took a few hours to install it  -
with its base station directly beneath my house's HVAC thermostat.  I
was immediately thrilled to actually KNOW the temp, wind, humidity, and
rainfall.  

 

Then I noticed that the indoor temperature reading was higher than the
house thermostat.  Having grown up with a father who kept our thermostat
OFF I was shocked that my house could be 68 degrees instead of the
luxurious 65 I keep it set at.  Shocked I tell you.  I mean, the new
"expensive" weather station MUST be the accurate temperature - right!?
So anyway, I started keeping paper records of temperature with four
different sensors and found that they actually tracked each other very
well with a total difference of 3 degrees.  Checking further I found
they all appeared to be operating within their particular specified
level of accuracy but I still had a problem.  I didn't know what the
temperature INSIDE my house was.

 

I started poking around the internet and found a site at an engineering
school that reminded me about precision and accuracy (and felt a little
better).  But then (after better than a week of obsessing with this) I
saw THIS on this same .edu site:  "Give a man a watch and he will always
know what time it is.  Give him two, and he'll never know for sure."  I
let out a deep sigh and gave up.

 

A few months later I registered my weather station with CWOP and within
a few days of QC I got the "dreaded" notice that my data (out door
temperature specifically) was "not within blah-blah-blah tolerances."  A
few minutes of examination I found that my outside temp TOO was about
three degrees above the three other temps I have around the yard.  

 

I made calibration changes to both my indoor and outdoor settings.  The
Davis Vantage Pro people may think I'm wrong to do it, but at least I'm
passing someone's scrutiny  (thanks, Philip et al)  But here's the
thing.  I still don't really know what the temperature is.  Sigh.

 

 

Mark Kimble

 

DW0129, KNYCHURC2 etc etc

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://server.gladstonefamily.net/pipermail/wxqc/attachments/20080402/12762219/attachment.html 


More information about the wxqc mailing list