[wxqc] Comparing stations

Paul Grace paulgrace at lookoutranch.com
Sun Oct 30 12:21:22 CDT 2011

Thanks for the additional detail!

-----Original Message-----
From: wxqc-bounces at lists.gladstonefamily.net
[mailto:wxqc-bounces at lists.gladstonefamily.net] On Behalf Of Mike Hardiman
Sent: Sunday, October 30, 2011 10:07
To: Discussion of weather data quality issues
Subject: Re: [wxqc] Comparing stations

Hi folks,

There seems to be a misconception out there that MADIS is "dumping" CWOP
stations based on these quality control tools.  To an NWS WFO, MADIS is
merely a data conduit, or data aggregator.  While they do offer mesonet data
in comma-delimited format at the various QC levels, the "AWIPS-ready" data
feed we get at the NWS has no quality control filters imposed.  So needless
to say, plenty of users are getting your data no matter what those pages

At my office, I do manual Quality Control at two levels.  Consistent
'problem stations' where data is clearly out of whack (e.g., 105 degree
temperatures at a station whose elevation is over 8,000 feet, or dewpoints
that are clearly off-base -- easy to find in the desert) will get
'blacklisted' right out front and never get into AWIPS.  I can blacklist an
entire station, or just specific elements from a station. 
There's a second level of manual QC which is more fluid and it decides what
stations/elements go into our local LAPS analysis and Observation Grids
which provide unofficial forecast verification and 'bias-corrected' forecast
model grids.  Usually here I'll block out stations with less-than-ideal wind
siting that would not work well in a general analysis, and occasional
"hiccups" from otherwise good stations, that I can later go in and clear out

The MADIS QC ratings probably do affect what goes into their 15km RSAS
analysis grids.  Frankly, I'm not sure who uses this product, considering
there is the 4km RTMA available nationwide, the 4km LAPS analysis at the WFO
level, and the coarser-resolution MSAS surface analysis.  The software we
use to run local WRF model runs can use the LAPS analysis as part of its
initialization, and future releases promise to make it easier to incorporate
local observations into the initialization as well.

So, to reiterate, your obs are not going down the drain just because you've
got a bunch of red Xs or "thumbs down" icons!

For example, the original poster's station, DW1453 still shows up in

Granted, there's a "Caution" flag, but there's one for a nearby National
Ocean Service station as well.  The only thing that looks a little off
(based on the areal surface plot) is the winds are much lighter than the NOS
station and the ob at Ketchikan airport.  If I were deciding what to do with
the ob, the only thing in question would be the wind, and even then I'd
watch it over time, check out the terrain and location, and if the guy was a
spotter or something, I might send him a message to ask about the siting or
what he thought was the cause.

Wind is probably the most difficult element to site a station for, and in
many cases there's nothing a property owner can do, especially in a heavily
wooded area, or in suburban areas.  But even a less-than-ideally sited wind
observation isn't worthless ... an otherwise "sheltered" 
anemometer could be right underneath a thunderstorm microburst someday and
verify a warning!

It looks like the winds should be howling through that inlet, but you can
get weird nearly-stationary eddies in steep terrain areas.

Hope this helps clear the air a bit.

wxqc mailing list
Post messages to wxqc at lists.gladstonefamily.net To unsubcribe or change
delivery options, please go to:
To search the archives:

The contents of this message are the responsibility of the author.

More information about the wxqc mailing list