[wxqc] Analysis Question

Gerry Creager gerry.creager at tamu.edu
Thu Dec 11 15:00:50 CST 2008

Rich Mulvey wrote:
>    I think that the people who keep telling us that they're
> recalibrating to match the analysis have a fundamental misunderstanding
> about what the analysis is telling them.
>    Never, EVER recalibrate your station to match the analysis, unless
> you have verified that your station is, in fact, incorrect.  The
> analysis is a projection of the conditions at your station.  It does not
> take into account microclimates.  It does not take into account local
> conditions that may be affecting the reports, like if someone put their
> thermometer on a rooftop without a suitable enclosure.  And so on.
>    Just because a station is a few miles from an airport does not mean
> that your station's readings will be the same as the airport's.  For
> example, my station is almost exactly 5 straight-line miles from KROC. 
> My temps are often 5 degrees or more different from it.  My humidity is
> often wildly different because I'm in a narrow stream valley.  My
> pressure tracks KROC's trends, but not its value.  And so on.  And yet
> my readings are accurate, as determined by other independent
> instruments.  So my analysis occasionally flags my readings as out of
> range, even though they're not.
>    You should be using the analysis as a starting point, to determine if
> your station is having problems.  If you consistently get temps or
> humidity out of range, then check to see if you have a siting issue with
> your thermometer or humidity sensor.  Maybe the fan in your aspirated
> pagoda has died.  Maybe a spider has spun a web around the humidity sensor.
>    But if you're confident that your readings are correct, then ignore
> the analysis until it starts showing even larger divergences than your
> norm.  That should prompt you to check your equipment again.  Until
> then, don't worry about it.

OR, you might find that your station is fine, but a nearby one is out of 
kilter.  We've seen that with AWOS before...


More information about the wxqc mailing list