0-6km shear vector, 0-6km mean wind and storm motion...

[Andrea Griffa]

Hi all. I'd like to discuss with you guys how to plot on a hodograph the 0-6km shear vector and the mean wind and storm motion vector(to then calculate storm relative winds).

I've read on "the Comet program" you can determine the direction of the mean wind shear vector (but not the magnitude) simply by drawing a line from the point that plots the surface wind to the point plotting the 6-km (20-kft) wind. Again, we use the 0-6 km layer as the layer most effecting the storm.
Always from "the comet program": calculating the mean wind shear vector is simply a matter of averaging the x and y components of each of the single layer wind shear vectors.
Storm motion can be either observed or anticipated. Besides using the head of the storm motion vector as the origin, you can draw vectors from the head of the storm motion vector to the vector heads (points along hodograph) of your ground-relative winds.

But how can you determine the storm motion vector? is that the same thing to do for the mean wind?

I attach an hodo to work on. How could you draw o-6km shear, mean wind and storm motion?

 

[Jeff Snyder]

Concerning the 0-6km shear vector... Shear is the difference in wind (speed and direction) between two levels. For example, the 0-6km wind shear = V6km - V0km, where V6km is the 6km wind vector, and V0km is the 0km (surface) wind vector. So:
0-6km shear vector = 6km wind vector - 0km wind vector
I think it's common to use a thin layer-average wind instead of stricly using the 0km and 6km wind. For example, you may average the 0-1km wind and 5.5-6.5km wind.

So, for the sounding you gave, the 0-6k shear vector is given as the bright red vector:

Storm motion is more complicated... Remember, storm motion has two components -- propagation and translation. The translational component can be thought of as a leaf floating down the river -- it moves with the flow. So, if the winds at all levels are out of the southwest at 40mph, the translational component of storm motion is 40mph towards the northeast. The second component, propagation, is what makes this a little messy. For example, think of a gust front spread out equally in all directions from a downdraft. If we assume no low-level flow, you'll find convergence on the leading edge of that gust front. If we give 20mph southwesterly low-level flow, then there will be enhanced convergence along the southwest side of the gust front. If the cold pool is strong enough to force the parcels to their LFCs, then you'll likely find that convection will preferentially develop along that southwest edge. Without any translation (moving with the flow), the storm will look like it's actually "moving" southwest. For supercells, the presence of the mesocyclone results in quasi-horiztonal pressure perturbations that act to push or pull the updraft to the right (for cyclonic mesos) and to the left (for anticyclonic mesos).

As far as "methods"... For supercells, there a couple of rule-of-thumb methods -- namely the 15R85 (15 degrees to the Right at 85% of the mean wind) and 30R75 (30 degrees to the Right at 75% of the mean wind). More recently, however, has been the development of the Bunkers Method. I don't know a whole lot about the Bunkers method, so I'll let others explain that one.

The mean-wind technique (averaging the winds between certain levels) is probably the easiest, but doesn't take into account dynamic processes / motions deviating from translation. For the mean wind technique, just average the u and v components (lon. and lat.) of the 2km and 6km winds (for example). Of course, is a storm is relatively shallow, it's motion may be influence more by low-level winds than mid-level winds. Likewise, if we're in or near the tropics, where the equilibrium level is often very high, the upper-level flow may affect storm motion more than the low-levels.

You said you went through the COMET module, but I'm not sure if this is the one you are talking about -- http://meted.ucar.edu/convectn/ic411/

I also suggest: http://meted.ucar.edu/mesoprim/hodograf/
ftp://ftp.cira.colostate.edu/braun/visit/...etraining-3.ppt
http://www.spc.noaa.gov/publications/edwar...ards/motion.pdf

 

[Helge Tuschy]

Hi Jeff,

I also have a short question for you.

For using the ID method, you have to draw the "non-pressure weighted mean wind ".
Can you please give me a tip how I can find this mean wind and how far it differs from the pressure weighted mean wind ? Maybe you know a useful link about this subject

Thanks a lot,

Helge

 

[Kiel Ortega]

I've been playing with python and some code from a professor to develop some sounding and hodograph software to help practice coding. The ID method SUCKS to do; actually, it's not the ID method, but the fact you have to rotate the entire coordinate system to get correct calculations. It's not the easiest thing and took me a while to figure it out. But now it works (sorta, I linearly interpret the vertical wind field and get good results...but if I put in the actual obs, it's wrong ) I'll edit the ID method out of my code and post a link later (the code is on my machine at work).


!!!!UPDATE!!!!
I put the code up at http://weather.ou.edu/~kortega/soundings/idmethod.py I don't have very good comments right now b/c I was just trying to get it up before I went out for the evening...if I get some time next week (which will probably happen) I'll update the file to explain more fully what I did. However, for the ID method, if you read the Bunkers et al. paper, it follows right along with them.

 

[Andrea Griffa]

Concerning the 0-6km shear vector... Shear is the difference in wind (speed and direction) between two levels. For example, the 0-6km wind shear = V6km - V0km, where V6km is the 6km wind vector, and V0km is the 0km (surface) wind vector. So:
0-6km shear vector = 6km wind vector - 0km wind vector
I think it's common to use a thin layer-average wind instead of stricly using the 0km and 6km wind. For example, you may average the 0-1km wind and 5.5-6.5km wind.

So, for the sounding you gave, the 0-6k shear vector is given as the bright red vector:
http://www.tornadocentral.com/now/0-6kmshearvector.png

Storm motion is more complicated... Remember, storm motion has two components -- propagation and translation. The translational component can be thought of as a leaf floating down the river -- it moves with the flow. So, if the winds at all levels are out of the southwest at 40mph, the translational component of storm motion is 40mph towards the northeast. The second component, propagation, is what makes this a little messy. For example, think of a gust front spread out equally in all directions from a downdraft. If we assume no low-level flow, you'll find convergence on the leading edge of that gust front. If we give 20mph southwesterly low-level flow, then there will be enhanced convergence along the southwest side of the gust front. If the cold pool is strong enough to force the parcels to their LFCs, then you'll likely find that convection will preferentially develop along that southwest edge. Without any translation (moving with the flow), the storm will look like it's actually "moving" southwest. For supercells, the presence of the mesocyclone results in quasi-horiztonal pressure perturbations that act to push or pull the updraft to the right (for cyclonic mesos) and to the left (for anticyclonic mesos).

As far as "methods"... For supercells, there a couple of rule-of-thumb methods -- namely the 15R85 (15 degrees to the Right at 85% of the mean wind) and 30R75 (30 degrees to the Right at 75% of the mean wind). More recently, however, has been the development of the Bunkers Method. I don't know a whole lot about the Bunkers method, so I'll let others explain that one.

The mean-wind technique (averaging the winds between certain levels) is probably the easiest, but doesn't take into account dynamic processes / motions deviating from translation. For the mean wind technique, just average the u and v components (lon. and lat.) of the 2km and 6km winds (for example). Of course, is a storm is relatively shallow, it's motion may be influence more by low-level winds than mid-level winds. Likewise, if we're in or near the tropics, where the equilibrium level is often very high, the upper-level flow may affect storm motion more than the low-levels.

You said you went through the COMET module, but I'm not sure if this is the one you are talking about -- [url=http://meted.ucar.edu/convectn/ic411/]http://meted.ucar.edu/convectn/ic411/[/url]

I also suggest: [url=http://meted.ucar.edu/mesoprim/hodograf/]http://meted.ucar.edu/mesoprim/hodograf/[/url]
[url=ftp://ftp.cira.colostate.edu/braun/visit/Supercell_Teletraining-3.ppt]ftp://ftp.cira.colostate.edu/braun/visit/...etraining-3.ppt[/url]
[url]http://www.spc.noaa.gov/publications/edwards/motion.pdf[/url][/quote]

Thanks Jeff for those precious considerations. I tried to calculate storm motion and VLM and VRM with the BKF2 method and it doesn't seem so difficult.
But, the most important thing is more simple: when you draw down 0-6 km shear, if I'm not wrong, you can't calculate the vector magnitude just from calculating the length of vector. And how do you calculate it? Is it the same thing of mean wind(making the average of all quantities)?

Yopu said it's common to use a thin layer-average wind instead of stricly using the 0km and 6km wind. For example, you may average the 0-1km wind and 5.5-6.5km wind.
In this case? How could you compute 0-6 km shear value?


[img]http://digilander.libero.it/andreblu81/0-6kmshearvector.png

 

[Jeff Snyder]

Magnitude of shear vector = length of vector = SquareRoot [ (v2-v1)^2 + (u2-u1)^2 ]
In other words, graphically, the magnitude of the shear vector is just the length of that vector. If you're looking at numbers, you can apply the vector magnitude formula (squareroot[blah blah blah]).

For the 5.5km-6.5km average, just average the u and v components of the winds within that layer, which is usually pretty easy by using the wind data provided by a sounding. In most cases, I wouldn't worry too much about this, but you may want to consider it if there's a strong wind discontinuity near 6km. For example, if winds increase drastically immediately above 6km, it may be worth doing a layer-average since any storm probably doesn't care if the 6km winds are only slightly strong if the winds at, for example, 6.2km are considerably stronger. I suppose if the storm top is exactly 6km the >6km winds probably don't influence the storm too much, but I also suppose that's why some are switching to updraft-relative winds not so much in terms of the storm's motion, but in terms of the storm depth (such as the effective bulk shear on the SPC mesoanalysis page).

 

[Helge Tuschy]

!!!!UPDATE!!!!
..... if you read the Bunkers et al. paper, it follows right along with them.[/quote]

Hi Kiel,

thanks for your help. I'll try to get an answer, comparing the Bunkers papers with your code but I think I would be glad about some sentences about what exactly you did in your code ...when you have time later on...

Cheers, Helge

 

[Kiel Ortega]

Hi Kiel,

thanks for your help. I'll try to get an answer, comparing the Bunkers papers with your code but I think I would be glad about some sentences about what exactly you did in your code ...when you have time later on...

Cheers, Helge

Okay, I have tried my best! The descriptions help me (but that's meaningless to others)...any specific questions, PM me