Forums

Back

Re: [aor-rtma] Calm winds near Guadalupe Pass

RSS (Opens New Window)(Opens New Window)
JC
Jacob Carley, modified 8 Years ago.

Re: [aor-rtma] Calm winds near Guadalupe Pass

Youngling Posts: 69 Join Date: 12/17/14 Recent Posts
Hi All,

Thanks for the excellent discussion.  I/we just wanted to provide a quick update.  After looking into this particular case it was found that there was a bug in one of the RTMA/URMA's QC routines for mesonet winds which would have helped address the issue.  The QC routine is designed to wind toss very light mesonet wind observations when the background is much higher than the ob (Tyndall and Horel, 2013; WaF), and it was not applied properly for wind speed.  This is fixed in the v2.6 RTMA/URMA package we are in the process of spinning up now for field evaluation.

However, as has been discussed, the big issue is improving the obs QC such that stuck instrument problems can be detected and flagged appropriately.  We're looking into ways in which we can start doing this effectively for future RTMA/URMA packages (which would also positively impact other systems in the production suite).

Thanks!
Jacob

On Thu, Apr 27, 2017 at 4:03 PM, Jeffrey Craven - NOAA Federal <jeffrey.craven@noaa.gov> wrote:
John thanks for the thorough explanation.  I just happened to notice it because that location tends to forecast rather high winds in NBM and I am always hoping to see a gust there that exceeds KGDP.   So the 7 knot URMA speed caught my attention.

Not a trivial problem so thanks for your help.

JPC

On Thu, Apr 27, 2017 at 2:55 PM John Horel <u0035056@umail.utah.edu> wrote:
Jeff-

We've been working for some time on how to implement additional data checks to the real-time publicly-accessible data such as from RAWS. The attached image is a screenshot of what will be a way to monitor issues by networks and stations. All of the data checks will be accessible by early summer through the SynopticLabs/MesoWest api that populates many NWS web pages for example via NWS Western Region.  We're meeting with MADIS folks in a couple of weeks- how that data checking might be transmitted to NCEP for the  RTMA/URMA and other applications will be a topic we'll discuss with them.

Progress has been slow, because in part doing even basic data checking can get pretty complicated given the volume of data. Even to catch the no-brainers like this one at GDBT2- the wind sensor is likely busted or offline- can be tough. You'll see in the screenshot that we track when the sensor begins to appear to be stuck until the situation is resolved. If an hour block is colored light yellow, then that means this "persistence" check is being triggered for that station. Out of the 2200 or so RAWS stations available, it looks like 28 are having that problem today. Then, I went too far and double checked WALN4 to verify that it was fixed earlier today (no color after 3 UTC), and we found another odd ball case not being handled correctly- it remains stuck on 0 through the most recent observation. 

No promises on how  or when this'll all get integrated operationally. 

Regards

John Horel
U/Utah

On Thu, Apr 27, 2017 at 11:42 AM, Jeffrey Craven - NOAA Federal <jeffrey.craven@noaa.gov> wrote:
Looks like this observation (GDBT2) is making it into URMA.    It is at about 7800 feet, and the EPZ sounding was showing 45 knots a couple of days ago at that level.  Interesting that it wasn't thrown out.

JPC

Inline image 2

Inline image 1

Inline image 3

Inline image 4

Jeff Craven
Chief, Statistical Modeling Branch
National Weather Service
Meteorological Development Laboratory (MDL)
Room 10410, SSMC2
Silver Spring, MD 20910
(816) 506-9783 cell/text
@jpcstorm

 


--
Jeff Craven
Chief, Statistical Modeling Branch
National Weather Service
Meteorological Development Laboratory (MDL)
Room 10410, SSMC2
Silver Spring, MD 20910
(816) 506-9783 cell/text
@jpcstorm

 


 

Attachments:

Attachment

Attachment

Attachment

Attachment