Jason Boggs
Oh, if all things are equal, GR2AE will always be better. Better resolution among other things.
Bob Hartig
EF5
Yeah, that being the case, give me level 2 data. It's far more nuanced than level 3. And the volume scan of AE might prove helpful, and could be great for record keeping.
However, for you guys who know more about level 2 than I do, I have a couple questions that have been bugging me about GR2 ever since I started using it:
1. With storms constantly morphing and changing both speed and direction, how frequently (as a general rule) do you update your storm motion vector for SRV? "All things being equal," that little bit of manual operation is my one reservation for field use. GR3 takes care of this concern automatically.
2. Along the same lines: when you set the motion vector for one cell, since the operation is specific to a single storm, how good is GR2 at showing other areas of rotation in storms other than the one you've selected? What I mean is, you set your marker on cell A, which is moving northeast at 47 knots. Great, but what about cell B, which is presently moving east/northeast at 35 knots? And cell C, which is moving in a direction between both of these cells at a speed of 40 knots? Maybe you've also got cells D, E, F, G, and so on, each behaving in its own unique fashion. I don't know what kind of number-crunching GR3 goes through to do it, but it will manage to do a reasonable job of depicting rotation in all of these storms (assuming they all have mesocyclones). How would you compare GR2 in that respect?
However, for you guys who know more about level 2 than I do, I have a couple questions that have been bugging me about GR2 ever since I started using it:
1. With storms constantly morphing and changing both speed and direction, how frequently (as a general rule) do you update your storm motion vector for SRV? "All things being equal," that little bit of manual operation is my one reservation for field use. GR3 takes care of this concern automatically.
2. Along the same lines: when you set the motion vector for one cell, since the operation is specific to a single storm, how good is GR2 at showing other areas of rotation in storms other than the one you've selected? What I mean is, you set your marker on cell A, which is moving northeast at 47 knots. Great, but what about cell B, which is presently moving east/northeast at 35 knots? And cell C, which is moving in a direction between both of these cells at a speed of 40 knots? Maybe you've also got cells D, E, F, G, and so on, each behaving in its own unique fashion. I don't know what kind of number-crunching GR3 goes through to do it, but it will manage to do a reasonable job of depicting rotation in all of these storms (assuming they all have mesocyclones). How would you compare GR2 in that respect?
The storm motion used to calculate the storm-relative velocity displayed in GR3 is the average motion of all identified storms from the previous volume scan. It's calculated at the radar product generator at the NWS/DOD site. Overall, it may not be too bad, but there are instances where storms whose motions deviate significantly from the mean will have rotational signatures that do not stand out visually because of this.
When wind fields are strong, supercells may not deviate much from the mean cell motion. When wind fields are weaker, they can. So the mean could work okay in many strong wind field situations, and less well (generally) otherwise. Another situation where the mean motion does not work well is with an upper low that has convection wrapped around it. The mean motion in such a case will often not be anywhere near the motion of the storms you may be interested in investigating.
This is all made even more complicated by where the storms are relative to the radar for a given storm motion. For example, if your storms A B and C are all to the northwest of the radar, the circulations would likely all be pretty evident, and the couplets would appear relatively symmetric, if the mean or manually derived motion was anywhere near the motion of any of those storms. On the other hand, if those same storms were northeast of the radar moving away, the same average or manually tracked motion could result in one or more of the storms' circulations looking "unbalanced".
If the bandwidth was there, I'd always choose GR2 type displays. The occasional extra work of tracking storms would be more than offset by the higher resolution, and the ability to better view the rotation in storms with significantly deviant motion.
When wind fields are strong, supercells may not deviate much from the mean cell motion. When wind fields are weaker, they can. So the mean could work okay in many strong wind field situations, and less well (generally) otherwise. Another situation where the mean motion does not work well is with an upper low that has convection wrapped around it. The mean motion in such a case will often not be anywhere near the motion of the storms you may be interested in investigating.
This is all made even more complicated by where the storms are relative to the radar for a given storm motion. For example, if your storms A B and C are all to the northwest of the radar, the circulations would likely all be pretty evident, and the couplets would appear relatively symmetric, if the mean or manually derived motion was anywhere near the motion of any of those storms. On the other hand, if those same storms were northeast of the radar moving away, the same average or manually tracked motion could result in one or more of the storms' circulations looking "unbalanced".
If the bandwidth was there, I'd always choose GR2 type displays. The occasional extra work of tracking storms would be more than offset by the higher resolution, and the ability to better view the rotation in storms with significantly deviant motion.
rdale
EF5
1) Maybe every half-hour unless I see something that is really deviating from the norm. THose small differences really aren't that big of a deal. If it's a weak storm, then it won't matter because the storm is weak. If it's a strong storm, 100kts of GtG is still going to look like 100kts of GtG even if the exact -/+ are offset.
2) Again related to storm strength. If they are moving in a very different fashion, they probably aren't that strong of storms. But the differences aren't going to matter. It might show +32/-5 instead of +17/-20. You still see the rotation.
2) Again related to storm strength. If they are moving in a very different fashion, they probably aren't that strong of storms. But the differences aren't going to matter. It might show +32/-5 instead of +17/-20. You still see the rotation.
Bob Hartig
EF5
Thanks, Rob and GPhillips. That's helpful input from both of you.
Jason Blum
EF0
I own both, and would gladly choose 2AE if it weren't for the fact that my data connections are rarely solid or fast enough to reliably update my polling. I do much better with GRLevel 3 in the field, and stick to 2AE for home-use.
Tyler Allison
EF5
All things being equal...sure GRAE is better. But that's the rub...all things are not equal and I call BS on anyone who says they use GRAE in the field on a routine basis. (I haven't seen anyone say that...yet) At least not until they get 3G or better more than 10 miles outside cities. Heck, down town Chicago you can't even get 3G on a reliable basis.
Bill Schintler
EF4
GR3 or GR2analyst for active chases
In 2008 I had continuous Level-2 Analyst coverage in the field at all times for 3 weeks of chasing in TX/OK/KS/CO/NE - except for southeast OK in the hills... Of course, that was before the resolution (and subsequent data bandwidth) increase. Depending on signal strength, I can still stream high-rez data but don't have much bandwidth left over for anything else - bill
In 2008 I had continuous Level-2 Analyst coverage in the field at all times for 3 weeks of chasing in TX/OK/KS/CO/NE - except for southeast OK in the hills... Of course, that was before the resolution (and subsequent data bandwidth) increase. Depending on signal strength, I can still stream high-rez data but don't have much bandwidth left over for anything else - bill
As Bill said, SuperRes killed using AE in field for me. I used it in 2007 a decent amount, but as soon as the local radars went SuperRes, that killed doing it while mobile. Not complaining, love the SuperRes, but it put it out of reach for chasing.
Jason Blum
EF0
I also love using GRLevel2:AE to reconstruct a chase afterward, using archived data. Last year we were near a very HP storm which produced some debris and damage to roofs along a pretty suspicious path. We suspected it had dropped a tornado, but there were never any warnings issued (it was in a pretty isolated area) and we never saw anything conclusive because of the torrential rain. So the next day it was nice being able to load the archived data in and analyze everything with GRLevel2:AE.
