Page 1 of 1

Help me understand Ensemble Member MSLP graphics

Posted: Thu Aug 27, 2020 3:30 pm
by Spacecoast
Dumb question from amateur..
Looking at Tropical Tidbits Ensemble Member MSLP maps, could someone explain what the red (or blue) numbers signify?
I tried searching for a quick tutorial on how to read / interprete these ensemble maps, with no luck.
Also, the color intensities, that range fom blue(0.0) to orange (2.6)... does that reflect normalized spread?
What should I make of a deeper orange coloration vs no colors for and individual member?

I apologize for asking, can anyone point me to a site that explains this stuff?

Re: Help me understand Ensemble Member MSLP graphics

Posted: Thu Aug 27, 2020 3:35 pm
by RL3AO
The numbers represent the sea level pressures and locations of the individual ensemble members. The shading is the normalized spread of the mean sea level pressures. Basically telling you where the biggest spread of pressure is.

Re: Help me understand Ensemble Member MSLP graphics

Posted: Thu Aug 27, 2020 3:43 pm
by Extratropical94
The normalized spread also takes into account the previous runs of the forecast system. For instance, the Euro uses the 30 most recent 00Z or 12Z runs to determine a mean standard deviation for every location and forecast lead time, which is then used to normalize the spread of the current ensemble run.

You might want to have a look at https://confluence.ecmwf.int/display/FU ... and+Spread for more detailed explanations of ensemble products.

Also, there are no dumb questions. :wink:

Re: Help me understand Ensemble Member MSLP graphics

Posted: Thu Aug 27, 2020 3:49 pm
by Spacecoast
Thank you for your response. Still a little confused.

The red / blue numbers seem to range from 0 to 90's, so how does that relate to mean MSLP? I was thinking these numbers indentified the particular ensemble member.
Shouldn't MSLP in mb be three or four digit values in the range of 800 - 1020?

Reagarding the normalized spread shading, is that he same as normalized standard deviation? In other words small values (or paler coloration) on the shaded fields mean high confidence, large values (brighter oranges) mean low confidence?

Re: Help me understand Ensemble Member MSLP graphics

Posted: Thu Aug 27, 2020 3:57 pm
by Extratropical94
Spacecoast wrote:Thank you for your response. Still a little confused.

The red / blue numbers seem to range from 0 to 90's, so how does that relate to mean MSLP? I was thinking these numbers indentified the particular ensemble member.
Shouldn't MSLP in mb be three or four digit values in the range of 800 - 1020?

Reagarding the normalized spread shading, is that he same as normalized standard deviation? In other words small values (or paler coloration) on the shaded fields mean high confidence, large values (brighter oranges) mean low confidence?


The numbers are shortened so they don't take up that much space. 75 means 975, 21 means 1021 (or 921). So if you see a blue 1021, it's gonna be a 1021 mb high, whereas a red 21 is a 921 mb low.
Yes, the normalized spread is the normalized standard deviation, with the past 30 runs acting as the normalization reference period. And you are correct, small values indicate a high confidence (= a tightly clustered ensemble). Basically, values below 1 show that the standard deviation is lower for this forecast time and region than in the past 30 runs. Values above 1 mean higher spreads then usual. A higher spread can exist for two reasons, either low confidence in the position of a low/high, or low confidence in its pressure (or both).

Re: Help me understand Ensemble Member MSLP graphics

Posted: Thu Aug 27, 2020 4:00 pm
by Spacecoast
Extratropical94 wrote:The numbers are shortened so they don't take up that much space. 75 means 975, 21 means 1021 (or 921). So if you see a blue 1021, it's gonna be a 1021 mb high, whereas a red 21 is a 921 mb low.
Yes, the normalized spread is the normalized standard deviation, with the past 30 runs acting as the normalization reference period. And you are correct, small values indicate a high confidence (= a tightly clustered ensemble). Basically, values below 1 show that the standard deviation is lower for this forecast time and region than in the past 30 runs. Values above 1 mean higher spreads then usual. A higher spread can exist for two reasons, either low confidence in the position of a low/high, or low confidence in its pressure (or both).


Bingo! Thank you very much. Exactly the explanation I was looking for.
This helps me immensely.