86 replies to this topic

[Bob Chill]

I think this is a good idea for a thread. I think some folks are afraid to ask basic questions about models and weather for fear of getting ridiculed. There are alot of finer details that I really want to understand better but it never seems appropriate to ask during storm threads.

Here are some guidelines I would like to have in the thread:

1. Let's try not to flood the thread with too many active questions at a time. It shouldn't take too long to resolve each one before moving on.
2. No banter please. Just Q&A and teaching/learning here.
3. No questions are too stupid to ask. More knowledge improves post content across the board.
4. All topics are fine. Not just winter. Severe, topical, teleconnections, ENSO, anything is welcome.
5. If someone posts a link in response to a question, read the link before asking anymore questions. Don't be lazy.
6. Do not reply to a question or be on the teaching side of a discussion if you don't have a clear understanding of the question at hand. Bad information is at the root of alot of bad posts here.


Ok, since it's my thread, I'm asking the first question. Below is a common sounding post and I don't know how to read it correctly at all. To narrow down my question, I know what the wind axis, hpa axis, and temp axis are. I do not know how to property interpret the colored lines and I don't understand what the light grey grid represents.


 Sounding.png   52.41K   2 downloads

[wxtrix]

here's a link to a tutorial to read a Skew-T (the type of sounding you posted--there are other types): http://www.atmos.mil...SkewT_Home.html

other reading:

http://texasstormchasers.com/tag/how-to-read-a-skew-t/
http://www.atmos.mil...kewT_HowTo.html
http://www.americanw...post__p__435795
http://www.spc.noaa.gov/exper/soundings/help/index.html
http://www.theweatherprediction.com/thermo/soundings/

[ThreeRedheads]

Ellinwood posted this on his site a couple of weeks ago. It was helpful:

 sounding_primer.png 1000×800 .png   465.86K   44 downloads

*green line is dewpoint. It was cut off on my screen cap of the legend.

[ThreeRedheads]

Thanks for the thread, Bob. There's A LOT I need to learn.

Trixie-great links! Can't say I understand it all, but it's a starting point.

[Bob Chill]

You see, this is exactly what I'm talking about. My question is answered. Thanks gals, excellent stuff. I will thoroughly read everything as I find the time. E-woods post is an excellent starting point for anybody interested in learning this stuff.

[WxUSAF]

Guide to the gray lines:

Solid lines that slant upward and to the right from the surface are temperature--> hence the "skew" of skew-T
Solid lines parallel to the bottom are pressure levels
Solid curved lines that curve upward and to the left are dry adiabats
Dashed curved lines that curve upward and then slightly bend left are moist adiabats
Dotted straight lines that slant upward and right are water vapor mixing ratios

[wxtrix]

another thing that helps snow weenies read models better is the 500 mb chart: http://www.theweathe...rts/500/basics/

[Bob Chill]

Guide to the gray lines:

Solid lines that slant upward and to the right from the surface are temperature--> hence the "skew" of skew-T
Solid lines parallel to the bottom are pressure levels
Solid curved lines that curve upward and to the left are dry adiabats
Dashed curved lines that curve upward and then slightly bend left are moist adiabats
Dotted straight lines that slant upward and right are water vapor mixing ratios

Excellent. Just excellent. I'm pretty good with temp and p levels but I really had no idea what the curved lines meant and I was a bit afraid to ask until today.

Thank you for the simple and straightforward definition of the lines.

[Fozz]

This thread should probably be pinned.

Now that spring is on the way, I will need a lot of learning with severe weather since I know next to nothing beyond the noob level.

[winterymix]

Guide to the gray lines:

Solid lines that slant upward and to the right from the surface are temperature--> hence the "skew" of skew-T
Solid lines parallel to the bottom are pressure levels
Solid curved lines that curve upward and to the left are dry adiabats
Dashed curved lines that curve upward and then slightly bend left are moist adiabats
Dotted straight lines that slant upward and right are water vapor mixing ratios

A few more questions for meteorologist(s):

Which models have the best verification scores at the following time intervals before a possible event of interest?
Best models 10 days before a potential event
Best models 7 days before a potential event
Best models 4 days before a potential event
Best models 2 days before a potential event
Best models 1 day before a potential event

When are SREF maps most useful?
The NAM is not really horrid, is it? Should it only be run to 48 hours?

Best sources for ecmwf maps and extended maps. Question: places like weather underground that post extended euro maps; is this in-house proprietary solution being used or is there a standard methodology?

Best sources of GFS ensemble maps?
Fastest sources for updates of the Canadian models? Is it practical to find the GGEM out to ten days?
Are ensembles of Canadian models useful and where is a good place to find those?
Help needed with model biases:
I found this links but are there better sources for updated model bias information?
Outdated, perhaps still useful:
http://www.hpc.ncep....s/biastext.html
http://www.hpc.ncep.noaa.gov/mdlbias/
http://www.hpc.ncep....l2.shtml#biases

[dtk]

Which models have the best verification scores at the following time intervals before a possible event of interest?
Best models 10 days before a potential event
Best models 7 days before a potential event
Best models 4 days before a potential event
Best models 2 days before a potential event
Best models 1 day before a potential event

This is tough to answer actually since it really depends on the type of event/season, etc. For the day 9-10 range, you have to use ensemble. The deterministic models only score about 0.5 or so for 500 hPa height AC (generally, 0.6 is used as a cutoff to define forecasts that have some skill). Errors are large at this lead time.

For days 4-7 (or so), ECMWF has higher scores than the other operational globals. The UKMet and GFS generally score 2nd behind the EC, with the Canadian behind that (especially days 6-7)..and then others even behind that. That's not to say there aren't occasions where the other models beat the EC, because it does happen. These metrics are also typically hemispheric, and each model has their own strengths/weaknesses by region, regime, season, etc. The EC is also less prone to "drops" in skill compared to the other operational models.

I can't really comment much on the short range, though the ECMWF is going to be a good bet (you don't score well at day 5 without doing well at day 1). It's being run at high enough spatial resolution to take seriously for many different types of phenomena.

Best sources of GFS ensemble maps?
Fastest sources for updates of the Canadian models? Is it practical to find the GGEM out to ten days?
Are ensembles of Canadian models useful and where is a good place to find those?

What kinds of GFS ensemble maps are you looking for? We generate lots of products based on something called the NAEFS, which combines the GEFS and Canadian ensemble members.

Why would you look at the GGEM out to ten days? Why would you look at any deterministic model out to ten days (other than for "fun")?

The Canadian ensemble is extremely useful if you're familiar with it. It's the only major operational global ensemble that is truly multi-model (the GEFS and EC ensemble have parameterizations to mimic model/error and uncertainty)....along the lines of the SREF. That is not to say it is more skillful than the EC EPS and GEFS, however. I think that it is prone to being a bit over dispersive, (i.e. it can exhibit too large of spread on occasion).

Help needed with model biases:
I found this links but are there better sources for updated model bias information?
Outdated, perhaps still useful:
http://www.hpc.ncep....s/biastext.html
http://www.hpc.ncep.noaa.gov/mdlbias/
http://www.hpc.ncep....l2.shtml#biases

The problems with these kinds of lists is that the models are updated fairly frequently.....meaning their biases change fairly often. As an example, the version of the GFS that we run now is nothing like the version we ran even as recently as two years ago. Too many myths exist about the models based on how things were ten years ago. I've tried to dispel some of the most egregious ones in other threads.

[WxUSAF]

I see dtk replying, so I hope he fields the verification score question since I'm not going to.

I think the SREFs have some value inside of 72 hours and certainly inside 48. I really enjoy looking at the plume charts to get a sense of the spread and clusters of the different solutions.

The NAM certainly struck out this weekend, but I'm certainly not as down on it as zwyts and Ian. Much of my forecasting in the AF was very short-range, so perhaps I just didn't notice if/when the NAM blew a 72 or 48 hour forecast. But it always did pretty well for me, and indeed, we were obligated to use it unless it was really REALLY out to lunch.

I use Raleighwx's website for most of my model analysis, but I don't have a favorites menu full of wx links like many on here.

I wouldn't worry about GGEM ensembles. The GGEM doesn't go out to 10 days on its home website, so I wouldn't bother looking for it even if it exists.

[ETauntonMA]

Also, there is a whole forum to ask any questions you may have.

http://www.americanw...eteorology-101/

[Bob Chill]

Also, there is a whole forum to ask any questions you may have.

http://www.americanw...eteorology-101/

I wasn't really trying to re-invent the wheel here. There are alot of folks in subforums who either don't want to or are scared to post in the main forum. Plus, we can get into mby stuff here because (just like every other area) the ma has it's own nuances with its climo and we can really get into detail with both winter storms and severe in our specific regions.

Mets, appreciate the answers to winterymix's model questions. I've slowly learned over time to better understand what the models are doing with specific setups. The nam is definitely superior with CAD events. I always put more stock in the nam's handling of the surface and 850 temps when flow is from the n & ne in the lower levels.

Euro struggles a bit in the lr with split flow. Especially with vorts in the southern stream. It catches up and outperforms the other globals once we are at the day 5 range but beyond that it's important to understand that it likes to move things out of the sw slower and that has a large effect on the evolution down the line.

These obs of the nam and euro are just what I've seen over the least 5 years or so. Please correct me if I'm posting bad info here.

[winterymix]

To Bob Chill, DTK and WxUSAF: Thank you for the collegial responses.
To DTK: You have a lot to teach us. Regarding frequent and evolving updates to model biases,
perhaps you can share favorite links at the weather.gov site?

To ETauntonMA: This place can be rude. Just recently, during a debate here about the relative
worth of the NAM vs. EURO for the weekend suppressed storm, someone here stated that the
discussion was "over my head". As the two METS mentioned above prove, when one is secure in
their knowledge, they can maintain a collegial tone.

[WinterWxLuvr]

Great thread Bob. I think Fozz is right. Pin this sucker, keep it right at the top of the page. I'm on the phone right now but I can't wait to examine the replies to your question.

[HubbDave]

Also, there is a whole forum to ask any questions you may have.

http://www.americanw...eteorology-101/

I think having one in a subforum is a good idea. Even if it mirrors one in the main thread.

This is helpful

[ThreeRedheads]

We can use both areas of the site to gain more knowledge. The advantage to having our own thread is the fact that we're more familiar with each other and may feel more comfortable asking questions.

I have so many questions, I don't know where to start!

[smokeybandit]

I'm curious of the model maintenance. Throughout the various seasons, there's always discussion of "Model X does horrible for this type of system" or "Model Y always over overdoes the precipitation."

How often are the algorithms and/or inputs tweaked? I'd think the owners of the models would want to adjust them as to not be so far off in certain situations.

[dtk]

I'm curious of the model maintenance. Throughout the various seasons, there's always discussion of "Model X does horrible for this type of system" or "Model Y always over overdoes the precipitation."

How often are the algorithms and/or inputs tweaked? I'd think the owners of the models would want to adjust them as to not be so far off in certain situations.

These numerical models are very complicated, nonlinear beasts. It's easy to say "tweak" something to target a particular problem, but these tweaks always have unintended consequences. Many things within the models themselves have feedback processes. For example, you can't modify things within the algorithms that handle cloud processes, without also impacting precipitation, and radiation, and surface fluxes, and....

Now, we do in fact try to target problem areas. However, we can't just make modifications to the models on a whim (or very frequently). NCEP has a huge customer base, and many factions within have a say as to whether or not certain things can be changed. There is a very rigorous process for testing and evaluating changes prior to implementation, and it happens in many stages (and chews up a lot of resources). Because of the scope of what we do, and the amount of testing that needs to be done, we typically make changes to the major systems at most once/year.

[eurojosh]

Ellinwood posted this on his site a couple of weeks ago. It was helpful:



*green line is dewpoint. It was cut off on my screen cap of the legend.

This was a good question and a great graphic - thanks to you both.

[Ian]

The NAM certainly struck out this weekend, but I'm certainly not as down on it as zwyts and Ian. Much of my forecasting in the AF was very short-range, so perhaps I just didn't notice if/when the NAM blew a 72 or 48 hour forecast. But it always did pretty well for me, and indeed, we were obligated to use it unless it was really REALLY out to lunch.

some of my commentary was strong just because. i said the nam has uses. i don't particularly believe one of them is to get a great idea of what a coastal is going to do 78 hours out. tho i do believe in 09-10 it was often very quick to get the stream interaction fairly correctly.. but those storms were fairly well modeled overall. if it's raining the next day the nam is usually first in my arsenal for timing etc.

[smokeybandit]

These numerical models are very complicated, nonlinear beasts. It's easy to say "tweak" something to target a particular problem, but these tweaks always have unintended consequences. Many things within the models themselves have feedback processes. For example, you can't modify things within the algorithms that handle cloud processes, without also impacting precipitation, and radiation, and surface fluxes, and....

Now, we do in fact try to target problem areas. However, we can't just make modifications to the models on a whim (or very frequently). NCEP has a huge customer base, and many factions within have a say as to whether or not certain things can be changed. There is a very rigorous process for testing and evaluating changes prior to implementation, and it happens in many stages (and chews up a lot of resources). Because of the scope of what we do, and the amount of testing that needs to be done, we typically make changes to the major systems at most once/year.

Thanks. I wouldn't expect willy nilly changes to the models, so once a year isn't much of a surprise. Plus I'd imagine you need an extended amount of feedback to best change the model. I can just see the predictable anomalies in the models really frustrating for those who maintain the models.

[usedtobe]

Thanks. I wouldn't expect willy nilly changes to the models, so once a year isn't much of a surprise. Plus I'd imagine you need an extended amount of feedback to best change the model. I can just see the predictable anomalies in the models really frustrating for those who maintain the models.

As DTK noted what really makes it tough are two things, one the physics of the atmosphere is very non linear so if you change something like the convective or cloud algorithms/parameterization schemes you don't always know how that might impact other things as condensation within clouds releases and changes the stability, less clouds impacts the amounts of radiation hitting the ground which can change the surface temperature and because they have many different kins of users they can't just make a change without consulting and notifying them of the change and what impact it might have on subsequent forecasts. While HPC the guys that work on the models were really responsive if the forecasters thought they saw something that looked to be a problem. They also have to do a lot of testing of any change before it is implemented. It can be a thankless task when people run down the model that you are working to improve.

[ETauntonMA]

I wasn't really trying to re-invent the wheel here. There are alot of folks in subforums who either don't want to or are scared to post in the main forum. Plus, we can get into mby stuff here because (just like every other area) the ma has it's own nuances with its climo and we can really get into detail with both winter storms and severe in our specific regions.

Mets, appreciate the answers to winterymix's model questions. I've slowly learned over time to better understand what the models are doing with specific setups. The nam is definitely superior with CAD events. I always put more stock in the nam's handling of the surface and 850 temps when flow is from the n & ne in the lower levels.

Euro struggles a bit in the lr with split flow. Especially with vorts in the southern stream. It catches up and outperforms the other globals once we are at the day 5 range but beyond that it's important to understand that it likes to move things out of the sw slower and that has a large effect on the evolution down the line.

These obs of the nam and euro are just what I've seen over the least 5 years or so. Please correct me if I'm posting bad info here.

To Bob Chill, DTK and WxUSAF: Thank you for the collegial responses.
To DTK: You have a lot to teach us. Regarding frequent and evolving updates to model biases,
perhaps you can share favorite links at the weather.gov site?

To ETauntonMA: This place can be rude. Just recently, during a debate here about the relative
worth of the NAM vs. EURO for the weekend suppressed storm, someone here stated that the
discussion was "over my head". As the two METS mentioned above prove, when one is secure in
their knowledge, they can maintain a collegial tone.

I think having one in a subforum is a good idea. Even if it mirrors one in the main thread.

This is helpful

That's why I said also. I agree that this is a great topic.

[winterymix]

...It can be a thankless task when people run down the model that you are working to improve.

It seems that folks like you and DTK have much to teach about the subtle nature of numerical modeling.
The NAM and to some extent, the GFS get trashed so much in relation to the European model. It can't be
just that simplistic. It seems to me that meteorologists earn their keep when they know how to blend
the relative strengths of each model. If I understood DTK correctly, he stated that yes, the European
model has excellent verification scores over long intervals but that skill doesn't always translate
to predicting sensible weather in certain geographic regions and in certain climate regimes.

In real life, in real NWS offices, do experienced meteorologists refer to a particular model
as a piece of garbage? (e.g. "Throw out the NAM; it is on crack!")

[Bob Chill]

As DTK noted what really makes it tough are two things, one the physics of the atmosphere is very non linear so if you change something like the convective or cloud algorithms/parameterization schemes you don't always know how that might impact other things as condensation within clouds releases and changes the stability, less clouds impacts the amounts of radiation hitting the ground which can change the surface temperature and because they have many different kins of users they can't just make a change without consulting and notifying them of the change and what impact it might have on subsequent forecasts. While HPC the guys that work on the models were really responsive if the forecasters thought they saw something that looked to be a problem. They also have to do a lot of testing of any change before it is implemented. It can be a thankless task when people run down the model that you are working to improve.

I think both you and dtk have really provided some valuable insight to mechanics of models and I find it dissappointing that so many posters feel the need to say "models are terrible...blah blah blah" all the time. It is extremely unrealistic to think the physics of the atmosphere can be perfectly replicated in a computer model (especially past a few days in many cases). What the models are able to do already is astounding imo.

I think sometimes we forget (or don't even consider) the fact that the surface of the earth is almost 200 million square miles and the volume of our effective atmoshpere is something close to 1 billion cubic miles. Saying a model is god awful because the snow missed you by even 200 miles is a ridiculous statement in alot of ways. Before criticizing a model, one should be able to thoroughly read the model runs and understand why it is likely wrong. If you only interpret a model absolutely verbatim, there is no room to criticize anything.

dtk- your technical model insight here is invaluable. Please continue to share as much as possible.

[nj2va]

Ellinwood posted this on his site a couple of weeks ago. It was helpful:

 sounding_primer.png 1000×800 .png   465.86K   44 downloads

*green line is dewpoint. It was cut off on my screen cap of the legend.

This is awesome, thanks for posting this. So in your example, at the 500mb level, the temp is ~-29C?

[UIWWildthing]

I will be reading this thread. A lot. Thanks for posting this!

[baroclinic_instability]

I think both you and dtk have really provided some valuable insight to mechanics of models and I find it dissappointing that so many posters feel the need to say "models are terrible...blah blah blah" all the time. It is extremely unrealistic to think the physics of the atmosphere can be perfectly replicated in a computer model (especially past a few days in many cases). What the models are able to do already is astounding imo.

I think sometimes we forget (or don't even consider) the fact that the surface of the earth is almost 200 million square miles and the volume of our effective atmoshpere is something close to 1 billion cubic miles. Saying a model is god awful because the snow missed you by even 200 miles is a ridiculous statement in alot of ways. Before criticizing a model, one should be able to thoroughly read the model runs and understand why it is likely wrong. If you only interpret a model absolutely verbatim, there is no room to criticize anything.

dtk- your technical model insight here is invaluable. Please continue to share as much as possible.

Yes, you bring up a good point. I do think individuals should at least have a little numerical modeling/programming experience before they discredit the model completely. From a meteorologist standpoint, even we should not "bash" the models, but some level of criticism is not only ok but actually useful to both devs/users. I try and stray away from outright "bashing" since I am well aware of how incredible numerical modeling really is. The fact we can simulate weather with relative skill out to day 6 is simply incredible, and it is a nod to human ingenuity (especially the data assimilation aspect). What bothers me are the individuals who bash the models and use them as an excuse for their own busted forecasts when they never put any thought into it to begin with. Weenies with zero experience in anything related to the physics of the atmosphere or numerical modeling who mindlessly bash the models are also especially egregious.

That said, certain models, under certain circumstances, truly do have limited utility and can, in some cases, be more harmful than good.

[CoastalWx]

Yes, you bring up a good point. I do think individuals should at least have a little numerical modeling/programming experience before they discredit the model completely. From a meteorologist standpoint, even we should not "bash" the models, but some level of criticism is not only ok but actually useful to both devs/users. I try and stray away from outright "bashing" since I am well aware of how incredible numerical modeling really is. The fact we can simulate weather with relative skill out to day 6 is simply incredible, and it is a nod to human ingenuity (especially the data assimilation aspect). What bothers me are the individuals who bash the models and use them as an excuse for their own busted forecasts when they never put any thought into it to begin with. Weenies with zero experience in anything related to the physics of the atmosphere or numerical modeling who mindlessly bash the models are also especially egregious.

That said, certain models, under certain circumstances, truly do have limited utility and can, in some cases, be more harmful than good.

Like I said with the gambler analogy...with certain models...gotta know when to hold them....know when to fold them. I actually like the NAM for things like identifying heavy rain, CAD, and also rapid cyclogenesis as a strong PV interacts with the Gulf Stream. However, I don't really take seriously what the NAM shows until within 48 hrs and even then...it is met with questioning unless other guidance agrees.

[Coach McGuirk]

I think this is a good idea for a thread. I think some folks are afraid to ask basic questions about models and weather for fear of getting ridiculed. There are alot of finer details that I really want to understand better but it never seems appropriate to ask during storm threads.

Here are some guidelines I would like to have in the thread:

1. Let's try not to flood the thread with too many active questions at a time. It shouldn't take too long to resolve each one before moving on.
2. No banter please. Just Q&A and teaching/learning here.
3. No questions are too stupid to ask. More knowledge improves post content across the board.
4. All topics are fine. Not just winter. Severe, topical, teleconnections, ENSO, anything is welcome.
5. If someone posts a link in response to a question, read the link before asking anymore questions. Don't be lazy.
6. Do not reply to a question or be on the teaching side of a discussion if you don't have a clear understanding of the question at hand. Bad information is at the root of alot of bad posts here.


Ok, since it's my thread, I'm asking the first question. Below is a common sounding post and I don't know how to read it correctly at all. To narrow down my question, I know what the wind axis, hpa axis, and temp axis are. I do not know how to property interpret the colored lines and I don't understand what the light grey grid represents.


 Sounding.png   52.41K   2 downloads

The sounding is snow but slightly above freezing.

[mapgirl]

This thread is great - something I would love to see in the future are more in depth explanations of threats and all the pieces involved so those of us who don't fully grasp all the working parts are able to look at a model and figure out what is going on.

am19psu was doing that for me last week with the snow threat, he broke it down for me piece by piece, and after a few model runs I was able to see the different players myself.

[CoastalWx]

This thread is great - something I would love to see in the future are more in depth explanations of threats and all the pieces involved so those of us who don't fully grasp all the working parts are able to look at a model and figure out what is going on.

am19psu was doing that for me last week with the snow threat, he broke it down for me piece by piece, and after a few model runs I was able to see the different players myself.

It's good to also break down technical terms that we throw around here, into simpler terms. Sometimes I'm guilty of that myself, but it's good to let the enthusiasts know what we mean by deformation or frontogenesis...you know, things like that. All these words sound cool and all, but I'm sure some people have no idea what we mean, other than the fact that it sounds good for heavy snow..lol. People shouldn't be afraid to ask, heck even I'm still learning about various things in meteorology. The day you stop learning and don't care anymore, is the day you should probably hang it up.

[NE Balti Zen]

<sup>Here is something that I am hoping someone can explain in more layman's terms...I have read on the Jeff Haby site the definitions of a "Vort" and "Shortwave" and "Trough" (and read explanations of those terms elsewhere) but was wondering if a red tagger (or learned non-red tagger) could generally sum them up with relation to developing storms. For instance, it used to be, before discovering eastern, I would look at a map and only focus on the "L" with regard to a developing storm. It is clear now that how strong the "L" (or weak or if it forms) is dependent, maybe, on the above concepts, do I have that right? So, for example, the issues with the modeling of the storm last week were that there were a number of potential different energies (vorts?) that might lead to the development of the "L" that became the storm that got the southern VA folks snow. Is that generally right?

A related question...when I look at the 500 heights/vort maps, what should I be gleaning from where the "X" is placed with regard to the various vorts on the map?

I apologize ahead of time if this question was too disorganized (or stupid), but it is stuff that I am having trouble sorting out and, despite plenty of reading on it, have not been able to sort it out satisfactorily yet.</sup>