Radial Velocity question

[Joel Wright]

Here's a question for the radar experts around here. This image is from June 28, 1998 when a strong mesovortex passed right over the town of Erie, IL causing extensive tree damage. (This is several years before I moved here). I had always heard about how bad this storm was, and decided to look it up in the archived data.

Anyway, as you watch the loop, you can see the green colors (in bound) show up right in the center of the red (out bound). What would cause this? Wouldn't this imply the wind was blowing towards the radar, from the east? This doesn't look right at all. Wind should have been blowing from the west or northwest on the south side of the mesovortex. There shouldn't be any green (in bound) winds in this area. You can see some green further north where you'd expect to see it, northeast of the mesovortex. Is this because the velocities are above the 60kt max on the legend? That's the only thing I can think of that would explain it. I doubt wind is moving perpendicular to the radar beam (either north or south) due to the location of the green just south of the mesovortex.
Anyway, just kind of curious. I don't recall seeing in bound velocities showing up in the center of outbound velocities before. Any help is appreciated.

 

[IAN GIAMMANCO]

That actually looks like folded or aliased velocity data... the 88-D depending on its selected scanning strategy can only sample velocities up to a certain value (i.e. Nyquist Velocity, function of the PRF and radar characteristics). However obviously there can be higher values, so basically what happens is the values "fold" over onto the other side of the scale. So the higher Doppler velocities are now depicted as flow toward the radar whne they really are still away but at a higher value than the radar can process. I'm sure others have some good links to radar characteristics and the basic principles of Doppler radars... this is an often occurence during hurricanes.

 

[Chris Nuttall]

That actually looks like folded or aliased velocity data... the 88-D depending on its selected scanning strategy can only sample velocities up to a certain value (i.e. Nyquist Velocity, function of the PRF and radar characteristics). However obviously there can be higher values, so basically what happens is the values "fold" over onto the other side of the scale. So the higher Doppler velocities are now depicted as flow toward the radar whne they really are still away but at a higher value than the radar can process. I'm sure others have some good links to radar characteristics and the basic principles of Doppler radars... this is an often occurence during hurricanes.

That's what I think it probably is, especially with the way the very high inbounds are embedded inside a broad region of high outbounds. If I'm reading the loop right, the radar is to the WSW, and that velocity structure does not really look like any kind of rotation. If that's occurring along the radial, it could be a convergent signature. But, I think it's simply folded, especially if you follow the feature through the whole loop. It looks like it happens at the beginning with high outbounds embedded in high inbounds.

Before I committed to anything though, I'd want to see the reflectivity, so I know how that matches up and get a better picture of the storm structure.

 

[Joel Wright]

Damn it. I should have posted the reflectivity along with it. My mistake. Sorry guys. Anyway, here's the 0.5 degree reflectivity...

 

[Robert Dewey]

The radial velocity at >07:19 does show some rotation just north of the folded velocity values, which coincides with the reflectivity. Looks like a supercell embedded within a squall line... and I'm guessing the storm-relative velocity may have shown a stronger couplet.

 

[Greg Blumberg]

I agree with Robert about seeing the rotation in the couplet. The reflectivity matches up pretty well with the velocity and there is a definite hook echo/inflow notch I'm seeing. Determining for sure, will require the storm-relative velocity. That product is favored for mesocyclone detection as compared to the radial velocity. You've got to look at how the storm looks when you subtract the storm motion out of the velocity data.

 

[Jason McClintock]

I agree with Robert about seeing the rotation in the couplet. The reflectivity matches up pretty well with the velocity and there is a definite hook echo/inflow notch I'm seeing. Determining for sure, will require the storm-relative velocity. That product is favored for mesocyclone detection as compared to the radial velocity. You've got to look at how the storm looks when you subtract the storm motion out of the velocity data.

Yes you need storm relative. But I say its an embedded super with aliased velocity causing the inbound to show up in that broad area of outbound.

 

[Glenn Rivers]

After looking at both of the images, I am convinced that there is range folding. If one looks at the transition from red to green, the highest outbound velocities enclose the green patch, with has the highest inbound velocities on the outer edge, decreasing inwards. The "bulls-eye” of the green patch is depicted as having the lowest inbound velocities, which seems unphysical.

For the sake of continuity, if one reads the outermost band of green in the patch as the first range folded outbound contour and the next band as the next range folded outbound contour, then one gets a physically consistent and spatially coherent velocity field with some phenomenally high outbound values of up to 90 Knots.

This is still a very stong meso, considering that outbound velocities of up to 90 Knots are just south of velocities that are very light or even slightly inbound. There is probably 80-100 Knots of shear in this velocity cuplet gate to gate. Very impressive.

I am not sure that this meso looks like a "bookend" vortex found on the North end of a bow-echo. If the velocity cuplet remains so pronounced at mid-levels, this vortex may indeed result from this apparent embedded supercell!

What kind of damage did this wicked storm do!!??

P.S. I hope this helps.

 

[Joel Wright]

First off, thanks everyone for some great information! I had never heard of range folding before. Very interesting.

Below is a slightly longer loop. You can see a fairly persistant hook-like feature that follows the path of an outflow boundary. I checked the archives and there was never a tornado reported with this feature. Just a 78kt wind report from Erie.

I am going to try to dig up the storm-relative velocity scans from this storm and see how it compares. The archive site is running very slow this morning though, so it might be awhile.

BTW, these radar archives are from this website.. http://www.ncdc.noaa.gov/nexradinv/
I'm sure many of you know about this site, but for those who don't it's a very useful tool for looking up past events. The database goes back to when the 88Ds first became operational.

Event: Tstm Wind
Begin Date: 28 Jun 1998, 01:30:00 AM CST
Begin Location: Erie
Begin LAT/LON: 41°39'N / 90°05'W
End Date: 28 Jun 1998, 01:30:00 AM CST
End Location: Erie
End LAT/LON: 41°39'N / 90°05'W
Magnitude: 78
Fatalities: 0
Injuries: 0
Property Damage: $ 0.0
Crop Damage: $ 0.0

 

[Danny Neal]

It is also very interesting to see how this embedded supercell is following right along the outflow boundary sagging south out of Northern Illinois until about the I 80 corridor..... If I recall this date correctly (I was 10 years old) I believe there were also impressive wind damage reports through the west and southwest sides of the Chicago area (Orland Park, Tinley Park, Mokena, New Lenox, Downers Grove, etc areas) Coincidentally it looks like that is the very area the OFB may have stalled out. Very cool. Thanks for posting this Joel.

 

[Jeff Duda]

This is a case of range folding, but there is also a definite embedded supercell in there. I bet the 78kt wind is from the mesocyclone basically making it to the ground.

 

[Jeff Snyder]

First off, thanks everyone for some great information! I had never heard of range folding before. Very interesting.

Just a minor correction -- the issue here is velocity aliasing, not range folding. Range folding effects show up as second trip (or third trip for some radars). As noted above, if the actual radial velocity exceeds the maximum unambiguous velocity, the calculated radial velocity will "wrap around" for one extreme (i.e. max outbound) to the opposite extreme (i.e. maximum inbound). As such, a classic sign of folded velocity is to see max inbounds embedded in an area of max outbounds (particularly along a radial). If you use GR2, you can try to unfold the velocity so that it displays correctly.

FWIW, range folding occurs when a strong echo lies beyond the maximum unambiguous range. In such a case, the echo is displayed near the radar (at a distance of range_actual - range_unambiguous). The so-called "Doppler Dilemma" addresses the battle between a high maximum unambiguous velocity and a high maximum unambiguous range; increasing one necessarily decreases the other. There are "workarounds" to this, fortunately -- for example, one can utilize a dual-PRF/dual-PRT scanning strategy to increase the maximum unambiguous velocity without decreasing the maximum unambiguous range.

 

[Joel Wright]

I tried to get the 0.5 degree storm-relative velocity information, but apparently it isn't available for this particular period.

I was able to get the base velocity data. It looks like it more accurately shows the correct velocity distribution.

 

[Jeff Duda]

If you have GR2 or GR2AE you can get SRM from the data. I usually go to the ncdc site and request a few hours.

 

[Glenn Rivers]

Correction about range folding

I posted here earlier today, sugesting that the problem was range folding. I was wrong. I forum member explained to me that the problem is a "velocity aliasing" error. Still, what an impressive velocity signature!