Late April severe weather risk ~Mon thru next Mon 4/24-5/01

[Stebo]

1 minute ago, bjc0303 said:

I have specifically read/heard that it makes no statistical difference in model verification (when talking about before/after the wave moves onshore).

In which case there would be qualitative data to back that up. Considering like Hoosier mentioned most NWS offices refer to RAOB data often in AFDs, I think what you heard isn't correct or isn't the whole story.

[bdgwx]

I have to be honest. I've read the same thing from NOAA presentations and other data deniability experiments. RAOBs are important for long range forecasts (because they sample the stratosphere), but in most other cases there is usually a replacement dataset (like commercial aircraft) that can take the place of RAOBs. In fact, in the case of aircraft it is said that their instrumentation (at least the newest generation) is as good as RAOBs and often better. I've seen dtk mention that these special RAOB soundings for approaching storms don't have a significant effect on the models.  

[bjc0303]

1 minute ago, bdgwx said:

I have to be honest. I've read the same thing from NOAA presentations and other data deniability experiments. RAOBs are important for long range forecasts (because they sample the stratosphere), but in most other cases there is usually a replacement dataset (like commercial aircraft) that can take the place of RAOBs. In fact, in the case aircraft it is said that their instrumentation (at least the newest generation) is as good as RAOBs and often better. I've seen dtk mention that these special RAOB soundings for approaching storms don't really have any effect on the models.  

That can't be right. Or is NOAA run by a bunch of OU sophomores who don't understand weather? Hmmmm.

[CGChaser]

*returning to chat*

[Stebo]

3 minutes ago, bdgwx said:

I have to be honest. I've read the same thing from NOAA presentations and other data deniability experiments. RAOBs are important for long range forecasts (because they sample the stratosphere), but in most other cases there is usually a replacement dataset (like commercial aircraft) that can take the place of RAOBs. In fact, in the case of aircraft it is said that their instrumentation (at least the newest generation) is as good as RAOBs and often better. I've seen dtk mention that these special RAOB soundings for approaching storms don't have a significant effect on the models.  

If that is the case that would be quite surprising. Even at that though, it isn't satellite, it is still instrumentation from aircraft. The thing with aircraft data though, it is mostly at one level, most aircraft are cruising across the Pacific at 20000' to 30000'. At that point you are above a good portion of the the atmosphere.

[brettjrob]

Here are a couple papers tangentially related to the RAOB debate:

Zapotocny et al. (2008) - found RAOBs to be as important as the most helpful satellite data, but this was with the modeling systems of 2007.

Privé et al. (2014) - found that launching radiosondes at 06z and 18z would lead to a "small but statistically significant" improvement in NWP, particularly for the off-hour runs.

From what I recall in lectures, seminars, and conversations here at the NWC, leading DA scientists tend toward the idea that "wait for it to move onshore" is a fallacy, with our modern DA systems. But we're talking about a very narrow issue, and one which I'm not sure any papers have directly addressed. Someone like dtk who works on DA operationally full-time would probably know best.

[bjc0303]

And for clarification since it appears from a few posts to be lost on a few of you, I was literally only talking about the fallacy mentioned above.

 

[andyhb]

Well this was really the discussion I was looking forward to heading towards hour 0 with this...

[Stebo]

5 minutes ago, brettjrob said:

Here are a couple papers tangentially related to the RAOB debate:

Zapotocny et al. (2008) - found RAOBs to be as important as the most helpful satellite data, but this was with the modeling systems of 2007.

Privé et al. (2014) - found that launching radiosondes at 06z and 18z would lead to a "small but statistically significant" improvement in NWP, particularly for the off-hour runs.

From what I recall in lectures, seminars, and conversations here at the NWC, leading DA scientists tend toward the idea that "wait for it to move onshore" is a fallacy, with our modern DA systems. But we're talking about a very narrow issue, and one which I'm not sure any papers have directly addressed. Someone like dtk who works on DA operationally full-time would probably know best.

Yeah and in a situation like this a small change can make a massive difference. We have both been following weather for ages, we have seen changes occur more often than not once the energy comes ashore.

I have no problem admitting I am wrong if data does show that but I'd like to actually see data to clarify it vs being told by someone that they heard things. It is very easy to mishear stuff, qualitative data doesn't lie though.

The two studies above are a great start though on the subject.

 

[bdgwx]

Whether it's a myth or not the point is that Friday/Saturday is still 4-5 days away. That is an eternity in weather forecasting. A lot can change between now and then. Plus, we are in a climatologically favorable time of year for big outbreaks and we have an energetic longwave trough that is practically a given so I'm firmly in the wait and see camp.

[bjc0303]

2 minutes ago, Stebo said:

Yeah and in a situation like this a small change can make a massive difference. We have both been following weather for ages, we have seen changes occur more often than not once the energy comes ashore.

Sure they can.

I think your second point is more coincidence than anything. Unless you're dealing with high-predictability patterns/features, you're going to see change. We've seen change with almost every model run, independent of the wave "coming ashore."

Just now, bdgwx said:

Whether it's a myth or not the point is that Friday/Saturday is still 4-5 days away. That is an eternity in weather forecasting. A lot can change between now and then. Plus, we are in a climatologically favorable time of year for big outbreaks and we have an energetic longwave trough that is practically a given so I'm firmly in the wait and see camp.

I'm in the wait and see camp, although it's looking increasingly likely to be a run-of-the-mill type severe event as brett said. I agree; I'd bet a lot WILL change between now and then, but simply being in a climatologically favorable period of time for outbreaks isn't enough when you have a gulf-penetrating cold front with offshore flow forecast even 36 hours before the event. 

 

Believe me, I wish as much as the next guy that future model runs will change their tune. I'm sure they will - and as is clear when looking at the models, a 3-6 hour faster arrival of the wave would lend itself to a pretty impressive large-scale environment. There's a lot of uncertainty here - the wave's distance, aside from the fact we're 100 hours out, doesn't seem like it's going to make a difference.

[MattPetrulli]

5 minutes ago, bdgwx said:

Whether it's a myth or not the point is that Friday/Saturday is still 4-5 days away. That is an eternity in weather forecasting. A lot can change between now and then. Plus, we are in a climatologically favorable time of year for big outbreaks and we have an energetic longwave trough that is practically a given so I'm firmly in the wait and see camp.

Yeah lets just stick to this event on this thread and have the myth debate on another.

[brettjrob]

4 minutes ago, Stebo said:

Yeah and in a situation like this a small change can make a massive difference. We have both been following weather for ages, we have seen changes occur more often than not once the energy comes ashore.

Yeah; the addition of tons of RAOBs along the west coast can only be a good thing once it does come ashore. It's just a question of magnitude. Obviously, the 96-h forecast when the energy is all offshore will be less accurate than the 72-h forecast when it's coming ashore, and most of that is simply the lead time decreasing. Did the increased RAOB sampling account for 1% or 10% of the total improvement in predictability over that period? That's the type of question people seem to be hung up on, for better or worse. It seems the only way to answer it would be to run special experiments with the GFS where you deny all RAOBs west of the Rockies the day the trough comes onshore, or something like that -- and for many cases. Not sure it's been done.

[Stebo]

8 minutes ago, brettjrob said:

Yeah; the addition of tons of RAOBs along the west coast can only be a good thing once it does come ashore. It's just a question of magnitude. Obviously, the 96-h forecast when the energy is all offshore will be less accurate than the 72-h forecast when it's coming ashore, and most of that is simply the lead time decreasing. Did the increased RAOB sampling account for 1% or 10% of the total improvement in predictability over that period? That's the type of question people seem to be hung up on, for better or worse. It seems the only way to answer it would be to run special experiments with the GFS where you deny all RAOBs west of the Rockies the day the trough comes onshore, or something like that -- and for many cases. Not sure it's been done.

Yeah that is the biggest issue, there is no real way of experimenting this outside of past data and at that the GFS has changed so much over the years so you can only look back so far. I do think people are misunderstanding though, I am not trying to imply that there will be some sort of massive unworldly change but in a highly static situation this week, any subtle changes could have enough of an implication to make a difference.

[Radtechwxman]

Well until you provide proof of this your argument is kind of collapsing on itself. I'm with everyone else. ROAB sampling is much more important than you are giving it credit to. If it was only 10% I dont think we would see the effort that is put forth. What happens with this lead system will lay the ground work for the big end of week system. Data gathered from this system over land will most definitely impact model solutions. There is still plenty of time for major changes for the good or bad. 

[bjc0303]

5 minutes ago, StormChaser4Life said:

Well until you provide proof of this your argument is kind of collapsing on itself. I'm with everyone else. ROAB sampling is much more important than you are giving it credit to. If it was only 10% I dont think we would see the effort that is put forth. What happens with this lead system will lay the ground work for the big end of week system. Data gathered from this system over land will most definitely impact model solutions. There is still plenty of time for major changes for the good or bad. 

Collapsing in on itself even with the fact that most leading DA folk tend to think it's a fallacy?

 

My point isn't that radiosondes are useless, folks. It's that the wave being sampled by radiosondes once it comes onshore isn't nearly the deal you are trying to make it out to be. If this thing ends up being forecast to be 4/27/2011 on steroids it won't be because the damn wave was sampled by a RAOB, holy ****.

[JoMo]

Moving on... The 00z NAM is coming in......

[bjc0303]

1 minute ago, JoMo said:

Moving on... The 00z NAM is coming in......

Whether it makes a meaningful difference or not, who knows, but the 00z NAM at the least seems ever so slightly quicker to return to a more favorable pattern for moisture return.

[cgidd8]

50 minutes ago, andyhb said:

Well this was really the discussion I was looking forward to heading towards hour 0 with this...

After saving up pto all fall/winter and eager to take some time off this week on a chasecation, this is my sentiment right about now.

[Drz1111]

1 hour ago, bjc0303 said:

Lol, who are you?

I'm actually wrapping up my degree, bud. What background do you have in meteorology?

I've never heard an official number from a paper so I could be wrong, but I'm fairly confident that it makes no difference. I've heard from respectable meteorologists (not some wx weenie on americanwx) repeatedly that the "better sampling" thing is a complete myth. Not to mention the poor spatial resolution radiosondes offer. 

Priceless.  You're not an OU sophomore, you're an OU junior.  You can always tell the kids speaking in absolutes, even in hard sciences - it's no different than the radical feminists over in the women's studies department.  One of the things that will make you a better scientist (or forecaster) as you get older is realizing that nothing is quite so absolute.

It's been several years since I was in grad school, but there are two implied complexities in your blanket statement that you're overlooking.  First:  what purpose does direct measurement have for providing data to a model?  If it's just extra points on a grid, then yes, obviously remote sensing is providing 90%+ of the data - that's inherent to the limits of the direct measurement grid.  But at least with the models I worked with (which were, admittedly, ocean models and not atmospheric), we used direct measurement to calibrate / ground-truth remotely-sensed data.  I have always assumed they're using it similarly in atmospheric models.  That means that the widely-spaced grid can have a surprisingly meaningful effect.

Second:  how do you verify what you're arguing?  If your proposition is:  "the wave coming onshore isn't going to alter the verification of the 72 hr forecast global upper-air pattern produced by a global model", then you are tautologically going to be correct (think about why).  if the question is whether or not it might decrease uncertainty as to forecast severe weather, that's a tougher question.  Do you really think that's ever been measured rigorously?  How would define your sample? (I'll bet anything it's rarely, if ever, been done.)

I have a lot more knowledge of EC cyclogenesis than I do of severe, but I can tell you that research has been done for EC cyclogenesis and it's come down on the side of it mattering, most famously with the 12/26/10 snowstorm.  It follows that it might matter for Plains severe.  But neither you nor I know, because not only has no one measured whether there's an effect, I'm not sure anyone has properly defined how you would ask the question.

Saying "someone told this to me and they weren't a weenie" is an appeal to authority, and not even a particularly strong one.  Pardon the French, but you need to bring stronger **** than that to the table once you graduate.

[1900hurricane]

Moisture return trends are encouraging on the NAM.

 

[jojo762]

2 minutes ago, 1900hurricane said:

Moisture return trends are encouraging on the NAM.

 

Was just about to post this exact GIF. The trends in the SFC Theta-E are even more encouraging (as you can see with the SFC barbs).

[bjc0303]

11 minutes ago, Drz1111 said:

Priceless.  You're not an OU sophomore, you're an OU junior.  You can always tell the kids speaking in absolutes, even in hard sciences - it's no different than the radical feminists over in the women's studies department.  One of the things that will make you a better scientist (or forecaster) as you get older is realizing that nothing is quite so absolute.

It's been several years since I was in grad school, but there are two implied complexities in your blanket statement that you're overlooking.  First:  what purpose does direct measurement have for providing data to a model?  If it's just extra points on a grid, then yes, obviously remote sensing is providing 90%+ of the data - that's inherent to the limits of the direct measurement grid.  But at least with the models I worked with (which were, admittedly, ocean models and not atmospheric), we used direct measurement to calibrate / ground-truth remotely-sensed data.  I have always assumed they're using it similarly in atmospheric models.  That means that the widely-spaced grid can have a surprisingly meaningful effect.

Second:  how do you verify what you're arguing?  If your proposition is:  "the wave coming onshore isn't going to alter the verification of the 72 hr forecast global upper-air pattern produced by a global model", then you are tautologically going to be correct (think about why).  if the question is whether or not it might decrease uncertainty as to forecast severe weather, that's a tougher question.  Do you really think that's ever been measured rigorously?  How would define your sample? (I'll bet anything it's rarely, if ever, been done.)

I have a lot more knowledge of EC cyclogenesis than I do of severe, but I can tell you that research has been done for EC cyclogenesis and it's come down on the side of it mattering, most famously with the 12/26/10 snowstorm.  It follows that it might matter for Plains severe.  But neither you nor I know, because not only has no one measured whether there's an effect, I'm not sure anyone has properly defined how you would ask the question.

Saying "someone told this to me and they weren't a weenie" is an appeal to authority, and not even a particularly strong one.  Pardon the French, but you need to bring stronger **** than that to the table once you graduate.

nope try again 

[bjc0303]

8 minutes ago, jojo762 said:

Was just about to post this exact GIF. The trends in the SFC Theta-E are even more encouraging (as you can see with the SFC barbs).

Those trends help one of what, in my opinion, is a two part problem. The other being cyclogenesis timing. 

[jojo762]

00z NAM brings a nice tongue of 60s DPs into southern OK by 12Z Friday... But shows significant cyclogenesis occuring on Thursday evening which complicates things...

[Stebo]

1 minute ago, bjc0303 said:

nope try again 

No offense dude, but in a science discussion you aren't going to get far with this snarky attitude. It is actually insulting to the rest of us who are trying to have a decent discussion here.

[bjc0303]

Just now, Stebo said:

No offense dude, but in a science discussion you aren't going to get far with this snarky attitude. It is actually insulting to the rest of us who are trying to have a decent discussion here.

I meant no to his guess of my class status within the department. The rest of his post was cool but I'm not arguing this anymore.

[bjc0303]

2 minutes ago, jojo762 said:

00z NAM brings a nice tongue of 60s DPs into southern OK by 12Z Friday... But shows significant cyclogenesis occuring on Thursday evening which complicates things...

GFS also shows this. Essentially veers the surface pattern to the east until renewed cyclo. occurs. On the GFS it's too late, on the euro it is as well but only by 3-6 hours or so.

 

Time will tell.

[Stebo]

Just now, bjc0303 said:

I meant no to his guess of my class status within the department. The rest of his post was cool but I'm not arguing this anymore.

Your class status is irrelevant though if you are about to graduate, I would suggest if you are going to make a qualitative post about something back it up with data and reasoning. You should be able to if you are as far along in the degree as you are suggesting.

[Stebo]

4 minutes ago, jojo762 said:

00z NAM brings a nice tongue of 60s DPs into southern OK by 12Z Friday... But shows significant cyclogenesis occuring on Thursday evening which complicates things...

Yeah that is a good start, the bigger part though is before that. The intrusion into the Gulf as noted by the imagery is lessening. In other words the models were going too strong with Tuesday/Wednesday's system and are starting to back off. Positive steps so far.