Ensembles

[Greg Fishel]

Had anyone had any luck with displaying ensemble data in PCGRIDDS? I have tried downloading each ensemble perturbation separately, but when processing the data, only the first member is recognized and the others are ignored. I am specifically interested in being able to overlay all the ensemble members for a given parameter onto one map. If you've had any luck, let me know. Thanks!

[jkrob]

Greg,

The Ensemble GRIB files are unique in the information they contain and the PCGRIDDS32 GRIB ingestor NGRB2PCG32 has not been configured to accomidate that data. I presume it could be but I'm not that familiar with what it is that makes the ensemble files different from the others and what it would take for NGRB2PCG32 to ingest it and also what, if any, modifications to PCGRIDDS32 would be required to display that data. I could add that to the wish list.

Jeff

[ncwxman]

Greg and Jeff,

I have found a way to display all the Ensemble members at the same time. You need to change the file names after NGRB2PCG32 has processed each ensemble member.

Change FROM



TO



When you have processed all members your file list will look like this:



Now use the FIL command in your macro. Like the following:

Code: Select all

loop
plan
map
stof
LNDF
LFO1 LFD1 LMN1
LLL2 LPL2 LTM2 LLN2
fhor
PMSL FIL1 CLR4 CI10&
PMSL FIL2 CLR2 CI10&
PMSL FIL3 CLR5 CI10&
PMSL FIL4 CLR6 CI10&
PMSL FIL5 CLR7 CI10&
PMSL FIL6 CLR4 dash CI10&
PMSL FIL7 CLR2 dash CI10&
PMSL FIL8 CLR5 dash CI10&
PMSL FIL9 CLR6 dash CI10&
PMSL FILA CLR7 dash CI10&
TXT3 GFS Ensemble MSLP
txt4          Ensemble Member                 Ensemble Member  
txt5 COLOR  - SOLID - DASHED         COLOR  - SOLID - DASHED 
txt6 yellow   N1      P1             green    N4      P4
txt7 red      N2      P2             blue     N5      P5 
txt8 cyan     N3      P3
endl

The key to making this work properly is when you LOAD your files.

Using the file list above, I loaded n1 first and then n2 and so forth.

FIL1 = n1
FIL2 = n2
FIL3 = n3
FIL4 = n4
FIL5 = n5
FIL6 = p1
FIL7 = p2
FIL8 = p3
FIL9 = p4
FILA = p5

I used the above macro and file structure to create the following image of the 120 hour forecast of MSLP from the GFS Ensembles.



Greg, I hope this helps and if you need any further assistance please let me know. I would be more than happy to help you.

Kevin McBride

[jkrob]

Kevin,

Way too cool, man...you rock, dude!!! (Does that sound odd coming from a 42 yr old??)

Anyway, the way you modified the file names looks like the way to go. I could look into modifying NGRB2PCG32 so it would add that 'tag' to the file name when it came across an ensemble GRIB file. BTW - what is the difference between n1 & p1? I believe there are other ensemble GRIB products but, like I said, I just haven't looked into them. Maybe I should.

Also, for the display, you could use the SLIN command to create true 'spaghetti' plots like NCEP does.

Again, GREAT job, Kevin!!

Jeff

[ncwxman]

Jeff,

Way too cool, man...you rock, dude!!! (Does that sound odd coming from a 42 yr old??)

Not if your 40!

Also, for the display, you could use the SLIN command to create true 'spaghetti' plots like NCEP does.

Ya, I thought about using the SLIN command last night when I was working on this but, I was a bit tired by the time I thought of it. That would likely work better.

These are 2 of the ensemble members. I'm not sure of all the technical differences between the ensemble members. I have read that it deals with how they handle the initial conditions and that effects the forecast. There are other differences to. Go to http://meted.ucar.edu/nwp/pcu1/ensemble/frameset.htm for more info.

Kevin

[ncwxman]

All,

Here is a spaghetti plot that I created.



I created the following macro to produce this plot.

Code: Select all

loop
emap
LNDF
LFO1 LFD1 LMN1
LLL2 LPL2 LTM2 LLN2
PMSL&SLIN 0992 FIL1 CLR5 dash
PMSL&SLIN 0992 FIL2 CLR5 dash&
PMSL&SLIN 0992 FIL3 CLR5 dash&
PMSL&SLIN 0992 FIL4 CLR5 dash&
PMSL&SLIN 0992 FIL5 CLR5 dash&
PMSL&SLIN 0992 FIL6 CLR5&
PMSL&SLIN 0992 FIL7 CLR5&
PMSL&SLIN 0992 FIL8 CLR5&
PMSL&SLIN 0992 FIL9 CLR5&
PMSL&SLIN 0992 FILA CLR5&
PMSL&SLIN 1008 FIL1 CLR4 dash
PMSL&SLIN 1008 FIL2 CLR4 dash&
PMSL&SLIN 1008 FIL3 CLR4 dash&
PMSL&SLIN 1008 FIL4 CLR4 dash&
PMSL&SLIN 1008 FIL5 CLR4 dash&
PMSL&SLIN 1008 FIL6 CLR4&
PMSL&SLIN 1008 FIL7 CLR4&
PMSL&SLIN 1008 FIL8 CLR4&
PMSL&SLIN 1008 FIL9 CLR4&
PMSL&SLIN 1008 FILA CLR4&
PMSL&SLIN 1020 FIL1 CLR2 dash
PMSL&SLIN 1020 FIL2 CLR2 dash&
PMSL&SLIN 1020 FIL3 CLR2 dash&
PMSL&SLIN 1020 FIL4 CLR2 dash&
PMSL&SLIN 1020 FIL5 CLR2 dash&
PMSL&SLIN 1020 FIL6 CLR2&
PMSL&SLIN 1020 FIL7 CLR2&
PMSL&SLIN 1020 FIL8 CLR2&
PMSL&SLIN 1020 FIL9 CLR2&
PMSL&SLIN 1020 FILA CLR2&
TXT3 GFS Ensemble MSLP
txt4 dashed = "N" Ensemble Members / solid = "P" Ensemble Members 
txt5 cyan = 992 mb, yellow = 1008 mb, red = 1020 mb
endl

Kevin

[jkrob]

Kevin & Greg,

Real quick...which FTP server & dirctory string do you get your Ensemble GRIB files from?

Jeff

[jkrob]

All,

Continuing this thread, here is thw Web page describing the Ensemble data. I just thought I'd pass this along to let you know what you were gettig into and all the processing involved to do it right...

http://www.emc.ncep.noaa.gov/gmb/ens/info/ens_grib.html

There is more there than what meets the eye. In other words it's more involved than just "ensemble data"...if you know what I mean.

BTW -
n1 = first negative perturbation
n2 = second negative perturbation
...
etc.

p1 = first positive perturbation
p2 = second positive perturbation
...
etc.

Thanks,
Jeff

[ncwxman]

Jeff,

I use the following server :
http://nomad5.ncep.noaa.gov/ncep_data/
It allows you to select a sub region of the total grid. This cuts file size down quite a bit and this speeds up the download if you don't need the entire grid. I use the the hi res.

Thanks for the additional info on the ensembles!

Kevin