[wxqc] Temperature "errors"

[Evan Bookbinder]

Gary,

I guess a little sarcasm is good now and again, but I found that your 
tongue-in-cheek comments were a bit out of line. Several members of this 
group have provided excellent documentation on what QC is, how it's done 
and what it means. These insightful documents likely would have 
circumvented the need for your email.

Ignoring some incorrect assumptions made on the development of frost and 
the killing of tender vegetation, let's make the obvious assumption that 
your observation of 32F is correct.

1.) Minimum temperatures are HIGHLY susceptible to terrain and land 
cover/usage. It is perfectly normal to have a 10 or more degree spread 
over several miles and a hundred feet elevation change on a clear, calm 
night. (I'm not going to get into a dissertation on cold air drainage 
and nocturnal inversions here).

2.) Not NOAA, nor anyone else, is rejecting anyone's data, contrary to 
your belief. We receive all data, good and bad. The QC tool merely 
provides you with the feedback necessary to determine whether an 
instrument could be failing or is incorrectly sited. These reports 
should never be used on a daily basis (due to local effects, frontal 
passages, etc...) and a long term average on the order of a few weeks 
should be used to determine systematic errors with your instrumentation. 
In this case, it's obvious that your local siting plays a huge roll in 
cold air drainage at night. The algorithm is nothing more than a 
mathematical equation used to relate your readings to surrounding sites. 
There's no standard "negative lapse rate" for cold air draining, so 
there's nothing that can be done to factor elevation into the 
algorithm.  So, if your daytime highs are QCing just fine, feel free to 
ignore the QC message on the min temperature.

Hope this helps,
Evan