Jump to content

Photo

Mid Atlantic Met Class Thread


86 replies to this topic

#71
das

  • 922 posts
  • Joined November 12, 2010

I don't have much to add other than repeating how good this thread is, and thank you for the information.

Coming from a software engineering background, I am curious what programming language most models are developed in? Also, what kind of computers are they using to process all of the algorithms?


The computing systems are really a system of systems with Supercomputers doing a lot of the heavy lifting and mainframes and client/server stuff taking a significant amount of local load. Here's a high-level diagram of the overall architecture, with some stuff redacted out:
EOS App Arch.png

Here's a slide deck that details most of the infrastructure and its tie in to numerical weather prediction modelling. There are some interesting computing systems links on slide 68.
https://skydrive.liv...B544483E3%21258

The local office stuff is typical of a remote or regional office. Local, high-res models with smaller domains are run at these sites so that 1) the main computing systems are not impeded and 2) it offers the local offices more flexibility in ad-hoc runs for significant or hyper-local forecasting needs. Here's a pic of our local office:
http://www.erh.noaa....r/computers.htm

Here's a draft roadmap for the NextGen exascale computing systems that will likely power all sorts of government-facing numerical and graphical predictive analysis systems, including those in NOAA, going forward. I only have the draft on this workstation. Ping me via PM if you want the latest and I'll dig it up when I can.
https://skydrive.liv...B544483E3%21257

#72
WxUSAF

  • 9,841 posts
  • Joined November 11, 2010

Speaking of CAPE and following up about the skew-T discussion, the amount of CAPE can be directly calculated from the skew-T by adding up the area between the parcel trajectory (along a moist adiabat) and the environmental temperature profile. So, if there's a lot of room between the two, you can know the CAPE is high before looking at the calculated parameters directly.

The "opposite" of CAPE is CIN (convective inhibition) and is also measured in J/kg. CIN occurs when the parcel trajectory will lead it to be colder than the environmental temperature. If the parcel is colder, it's not buoyant in the free atmosphere and needs to be forced to a height in the atmosphere where it is bouyant before it can freely begin to rise. When people speak of a "cap", there means the atmospheric profile has a certain amount of CIN to overcome.

#73
mapgirl

  • 18,270 posts
  • Joined November 12, 2010

Speaking of CAPE and following up about the skew-T discussion, the amount of CAPE can be directly calculated from the skew-T by adding up the area between the parcel trajectory (along a moist adiabat) and the environmental temperature profile. So, if there's a lot of room between the two, you can know the CAPE is high before looking at the calculated parameters directly.

The "opposite" of CAPE is CIN (convective inhibition) and is also measured in J/kg. CIN occurs when the parcel trajectory will lead it to be colder than the environmental temperature. If the parcel is colder, it's not buoyant in the free atmosphere and needs to be forced to a height in the atmosphere where it is bouyant before it can freely begin to rise. When people speak of a "cap", there means the atmospheric profile has a certain amount of CIN to overcome.


The easiest way for me to understand cap was to think of the atmosphere as a pot with a lid on it. The lid being the cap that needs to come off before any storms can develop and grow in size.

#74
Bob Chill

  • Onin Le Ikimod

  • 15,352 posts
  • Joined November 29, 2010

Speaking of CAPE and following up about the skew-T discussion, the amount of CAPE can be directly calculated from the skew-T by adding up the area between the parcel trajectory (along a moist adiabat) and the environmental temperature profile. So, if there's a lot of room between the two, you can know the CAPE is high before looking at the calculated parameters directly.

The "opposite" of CAPE is CIN (convective inhibition) and is also measured in J/kg. CIN occurs when the parcel trajectory will lead it to be colder than the environmental temperature. If the parcel is colder, it's not buoyant in the free atmosphere and needs to be forced to a height in the atmosphere where it is bouyant before it can freely begin to rise. When people speak of a "cap", there means the atmospheric profile has a certain amount of CIN to overcome.


Very good explanation. I found this skew diagram while reading through the links that trix posted. I don't think there could be a better visual to go along with your explanation.




#75
ohleary

  • Basho

  • 4,683 posts
  • Joined December 15, 2010

CAPE is an acronym for Convective Available Potential Energy, and is used as a metric when identifying the potential for convection (thunderstorm activity). It must be stressed that CAPE is the end all for determining severe weather risk, but is certainly important. Remember that air must have buoyancy (ability to rise) and then condense to form clouds, rain, hail, etc and CAPE provides a measure of this buoyancy. The higher this buoyancy, the quicker air will rise and begin the formation of thunderstorms. The CAPE value is measured in Joules per kilogram (J/Kg) and, generally, you need a minimum of 1,000j/kg of CAPE to see severe weather (Blanchard, 1998). CAPE can reach values of up to 5,000J/Kg, but that is reserved for rare events, such as large scale severe weather outbreaks. Here in the Mid Atlantic a good day is when CAPE gets to ~2,000J/Kg. There are two other types of CAPE talked about in forecast discussions and severe weather outlooks: MLCAPE (Mixed Layer CAPE) and SBCAPE (Surface-Based CAPE). Unfortunately I am not adept at explaining these two terms, but if you were to Google: SPC + MLCAPE or SPC + SBCAPE there are several good snippets of information. Hope this helps!


Reference:

Blanchard, D. O. (1998). Assessing the Vertical distribution of Convective Available Potential Energy. American Meteorological Society, 13, 870 - 877.:


MLCAPE and SBCAPE are just two different ways of initially lifting the parcel. With MLCAPE you average the temp and dewpoint in the lowest 100mb and lift from where they intersect. With SBCAPE, you lift from the surface temp and dewpoint to get your parcel trajectory. For example, in Bob's post above, the parcel looks to be SBCAPE where the traces start at the surface temp/dewpoint. If you try lifting from the average of the lowest 100mb temp/dewpoint, it looks like you would get slightly less CAPE.

#76
Bob Chill

  • Onin Le Ikimod

  • 15,352 posts
  • Joined November 29, 2010

Ellinwood and other severe guys, could you provide a little tutorial on what you guys look for irt t-storms and severe? I have a handle on some of the basics but I thought it was pretty cool that you identified today's threat pretty easily in advance. I'm interested in which charts and models you use.

Thanks!

#77
Ellinwood

  • storm snob

  • 5,598 posts
  • Joined November 12, 2010

Ellinwood and other severe guys, could you provide a little tutorial on what you guys look for irt t-storms and severe? I have a handle on some of the basics but I thought it was pretty cool that you identified today's threat pretty easily in advance. I'm interested in which charts and models you use.

Thanks!


Soundings, available moisture, dynamics and jet streaks are the most important things to look at... I'll get something drawn up this weekend but for now it's chase day!

#78
Bob Chill

  • Onin Le Ikimod

  • 15,352 posts
  • Joined November 29, 2010

Soundings, available moisture, dynamics and jet streaks are the most important things to look at... I'll get something drawn up this weekend but for now it's chase day!


Excellent. No rush of course. Thread won't be going anywhere and I'm looking forward to expanding my brain this year. Winter analysis is getting kinda boring in these parts. lol

Bring your camera and don't pull a Dorothy and Oz on us!

#79
sbos_wx

  • Jay

  • 13,919 posts
  • Joined November 12, 2010

Sometimes i think the things that inhibit severe can be better to know; dry air, cap, stable marine LL BL, too much shear, lack of mid level lapse rates, no trigger/lift, no cold pool aloft, bad timing, no curved hodo signature, no favorable jet... some obviously already mentioned

#80
ohleary

  • Basho

  • 4,683 posts
  • Joined December 15, 2010

Sometimes i think the things that inhibit severe can be better to know; dry air, cap, stable marine LL BL, too much shear, lack of mid level lapse rates, no trigger/lift, no cold pool aloft, bad timing, no curved hodo signature, no favorable jet... some obviously already mentioned


Not sure "better" is accurate. Alot of these are just the opposite of features you assess for severe risk...so if you learn what causes severe wx, knowing these is a no-brainer.

We are taught to assess severe parameters and look for severe where they overlap, like on a severe composite map. Then once the threat area is outlined, you assess the features that would hinder it. At least, that's what I'd do. :)

For example, if you look at severe parameters and find an area where they all overlap, there likely won't be a stable marine layer/lack of ML lapse rates/lack of shear/lack of trigger/no favorable jet/lack of hodo sig/etc there in the first place. You initially found an environment where the opposite of all of these are present. Make sense?

#81
Fozz

  • 15,471 posts
  • Joined November 14, 2010

I found this link (http://www.wundergro.../radar/help.asp) which nicely explains how radars work, but I do have a noob question:

I sometimes see people talking about level 3 or level 2 radars, which I assume refers to NEXRAD's data, but what is the difference between those "levels"? From what I've seen, people talk about level 3 data as if it's inferior, but why?

#82
NOVAForecaster

  • 547 posts
  • Joined February 15, 2012

I found this link (http://www.wundergro.../radar/help.asp) which nicely explains how radars work, but I do have a noob question:

I sometimes see people talking about level 3 or level 2 radars, which I assume refers to NEXRAD's data, but what is the difference between those "levels"? From what I've seen, people talk about level 3 data as if it's inferior, but why?


http://www.grlevelx.com/ this is what is being referred to by the "levels" of radar...

#83
dtk

  • 739 posts
  • Joined November 23, 2010

http://www.grlevelx.com/ this is what is being referred to by the "levels" of radar...


In many data sets, particularly for observations, the "level" usually is in reference to the amount of processing that has been done to the raw measurement. For NEXRAD, you can find descriptions for
Level2 Here: http://www.roc.noaa....vel2Info.aspx
and
Level3 Here: http://www.roc.noaa....Level3Info.aspx

#84
thunderman

  • 664 posts
  • Joined November 29, 2010

Great resource for comparing soundings and overall synoptic patterns to verified events.

http://www.spc.noaa....archive/events/

#85
thunderman

  • 664 posts
  • Joined November 29, 2010

Mesoscale book that I would highly recommend. Served me well at Millersville and has a fair amount of pictures/diagrams if you are a more visual learner. It doesn't really make a hard effort to avoid the mathematics, but it also tries to break things down conceptually so if you have the interest you will likely enjoy it.

http://www.amazon.co...r/dp/0470742135


BTW, this a GREAT thread!

#86
Bob Chill

  • Onin Le Ikimod

  • 15,352 posts
  • Joined November 29, 2010

This is probably a good thread to post all our favorite weather links. I'm terrible about organizing my own. I almost feel like paying someone to send me an organized bookmark folder with all the good ones.

Here's a few I really like:

Making composites for analog years:

http://www.esrl.noaa...data/getpage.pl

Major teleconnection forecasts and historical data:

http://www.cpc.ncep....nnections.shtml

One Stop Enso:

Nino: http://www.elnino.noaa.gov/index.html
Nina: http://www.elnino.noaa.gov/lanina.html
Historical Data: http://www.cpc.ncep....ensoyears.shtml

MJO ensemble forecast (GFS):

http://www.cpc.ncep....O/foregfs.shtml

Aweome climate data site (especially for historical location snowfall):

http://climate.usurf...oducts/data.php

Favorite MD radar (great resolution with temp overlay):

http://www.marylandw...klwxstatebr.php

Favorite Model Links:

GFS/NAM/GEFS/RUC/SREF: http://mag.ncep.noaa...=MODEL GUIDANCE

Euro: http://www.wundergro...i=0&ls=0&rad2=0




There are 100's of other great links out there. Please share your faves. I need to get my stuff organized and I'm always looking for the best links out there. I'm sure there are many other board members who would like to add to their list as well.

#87
LocoAko

  • Jake

  • 4,358 posts
  • Joined November 11, 2010

Mesoscale book that I would highly recommend. Served me well at Millersville and has a fair amount of pictures/diagrams if you are a more visual learner. It doesn't really make a hard effort to avoid the mathematics, but it also tries to break things down conceptually so if you have the interest you will likely enjoy it.

http://www.amazon.co...r/dp/0470742135


BTW, this a GREAT thread!


Seconded suggestion for this book! Great, comprehensive resource for all things mesoscale. :)




Reply to this topic



  

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users