Forums

Back

High Dew Point Bias

RSS (Opens New Window)(Opens New Window)
DB
David Bonnette, modified 1 Year ago.

High Dew Point Bias

Youngling Posts: 4 Join Date: 3/31/18 Recent Posts

Howdy! 

 

We have noticed on deeply mixed as well as light wind days that the URMA's dew point has a high bias in our forecast area; North & Central Texas/Dallas-Fort Worth. We have a uniquely high number of CWOP & backyard weather stations that report anomalously high dew points when compared with nearby METAR values (see last slide of the linked slides) that are being ingested into the RTMA & URMA.  We scrubbed through the kml files on a couple of these afternoons and have a list of obs that require blacklisting. Hopefully removing those obs will fix or at least improve the issue. This was reinforced the past couple of days when we noticed an obvious speckling in the NBM where each grid with a METAR is ~2-5 degrees lower than the surrounding grid boxes. 

 

Here is a quick slide presentation with more information and some images to illustrate the points above:

https://docs.google.com/presentation/d/1085DzsGcRXn1jAzH_Az2QKZs4_FSwtkufFxMQ-7HQIA/edit?usp=sharing

 

Thank you!

-David

WFO FWD

MM
Matthew Morris, modified 1 Year ago.

RE: High Dew Point Bias

Youngling Posts: 158 Join Date: 12/6/17 Recent Posts
Hi David,

Thanks for sharing this slide deck, which provides context for the bulk flagging request.  We are investigating this issue and will follow up with you once we learn more.

Thanks,
Matt

On Thu, Jul 7, 2022 at 5:48 AM VLab Notifications <VLab.Notifications@noaa.gov> wrote:

Howdy! 

 

We have noticed on deeply mixed as well as light wind days that the URMA's dew point has a high bias in our forecast area; North & Central Texas/Dallas-Fort Worth. We have a uniquely high number of CWOP & backyard weather stations that report anomalously high dew points when compared with nearby METAR values (see last slide of the linked slides) that are being ingested into the RTMA & URMA.  We scrubbed through the kml files on a couple of these afternoons and have a list of obs that require blacklisting. Hopefully removing those obs will fix or at least improve the issue. This was reinforced the past couple of days when we noticed an obvious speckling in the NBM where each grid with a METAR is ~2-5 degrees lower than the surrounding grid boxes. 

 

Here is a quick slide presentation with more information and some images to illustrate the points above:

https://docs.google.com/presentation/d/1085DzsGcRXn1jAzH_Az2QKZs4_FSwtkufFxMQ-7HQIA/edit?usp=sharing

 

Thank you!

-David

WFO FWD


--
David Bonnette RTMA/URMA Discussion Group Virtual Lab Forum https://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/23285825VLab.Notifications@noaa.gov
MM
Matthew Morris, modified 1 Year ago.

RE: High Dew Point Bias

Youngling Posts: 158 Join Date: 12/6/17 Recent Posts
Hi David,

The requested mesonet stations are now being blacklisted in RTMA/URMA.  Most stations were added to the blacklist last Friday, July 15th, with the remainder added yesterday, July 18th.  Hopefully, this will alleviate the biases observed in the DFW area.  We can also look into addressing this issue using an automated QC package that is under development for the future 3D-RTMA/URMA system (implementation expected in late 2024); this could take into account the station metadata (e.g., "aspirated" vs. "non-aspirated" stations).

Thanks,
Matt

On Mon, Jul 11, 2022 at 4:41 PM VLab Notifications <VLab.Notifications@noaa.gov> wrote:
Hi David,

Thanks for sharing this slide deck, which provides context for the bulk flagging request.  We are investigating this issue and will follow up with you once we learn more.

Thanks,
Matt

On Thu, Jul 7, 2022 at 5:48 AM VLab Notifications <VLab.Notifications@noaa.gov> wrote:

Howdy! 

 

We have noticed on deeply mixed as well as light wind days that the URMA's dew point has a high bias in our forecast area; North & Central Texas/Dallas-Fort Worth. We have a uniquely high number of CWOP & backyard weather stations that report anomalously high dew points when compared with nearby METAR values (see last slide of the linked slides) that are being ingested into the RTMA & URMA.  We scrubbed through the kml files on a couple of these afternoons and have a list of obs that require blacklisting. Hopefully removing those obs will fix or at least improve the issue. This was reinforced the past couple of days when we noticed an obvious speckling in the NBM where each grid with a METAR is ~2-5 degrees lower than the surrounding grid boxes. 

 

Here is a quick slide presentation with more information and some images to illustrate the points above:

https://docs.google.com/presentation/d/1085DzsGcRXn1jAzH_Az2QKZs4_FSwtkufFxMQ-7HQIA/edit?usp=sharing

 

Thank you!

-David

WFO FWD


--
David Bonnette RTMA/URMA Discussion Group Virtual Lab Forum https://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/23285825VLab.Notifications@noaa.gov

--
Matthew Morris RTMA/URMA Discussion Group Virtual Lab Forum http://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/23378136VLab.Notifications@noaa.gov
JS
Jack Settelmaier, modified 1 Year ago.

RE: High Dew Point Bias

Youngling Posts: 23 Join Date: 11/2/12 Recent Posts

This is a great write-up David.  I wonder how unique the situation of CWOP and backyard stations is to just the FWD area? 

 

Nice to see that you've detailed the issue and Matt has applied a mitigation step by excluding sites.

 

Matt, do you have any details on how the "automated QC package that is under development for the future 3D-RTMA/URMA system (implementation expected in late 2024)" differs from the present RTMA/URMA QC checks?  I've looked around a bit on this VLab site, but am not seeing any more details than what is shown in the link to a COMET doc of various gross and buddy checks.

MM
Matthew Morris, modified 1 Year ago.

RE: High Dew Point Bias

Youngling Posts: 158 Join Date: 12/6/17 Recent Posts
Hi Jack,

The initial work on the automated quality control (AutoQC) package has been to mirror the QC in the operational system.  Additional techniques will be incorporated prior to the first 3D-RTMA implementation in FY2024, such as temporal consistency checks.  One key advantage of using the AutoQC package is that the reject and accept lists can be updated in real-time, rather than needing to wait for a system upgrade.

Thanks,
Matt

On Fri, Aug 5, 2022 at 12:33 PM VLab Notifications <VLab.Notifications@noaa.gov> wrote:

This is a great write-up David.  I wonder how unique the situation of CWOP and backyard stations is to just the FWD area? 

 

Nice to see that you've detailed the issue and Matt has applied a mitigation step by excluding sites.

 

Matt, do you have any details on how the "automated QC package that is under development for the future 3D-RTMA/URMA system (implementation expected in late 2024)" differs from the present RTMA/URMA QC checks?  I've looked around a bit on this VLab site, but am not seeing any more details than what is shown in the link to a COMET doc of various gross and buddy checks.


--
Jack Settelmaier RTMA/URMA Discussion Group Virtual Lab Forum https://vlab.noaa.gov/web/715073/discussions-forums-/-/message_boards/view_message/23914452VLab.Notifications@noaa.gov