Forums

Back

RE: Map View and 1-D Viewer Not Matching Probabilities

RSS (Opens New Window)(Opens New Window)
TG
Thomas Grafenauer, modified 3 Years ago.

Map View and 1-D Viewer Not Matching Probabilities

Youngling Posts: 5 Join Date: 2/20/14 Recent Posts

Noticed that the probability from the 1-D viewer is not matching the probability from the map viewer within NBM4.1.   For example, I've attached images from the Jan 30 06z Init.  Notice that for Devils Lake, ND, Jamestown, ND, and Grand Forks, ND, the probability on the map viewer to exceed 48 knots is 44%, 62%, and 23% respectively.   The probability from the 1-D viewer is  33%, 85%, and 52% respectively.  https://drive.google.com/file/d/1UPWaE4Ym1LSuIr74cQfTARu7fzdbyytA/view?usp=sharing, https://drive.google.com/file/d/1W7pXqZTRa7EcrGFqwhlsUOuCVUSTRD8e/view?usp=sharing, https://drive.google.com/file/d/1dk1DFBzomMGDP-wROZffFTUzhGzETvQf/view?usp=sharing, https://drive.google.com/file/d/1txb_8T8XZP0PaBeaJ-YH8yGRc231D2TD/view?usp=sharing

DS
Dana Strom, modified 3 Years ago.

RE: Map View and 1-D Viewer Not Matching Probabilities

Youngling Posts: 87 Join Date: 12/8/14 Recent Posts
Hi Thomas,

The probabilities calculated within the 1D section are generated via a different set of calculations and processes as compared to the 24 hr Max Wind Gust element. I'm afraid you're trying to compare apples and oranges.

The graphs you show utilizes the hourly wind gust data and the standard deviations to come up with those probabilities. The 24 hr Max Wind Gust is an independent element that likely both uses different input and has a different calculation in the NBM (ask the NBM team if you're interested). 

The 1D page does contain the plot for the 24 hr Max Wind Gust probabilities that you can use to view the data. Below are the Plan view and corresponding probability chart for Jamestown ND.

image.png

image.png

I hope that clears up the confusion,
Dana

On Sun, Jan 30, 2022 at 11:16 AM Thomas Grafenauer <VLab.Notifications@noaa.gov> wrote:

Noticed that the probability from the 1-D viewer is not matching the probability from the map viewer within NBM4.1.   For example, I've attached images from the Jan 30 06z Init.  Notice that for Devils Lake, ND, Jamestown, ND, and Grand Forks, ND, the probability on the map viewer to exceed 48 knots is 44%, 62%, and 23% respectively.   The probability from the 1-D viewer is  33%, 85%, and 52% respectively.  https://drive.google.com/file/d/1UPWaE4Ym1LSuIr74cQfTARu7fzdbyytA/view?usp=sharing, https://drive.google.com/file/d/1W7pXqZTRa7EcrGFqwhlsUOuCVUSTRD8e/view?usp=sharing, https://drive.google.com/file/d/1dk1DFBzomMGDP-wROZffFTUzhGzETvQf/view?usp=sharing, https://drive.google.com/file/d/1txb_8T8XZP0PaBeaJ-YH8yGRc231D2TD/view?usp=sharing


--
Thomas Grafenauer Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum https://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20174485VLab.Notifications@noaa.gov


--
Dana Strom
Visualization Task Lead
Digital Forecast Services Division
Meteorological Development Lab
NOAA/National Weather Service

301-427-9451

Attachments:

Attachment

Attachment

TG
Thomas Grafenauer, modified 3 Years ago.

RE: Map View and 1-D Viewer Not Matching Probabilities

Youngling Posts: 5 Join Date: 2/20/14 Recent Posts
That makes sense, thanks for clearing that up!  

Is it possible to replace the SD calculation with the 24 hour max wind and 24 hour max wind gust calculations so that the output is consistent?

-tommy 

Tommy Grafenauer
Science and Operations Officer
National Weather Service - Grand Forks, ND
701-772-0720 x766



On Mon, Jan 31, 2022 at 8:40 AM Dana Strom <VLab.Notifications@noaa.gov> wrote:

Hi Thomas,

The probabilities calculated within the 1D section are generated via a different set of calculations and processes as compared to the 24 hr Max Wind Gust element. I'm afraid you're trying to compare apples and oranges.

The graphs you show utilizes the hourly wind gust data and the standard deviations to come up with those probabilities. The 24 hr Max Wind Gust is an independent element that likely both uses different input and has a different calculation in the NBM (ask the NBM team if you're interested). 

The 1D page does contain the plot for the 24 hr Max Wind Gust probabilities that you can use to view the data. Below are the Plan view and corresponding probability chart for Jamestown ND.

image.png

image.png

I hope that clears up the confusion,
Dana

On Sun, Jan 30, 2022 at 11:16 AM Thomas Grafenauer <VLab.Notifications@noaa.gov> wrote:

Noticed that the probability from the 1-D viewer is not matching the probability from the map viewer within NBM4.1.   For example, I've attached images from the Jan 30 06z Init.  Notice that for Devils Lake, ND, Jamestown, ND, and Grand Forks, ND, the probability on the map viewer to exceed 48 knots is 44%, 62%, and 23% respectively.   The probability from the 1-D viewer is  33%, 85%, and 52% respectively.  https://drive.google.com/file/d/1UPWaE4Ym1LSuIr74cQfTARu7fzdbyytA/view?usp=sharing, https://drive.google.com/file/d/1W7pXqZTRa7EcrGFqwhlsUOuCVUSTRD8e/view?usp=sharing, https://drive.google.com/file/d/1dk1DFBzomMGDP-wROZffFTUzhGzETvQf/view?usp=sharing, https://drive.google.com/file/d/1txb_8T8XZP0PaBeaJ-YH8yGRc231D2TD/view?usp=sharing


--
Thomas Grafenauer Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum https://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20174485VLab.Notifications@noaa.gov


--
Dana Strom
Visualization Task Lead
Digital Forecast Services Division
Meteorological Development Lab
NOAA/National Weather Service

301-427-9451


--
Dana Strom Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum http://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20204117 VLab.Notifications@noaa.gov

DS
Dana Strom, modified 3 Years ago.

RE: Map View and 1-D Viewer Not Matching Probabilities

Youngling Posts: 87 Join Date: 12/8/14 Recent Posts
Unfortunately not, The calculations on that plot are done by the interface, whereas the 24 hour max wind gust calculations are done internally within the NBM that we display.

If you're looking to see the probability of exceedances, you're going to want to use the chart where we display the data straight from the NBM. That's going to have more accurate information if you're comparing the two.

The plots that use the SD/probability curves (where available) to calculate probability of exceedances should only be used in absence of a direct exceedance product from the NBM. 

On Mon, Jan 31, 2022 at 10:10 AM Thomas Grafenauer <VLab.Notifications@noaa.gov> wrote:

That makes sense, thanks for clearing that up!  

Is it possible to replace the SD calculation with the 24 hour max wind and 24 hour max wind gust calculations so that the output is consistent?

-tommy 

Tommy Grafenauer
Science and Operations Officer
National Weather Service - Grand Forks, ND
701-772-0720 x766



On Mon, Jan 31, 2022 at 8:40 AM Dana Strom <VLab.Notifications@noaa.gov> wrote:

Hi Thomas,

The probabilities calculated within the 1D section are generated via a different set of calculations and processes as compared to the 24 hr Max Wind Gust element. I'm afraid you're trying to compare apples and oranges.

The graphs you show utilizes the hourly wind gust data and the standard deviations to come up with those probabilities. The 24 hr Max Wind Gust is an independent element that likely both uses different input and has a different calculation in the NBM (ask the NBM team if you're interested). 

The 1D page does contain the plot for the 24 hr Max Wind Gust probabilities that you can use to view the data. Below are the Plan view and corresponding probability chart for Jamestown ND.

image.png

image.png

I hope that clears up the confusion,
Dana

On Sun, Jan 30, 2022 at 11:16 AM Thomas Grafenauer <VLab.Notifications@noaa.gov> wrote:

Noticed that the probability from the 1-D viewer is not matching the probability from the map viewer within NBM4.1.   For example, I've attached images from the Jan 30 06z Init.  Notice that for Devils Lake, ND, Jamestown, ND, and Grand Forks, ND, the probability on the map viewer to exceed 48 knots is 44%, 62%, and 23% respectively.   The probability from the 1-D viewer is  33%, 85%, and 52% respectively.  https://drive.google.com/file/d/1UPWaE4Ym1LSuIr74cQfTARu7fzdbyytA/view?usp=sharing, https://drive.google.com/file/d/1W7pXqZTRa7EcrGFqwhlsUOuCVUSTRD8e/view?usp=sharing, https://drive.google.com/file/d/1dk1DFBzomMGDP-wROZffFTUzhGzETvQf/view?usp=sharing, https://drive.google.com/file/d/1txb_8T8XZP0PaBeaJ-YH8yGRc231D2TD/view?usp=sharing


--
Thomas Grafenauer Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum https://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20174485VLab.Notifications@noaa.gov


--
Dana Strom
Visualization Task Lead
Digital Forecast Services Division
Meteorological Development Lab
NOAA/National Weather Service

301-427-9451


--
Dana Strom Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum http://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20204117 VLab.Notifications@noaa.gov


--
Thomas Grafenauer Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum http://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20207743 VLab.Notifications@noaa.gov



--
Dana Strom
Visualization Task Lead
Digital Forecast Services Division
Meteorological Development Lab
NOAA/National Weather Service

301-427-9451

TG
Thomas Grafenauer, modified 3 Years ago.

RE: Map View and 1-D Viewer Not Matching Probabilities

Youngling Posts: 5 Join Date: 2/20/14 Recent Posts
Gotcha....good to know.   So, that leads me to one last question....is this true of other parameters like QPF, snow, etc.?

-tommy

Tommy Grafenauer
Science and Operations Officer
National Weather Service - Grand Forks, ND
701-772-0720 x766



On Tue, Feb 1, 2022 at 6:51 AM Dana Strom <VLab.Notifications@noaa.gov> wrote:

Unfortunately not, The calculations on that plot are done by the interface, whereas the 24 hour max wind gust calculations are done internally within the NBM that we display.

If you're looking to see the probability of exceedances, you're going to want to use the chart where we display the data straight from the NBM. That's going to have more accurate information if you're comparing the two.

The plots that use the SD/probability curves (where available) to calculate probability of exceedances should only be used in absence of a direct exceedance product from the NBM. 

On Mon, Jan 31, 2022 at 10:10 AM Thomas Grafenauer <VLab.Notifications@noaa.gov> wrote:

That makes sense, thanks for clearing that up!  

Is it possible to replace the SD calculation with the 24 hour max wind and 24 hour max wind gust calculations so that the output is consistent?

-tommy 

Tommy Grafenauer
Science and Operations Officer
National Weather Service - Grand Forks, ND
701-772-0720 x766



On Mon, Jan 31, 2022 at 8:40 AM Dana Strom <VLab.Notifications@noaa.gov> wrote:

Hi Thomas,

The probabilities calculated within the 1D section are generated via a different set of calculations and processes as compared to the 24 hr Max Wind Gust element. I'm afraid you're trying to compare apples and oranges.

The graphs you show utilizes the hourly wind gust data and the standard deviations to come up with those probabilities. The 24 hr Max Wind Gust is an independent element that likely both uses different input and has a different calculation in the NBM (ask the NBM team if you're interested). 

The 1D page does contain the plot for the 24 hr Max Wind Gust probabilities that you can use to view the data. Below are the Plan view and corresponding probability chart for Jamestown ND.

image.png

image.png

I hope that clears up the confusion,
Dana

On Sun, Jan 30, 2022 at 11:16 AM Thomas Grafenauer <VLab.Notifications@noaa.gov> wrote:

Noticed that the probability from the 1-D viewer is not matching the probability from the map viewer within NBM4.1.   For example, I've attached images from the Jan 30 06z Init.  Notice that for Devils Lake, ND, Jamestown, ND, and Grand Forks, ND, the probability on the map viewer to exceed 48 knots is 44%, 62%, and 23% respectively.   The probability from the 1-D viewer is  33%, 85%, and 52% respectively.  https://drive.google.com/file/d/1UPWaE4Ym1LSuIr74cQfTARu7fzdbyytA/view?usp=sharing, https://drive.google.com/file/d/1W7pXqZTRa7EcrGFqwhlsUOuCVUSTRD8e/view?usp=sharing, https://drive.google.com/file/d/1dk1DFBzomMGDP-wROZffFTUzhGzETvQf/view?usp=sharing, https://drive.google.com/file/d/1txb_8T8XZP0PaBeaJ-YH8yGRc231D2TD/view?usp=sharing


--
Thomas Grafenauer Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum https://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20174485VLab.Notifications@noaa.gov


--
Dana Strom
Visualization Task Lead
Digital Forecast Services Division
Meteorological Development Lab
NOAA/National Weather Service

301-427-9451


--
Dana Strom Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum http://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20204117 VLab.Notifications@noaa.gov


--
Thomas Grafenauer Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum http://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20207743 VLab.Notifications@noaa.gov



--
Dana Strom
Visualization Task Lead
Digital Forecast Services Division
Meteorological Development Lab
NOAA/National Weather Service

301-427-9451


--
Dana Strom Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum http://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20221423 VLab.Notifications@noaa.gov

DS
Dana Strom, modified 3 Years ago.

RE: Map View and 1-D Viewer Not Matching Probabilities

Youngling Posts: 87 Join Date: 12/8/14 Recent Posts
Yes. If they are available for the element/threshold you're interested in, you should use the direct output. However, let me be more specific:

For wind speed and gust, you can see the 10m data and std dev (dotted line and yellow spread). 

image.png

For the Tmax data, we are able to supply it the percentiles (box/whiskers) at which point the calculation has a lot more data to do those calculations for.

image.png

You should see better agreement with the calculations done for these elements than ones with only the standard deviations supplied. Remember thorough, exact agreement is not expected.

We supply the percentiles for every element but wind speed and gust. This is due to wind speed/gust in this plot being the hourly product with no percentiles to go along with it. 



On Tue, Feb 1, 2022 at 8:02 AM Thomas Grafenauer <VLab.Notifications@noaa.gov> wrote:
Gotcha....good to know.   So, that leads me to one last question....is this true of other parameters like QPF, snow, etc.?

-tommy

Tommy Grafenauer
Science and Operations Officer
National Weather Service - Grand Forks, ND
701-772-0720 x766



On Tue, Feb 1, 2022 at 6:51 AM Dana Strom <VLab.Notifications@noaa.gov> wrote:
Unfortunately not, The calculations on that plot are done by the interface, whereas the 24 hour max wind gust calculations are done internally within the NBM that we display.

If you're looking to see the probability of exceedances, you're going to want to use the chart where we display the data straight from the NBM. That's going to have more accurate information if you're comparing the two.

The plots that use the SD/probability curves (where available) to calculate probability of exceedances should only be used in absence of a direct exceedance product from the NBM. 

On Mon, Jan 31, 2022 at 10:10 AM Thomas Grafenauer <VLab.Notifications@noaa.gov> wrote:
That makes sense, thanks for clearing that up!  

Is it possible to replace the SD calculation with the 24 hour max wind and 24 hour max wind gust calculations so that the output is consistent?

-tommy 

Tommy Grafenauer
Science and Operations Officer
National Weather Service - Grand Forks, ND
701-772-0720 x766



On Mon, Jan 31, 2022 at 8:40 AM Dana Strom <VLab.Notifications@noaa.gov> wrote:
Hi Thomas,

The probabilities calculated within the 1D section are generated via a different set of calculations and processes as compared to the 24 hr Max Wind Gust element. I'm afraid you're trying to compare apples and oranges.

The graphs you show utilizes the hourly wind gust data and the standard deviations to come up with those probabilities. The 24 hr Max Wind Gust is an independent element that likely both uses different input and has a different calculation in the NBM (ask the NBM team if you're interested). 

The 1D page does contain the plot for the 24 hr Max Wind Gust probabilities that you can use to view the data. Below are the Plan view and corresponding probability chart for Jamestown ND.

image.png

image.png

I hope that clears up the confusion,
Dana

On Sun, Jan 30, 2022 at 11:16 AM Thomas Grafenauer <VLab.Notifications@noaa.gov> wrote:

Noticed that the probability from the 1-D viewer is not matching the probability from the map viewer within NBM4.1.   For example, I've attached images from the Jan 30 06z Init.  Notice that for Devils Lake, ND, Jamestown, ND, and Grand Forks, ND, the probability on the map viewer to exceed 48 knots is 44%, 62%, and 23% respectively.   The probability from the 1-D viewer is  33%, 85%, and 52% respectively.  https://drive.google.com/file/d/1UPWaE4Ym1LSuIr74cQfTARu7fzdbyytA/view?usp=sharing, https://drive.google.com/file/d/1W7pXqZTRa7EcrGFqwhlsUOuCVUSTRD8e/view?usp=sharing, https://drive.google.com/file/d/1dk1DFBzomMGDP-wROZffFTUzhGzETvQf/view?usp=sharing, https://drive.google.com/file/d/1txb_8T8XZP0PaBeaJ-YH8yGRc231D2TD/view?usp=sharing


--
Thomas Grafenauer Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum https://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20174485VLab.Notifications@noaa.gov


--
Dana Strom
Visualization Task Lead
Digital Forecast Services Division
Meteorological Development Lab
NOAA/National Weather Service

301-427-9451


--
Dana Strom Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum http://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20204117VLab.Notifications@noaa.gov

--
Thomas Grafenauer Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum http://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20207743VLab.Notifications@noaa.gov


--
Dana Strom
Visualization Task Lead
Digital Forecast Services Division
Meteorological Development Lab
NOAA/National Weather Service

301-427-9451


--
Dana Strom Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum http://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20221423VLab.Notifications@noaa.gov

--
Thomas Grafenauer Whole Story Uncertainty & Probabilities Viewer Virtual Lab Forum http://vlab.noaa.gov/web/wsup/forums/-/message_boards/view_message/20221445VLab.Notifications@noaa.gov


--
Dana Strom
Visualization Task Lead
Digital Forecast Services Division
Meteorological Development Lab
NOAA/National Weather Service

301-427-9451

Attachments:

Attachment

Attachment