Forums

Back

RE: Too low Winds in NBM (from URMA?) over Florida Keys

RSS (Opens New Window)(Opens New Window)
JS
Jack Settelmaier, modified 2 Years ago.

Too low Winds in NBM (from URMA?) over Florida Keys

Youngling Posts: 23 Join Date: 11/2/12 Recent Posts
WFO Key West shared a case with SRH STSD, describing an instance where the URMA-analyzed winds appear to be too low, when compared to available observation platforms in the nearby water and land areas.  

These too-low URMA-analyzed wind speeds appear to be similar to the too-low NBM wind speed forecasts.  If one uses the URMA analysis, vs nearby point locations, as observed data for verification purposes, resulting skill assessments favor the NBM over the NDFD.  

What appears odd in this case, is how the RTMA analysis seems to more closely match the nearby observations, but something in the URMA processes appears to have lowered the analyzed values.  Could it be the URMA made use of zero wind speed values at some CWOP sites?  URMA KML files appear to indicate the suspect zero wind speed from E9599 was rejected.
I wanted to test if that was true in looking at the parallel v2.10 RTMA/URMA data, in which CWOP data is NOT used, but it aged off before I could view that.

Link to Google Slides.
MM
Matthew Morris, modified 2 Years ago.

RE: Too low Winds in NBM (from URMA?) over Florida Keys

Youngling Posts: 169 Join Date: 12/6/17 Recent Posts
Hi Jack,

According to the figure labels, the RTMA figures on slides 9 and 12 are for wind gust, whereas the URMA figures on slides 10 and 13 are for wind speed.  Thus, this is not a valid comparison of the figures.  Are the values reported on slides 11 and 14 also comparing wind speed to wind gust?

Thanks,
Matt

On Fri, Jan 27, 2023 at 4:06 PM VLab Notifications <VLab.Notifications@noaa.gov> wrote:
WFO Key West shared a case with SRH STSD, describing an instance where the URMA-analyzed winds appear to be too low, when compared to available observation platforms in the nearby water and land areas.  

These too-low URMA-analyzed wind speeds appear to be similar to the too-low NBM wind speed forecasts.  If one uses the URMA analysis, vs nearby point locations, as observed data for verification purposes, resulting skill assessments favor the NBM over the NDFD.  

What appears odd in this case, is how the RTMA analysis seems to more closely match the nearby observations, but something in the URMA processes appears to have lowered the analyzed values.  Could it be the URMA made use of zero wind speed values at some CWOP sites?  URMA KML files appear to indicate the suspect zero wind speed from E9599 was rejected.
I wanted to test if that was true in looking at the parallel v2.10 RTMA/URMA data, in which CWOP data is NOT used, but it aged off before I could view that.

Link to Google Slides.

--
Jack Settelmaier RTMA/URMA Discussion Group Virtual Lab Forum http://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/27307683VLab.Notifications@noaa.gov


--
Matthew Morris
SAIC at NOAA/NWS/NCEP/EMC
5830 University Research Ct., Rm. 2038
College Park, MD 20740
301-683-3758
JS
Jack Settelmaier, modified 2 Years ago.

RE: Too low Winds in NBM (from URMA?) over Florida Keys

Youngling Posts: 23 Join Date: 11/2/12 Recent Posts
Thanks Matt, maybe I messed up.  Please do correct me, with any input you have on the values from the various systems.

Let me check on that tomorrow, assuming I still can get access to the files needed.

On Mon, Jan 30, 2023 at 11:47 AM VLab Notifications <VLab.Notifications@noaa.gov> wrote:
Hi Jack,

According to the figure labels, the RTMA figures on slides 9 and 12 are for wind gust, whereas the URMA figures on slides 10 and 13 are for wind speed.  Thus, this is not a valid comparison of the figures.  Are the values reported on slides 11 and 14 also comparing wind speed to wind gust?

Thanks,
Matt

On Fri, Jan 27, 2023 at 4:06 PM VLab Notifications <VLab.Notifications@noaa.gov> wrote:
WFO Key West shared a case with SRH STSD, describing an instance where the URMA-analyzed winds appear to be too low, when compared to available observation platforms in the nearby water and land areas.  

These too-low URMA-analyzed wind speeds appear to be similar to the too-low NBM wind speed forecasts.  If one uses the URMA analysis, vs nearby point locations, as observed data for verification purposes, resulting skill assessments favor the NBM over the NDFD.  

What appears odd in this case, is how the RTMA analysis seems to more closely match the nearby observations, but something in the URMA processes appears to have lowered the analyzed values.  Could it be the URMA made use of zero wind speed values at some CWOP sites?  URMA KML files appear to indicate the suspect zero wind speed from E9599 was rejected.
I wanted to test if that was true in looking at the parallel v2.10 RTMA/URMA data, in which CWOP data is NOT used, but it aged off before I could view that.

Link to Google Slides.

--
Jack Settelmaier RTMA/URMA Discussion Group Virtual Lab Forum http://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/27307683VLab.Notifications@noaa.gov


--
Matthew Morris
SAIC at NOAA/NWS/NCEP/EMC
5830 University Research Ct., Rm. 2038
College Park, MD 20740
301-683-3758

--
Matthew Morris RTMA/URMA Discussion Group Virtual Lab Forum http://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/27354573VLab.Notifications@noaa.gov


--
Jack Settelmaier
Digital Techniques Meteorologist
NOAA/NWS, Southern Region HQ
Fort Worth, TX 
Work Desk: 682 703 3685
Virtual Office (most core work hours): https://meet.google.com/ujm-ajkv-rhk
JS
Jack Settelmaier, modified 2 Years ago.

RE: Too low Winds in NBM (from URMA?) over Florida Keys

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

Indeed you were correct, that I mixed Wind Speed and Gusts.  I have now corrected slides 9-11 to now compare RTMA and URMA Wind Speeds.  [I've also updated slides 12-14, but have Hid them, since they are so alike and there is no point to make].

So, the only questions I have from this review are:
1) Why did URMA slightly LOWER the 14Z Wind Speed analysis value from what the RTMA had, in the vicinity of LONF1, especially since it appears the LONF1 observation of 11.13 knots (slide 15) was not rejected.
2) Also, can you explain the lower portion of Slide 15 with regard to the iterations.  Am I reading correctly, the (HRRR-supplied) background value was 10.19, and after the LONF1 observation of 11.13 was considered, and not rejected (aka accepted) the successive iterations actually lowered the analysis value down to an eventual 9.41 knows (which matches the slides 10-11 image/values). Was the LONF1 value not actually used?

On Mon, Jan 30, 2023 at 4:17 PM VLab Notifications <VLab.Notifications@noaa.gov> wrote:
Thanks Matt, maybe I messed up.  Please do correct me, with any input you have on the values from the various systems.

Let me check on that tomorrow, assuming I still can get access to the files needed.

On Mon, Jan 30, 2023 at 11:47 AM VLab Notifications <VLab.Notifications@noaa.gov> wrote:
Hi Jack,

According to the figure labels, the RTMA figures on slides 9 and 12 are for wind gust, whereas the URMA figures on slides 10 and 13 are for wind speed.  Thus, this is not a valid comparison of the figures.  Are the values reported on slides 11 and 14 also comparing wind speed to wind gust?

Thanks,
Matt

On Fri, Jan 27, 2023 at 4:06 PM VLab Notifications <VLab.Notifications@noaa.gov> wrote:
WFO Key West shared a case with SRH STSD, describing an instance where the URMA-analyzed winds appear to be too low, when compared to available observation platforms in the nearby water and land areas.  

These too-low URMA-analyzed wind speeds appear to be similar to the too-low NBM wind speed forecasts.  If one uses the URMA analysis, vs nearby point locations, as observed data for verification purposes, resulting skill assessments favor the NBM over the NDFD.  

What appears odd in this case, is how the RTMA analysis seems to more closely match the nearby observations, but something in the URMA processes appears to have lowered the analyzed values.  Could it be the URMA made use of zero wind speed values at some CWOP sites?  URMA KML files appear to indicate the suspect zero wind speed from E9599 was rejected.
I wanted to test if that was true in looking at the parallel v2.10 RTMA/URMA data, in which CWOP data is NOT used, but it aged off before I could view that.

Link to Google Slides.

--
Jack Settelmaier RTMA/URMA Discussion Group Virtual Lab Forum http://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/27307683VLab.Notifications@noaa.gov


--
Matthew Morris
SAIC at NOAA/NWS/NCEP/EMC
5830 University Research Ct., Rm. 2038
College Park, MD 20740
301-683-3758

--
Matthew Morris RTMA/URMA Discussion Group Virtual Lab Forum http://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/27354573VLab.Notifications@noaa.gov


--
Jack Settelmaier
Digital Techniques Meteorologist
NOAA/NWS, Southern Region HQ
Fort Worth, TX 
Work Desk: 682 703 3685
Virtual Office (most core work hours): https://meet.google.com/ujm-ajkv-rhk

--
Jack Settelmaier RTMA/URMA Discussion Group Virtual Lab Forum http://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/27361115VLab.Notifications@noaa.gov


--
Jack Settelmaier
Digital Techniques Meteorologist
NOAA/NWS, Southern Region HQ
Fort Worth, TX 
Work Desk: 682 703 3685
Virtual Office (most core work hours): https://meet.google.com/ujm-ajkv-rhk
MM
Matthew Morris, modified 2 Years ago.

RE: Too low Winds in NBM (from URMA?) over Florida Keys

Youngling Posts: 169 Join Date: 12/6/17 Recent Posts
Hi Jack,

Recall: The CONUS RTMA/URMA systems perform three outer-loops when minimizing the 2DVar cost function.

1) A difference of roughly 0.1 knots between RTMA and URMA is likely due to subtle differences in the observation usage.  For example, we perform a cross-validation technique to assess the system performance.  In this process, a subset (~10%) of stations are selected and subsequently withheld from the first two outer loops of the analysis procedure.  These observations are then reintroduced on the final outer loop.

2) At the end of each outer loop, the computed value can be considered as the intermediate analysis.  At this point, the QC is performed again, meaning that observation usage can change from rejected to accepted and vice versa.  In this case, you are correct that the HRRR-supplied background value was 10.19 kts, and the analysis lowered this slightly to 9.41 kts despite assimilating the LONF1 observation.  This can likely be attributed to the influence of nearby observations.

Thanks,
Matt

On Tue, Jan 31, 2023 at 5:54 PM VLab Notifications <VLab.Notifications@noaa.gov> wrote:
Matt,

Indeed you were correct, that I mixed Wind Speed and Gusts.  I have now corrected slides 9-11 to now compare RTMA and URMA Wind Speeds.  [I've also updated slides 12-14, but have Hid them, since they are so alike and there is no point to make].

So, the only questions I have from this review are:
1) Why did URMA slightly LOWER the 14Z Wind Speed analysis value from what the RTMA had, in the vicinity of LONF1, especially since it appears the LONF1 observation of 11.13 knots (slide 15) was not rejected.
2) Also, can you explain the lower portion of Slide 15 with regard to the iterations.  Am I reading correctly, the (HRRR-supplied) background value was 10.19, and after the LONF1 observation of 11.13 was considered, and not rejected (aka accepted) the successive iterations actually lowered the analysis value down to an eventual 9.41 knows (which matches the slides 10-11 image/values). Was the LONF1 value not actually used?

On Mon, Jan 30, 2023 at 4:17 PM VLab Notifications <VLab.Notifications@noaa.gov> wrote:
Thanks Matt, maybe I messed up.  Please do correct me, with any input you have on the values from the various systems.

Let me check on that tomorrow, assuming I still can get access to the files needed.

On Mon, Jan 30, 2023 at 11:47 AM VLab Notifications <VLab.Notifications@noaa.gov> wrote:
Hi Jack,

According to the figure labels, the RTMA figures on slides 9 and 12 are for wind gust, whereas the URMA figures on slides 10 and 13 are for wind speed.  Thus, this is not a valid comparison of the figures.  Are the values reported on slides 11 and 14 also comparing wind speed to wind gust?

Thanks,
Matt

On Fri, Jan 27, 2023 at 4:06 PM VLab Notifications <VLab.Notifications@noaa.gov> wrote:
WFO Key West shared a case with SRH STSD, describing an instance where the URMA-analyzed winds appear to be too low, when compared to available observation platforms in the nearby water and land areas.  

These too-low URMA-analyzed wind speeds appear to be similar to the too-low NBM wind speed forecasts.  If one uses the URMA analysis, vs nearby point locations, as observed data for verification purposes, resulting skill assessments favor the NBM over the NDFD.  

What appears odd in this case, is how the RTMA analysis seems to more closely match the nearby observations, but something in the URMA processes appears to have lowered the analyzed values.  Could it be the URMA made use of zero wind speed values at some CWOP sites?  URMA KML files appear to indicate the suspect zero wind speed from E9599 was rejected.
I wanted to test if that was true in looking at the parallel v2.10 RTMA/URMA data, in which CWOP data is NOT used, but it aged off before I could view that.

Link to Google Slides.

--
Jack Settelmaier RTMA/URMA Discussion Group Virtual Lab Forum http://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/27307683VLab.Notifications@noaa.gov


--
Matthew Morris
SAIC at NOAA/NWS/NCEP/EMC
5830 University Research Ct., Rm. 2038
College Park, MD 20740
301-683-3758

--
Matthew Morris RTMA/URMA Discussion Group Virtual Lab Forum http://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/27354573VLab.Notifications@noaa.gov


--
Jack Settelmaier
Digital Techniques Meteorologist
NOAA/NWS, Southern Region HQ
Fort Worth, TX 
Work Desk: 682 703 3685
Virtual Office (most core work hours): https://meet.google.com/ujm-ajkv-rhk

--
Jack Settelmaier RTMA/URMA Discussion Group Virtual Lab Forum http://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/27361115VLab.Notifications@noaa.gov


--
Jack Settelmaier
Digital Techniques Meteorologist
NOAA/NWS, Southern Region HQ
Fort Worth, TX 
Work Desk: 682 703 3685
Virtual Office (most core work hours): https://meet.google.com/ujm-ajkv-rhk

--
Jack Settelmaier RTMA/URMA Discussion Group Virtual Lab Forum http://vlab.noaa.gov/web/715073/home/-/message_boards/view_message/27383447VLab.Notifications@noaa.gov


--
Matthew Morris
SAIC at NOAA/NWS/NCEP/EMC
5830 University Research Ct., Rm. 2038
College Park, MD 20740
301-683-3758