VIL, VIL density and hail size

[Jason Boggs]

Obviously, VIL and VIL density play somewhat of a major role in determining hail size in a given storm. I have learned that VIL density varies from day to day on determining hail size. IS there a VIL or VILD value that you prefer when looking for various hail sizes.

My real question is this. When looking at a storm on Grlevel3, what do you look for in determining hail size? I have also learned that sometimes Grlevel3 grossly overdoes hail size.

I'm just looking for different ways to better determine hail sizes in various storms. Here is a radar grab to get a better feel for what I'm talking about.

 

[Aaron Wesson]

VILD values of 3.5 g m-3 and values greater than 4.0 g m-3, have been proven to produce large hail.

This is a very good PDF file
http://www.crh.noaa.gov/lmk/soo/docu/Vil_density.pdf

As for VIL
Instead of writting what has already been posted check these links out.

http://www.theweatherprediction.com/habyhints/249/

http://www.wrh.noaa.gov/wrh/99TAs/9902/index.html

 

[rdale]

First note that VIL _removes_ any hail from the equations, so if you are using Level III (NIDS) data - don't correlate VIL to hail.

VILD was thought to be a nice tool, but it ended up not working either.

There simply is no connection between VIL (or VILD) and hail size. Bigger numbers are likely to produce larger hailstones, but that's about it. The MEHS product from AE, using NSSL algorithms, is quite reliable however.

- Rob

 

[Van DeWald]

VIL is short for vertically integrated liquid, essentially the amount of liquid (not ice) in the column (based on reflectivity). But, one can roughly correlate a VIL of the day with severe hail. In other words, once you get a severe hail report, you can cross reference that with the current VIL and use that as a rough predictor of additional severe hail reports. For example, a 3/4 hail report had a VIL of 65. It would be safe to assume that a similar storm with another VIL of 65 may also be producing severe hail that day. In fact, back in the 90s, it was one of the top predictors used, and "VIL of the day" programs were used to get the current day's value based on freezing level and local climatology, etc. Digital VIL is also useful for finding the core of the storm when looking for storm verification, etc. Often, after a storm moves though an area and we haven't received storm verification, we can use the DVIL to pinpoint a very small area (often a couple square miles) to call citizens to inquire of hail size.

More recently, mets are correlating 50dbz heights to the freezing level and the -20C level to predict severe hail which actually works quite well. Several studies have been done which connect a given freezing level with expected severe hail sizes, etc.

VIL Density is just a correlation of VIL to the storm echo top. It basically boils down to this, if you have a high VIL but a low echo top, that results in a higher VIL Density and a greater threat for hail. Or, if you have a low VIL with a high echo top results in a low VIL Density and a much less chance for hail. Studies have shown that VILD values of 4 or greater are usually associated with severe hail, but from my experience, it can vary quite a bit from site to site. Is it a slam dunk that you get severe hail with values of 4 or higher? No, but it's just a tool like anything else.

As far as the hail size that you see in GR3, that's just the NSSL hail algorithm output. I take it with a grain of salt. Again, back in the day, it was about all we had for guidance for predicting hail size. But, again, with recent science advancements by correlating 50dbz heights with the freezing and -20C level, and also by calculating storm top divergence comes up with more accurate hail size forecast.

One other predictor for hail size is wet bulb zero value. Heights of 7000-10000 feet are ideal for large hail. When wet bulb zero heights get to 12000-14000, it results in a large amount of melting of the hailstones before reaching the ground. You might have an algorithm output of 2" hail, but only 3/4" stones might be reaching the ground due to the longer available melting time.

 

[Kiel Ortega]

As far as the hail size that you see in GR3, that's just the NSSL hail algorithm output. I take it with a grain of salt. Again, back in the day, it was about all we had for guidance for predicting hail size. But, again, with recent science advancements by correlating 50dbz heights with the freezing and -20C level, and also by calculating storm top divergence comes up with more accurate hail size forecast.

Actually, the -20C level and 0C and reflectivity heights is what the Hail Detection Algorithm does...just a little more complicated (though, there is no difference at the basic level of the methods except one is manual and the other is an algorithm). I think the biggest problem that the algorithms (both MESH and VIL) face are sampling issues and the method used to derive the number. If you use a cell-based method, you end up with highest reflectivity through the whole column which can give you incredible high value. A grid based method has some different problems, but you end up with more moderate values (see http://ams.confex.com/ams/pdfpapers/104885.pdf--some research I did as an undergraduate).

As far as VIL and VIL density, you might check out (though locked away from those without AMS access ):
Roger Edwards and Rich Thompson "Nationwide Comparisons of Hail Size with WSR-88D Vertically Integrated Liquid Water and Derived Thermodynamic Sounding Data"
http://ams.allenpress.com/perlserv/....1175/1520-0434(1998)013<0277:NCOHSW>2.0.CO;2
Steven Amburn and Peter Wolf "VIL Density as a Hail Indicator"
http://ams.allenpress.com/perlserv/....1175/1520-0434(1997)012<0473:VDAAHI>2.0.CO;2

 

[Dean Baron]

not sure how accurate this is but it seems to work alright for me most of the time. i look for VIL of atleast 45 and VILD of atleast 5.00. VIL can be extremely high but produce only marginally large hail depending on the amount of rain in the storm that can mess with the VIL values, but i noticed that when the VILD gets closer to 6.0 the hail size generally increases and anything over about 6.2 is about golf ball size hail. This is just for the central US, this all changes depending on which part of the country you're looking at. I also try to look use echo tops when i have an idea what the average echo top is for the day (depending on the setup).

 

[rdale]

Those numbers depend GREATLY on the temperatures aloft... Echo tops really tell you nothing regarding storm severity. I wouldn't be too comfortable making that blanket of a statement.

 

[Bobby Prentice]

<snip>
My real question is this. When looking at a storm on Grlevel3, what do you look for in determining hail size? I have also learned that sometimes Grlevel3 grossly overdoes hail size.

I'm just looking for different ways to better determine hail sizes in various storms.

The responses given so far have been quite good.

I will add that the next big national improvement in radar technology (including hail detection) will occur when the existing WSR-88D (NEXRAD) weather radar network is upgraded to dual-polarization capability beginning in 2009. The WSR-88D hail algorithm should show significant improvement:

http://www.nssl.noaa.gov/divisions/radar/dualpol/
http://www.nssl.noaa.gov/divisions/radar/dualpol/hazards.php

Polarimetric Doppler Radar

IDENTIFYING PROBLEMS AND FINDING SOLUTIONS:

Detecting Weather Hazards

PROBLEM: HAIL - Where is hail falling, how big is it, and where has it been? Conventional radar can only identify the probability of hail at any particular location within the storm.

SOLUTION: Polarimetric radar can identify the specific location of hail, estimate hail size, and track hail swaths.

SUCCESSES: Validation of polarimetric hail detection (Heinselman and Ryzhkov): NSSL scientists recently showed that a simplified version of NSSL's polarimetric hydrometeor classification algorithm (HCA) outperforms the NEXRAD Hail Detection Algorithm (HDA) currently in use by the NWS. They compared rain and hail reports collected by storm-intercept teams (mostly during the Joint Polarimetric Experiment) with data provided by the HDA and HCA to verify the accuracy of the algorithms. Scientists found the HCA had significantly fewer false alarms. This work was done in anticipation and support of the polarimetric upgrade to the WSR-88D network.

Hail size discrimination (Heinselman) Dual polarization radar not only supplies information for hydrometeor classification, but may also prove useful for gauging hail size within the storm. Past research indicates relationships between polarimetric variables that might allow for hail size to be categorized. Investigating the relation of polarimetric variables to hail size can be beneficial in improving warnings for hail producing storms, understanding the physical processes that lead to hail formation, and determining possible early signatures associated with hail formation and growth. NSSL scientists found hail size was best distinguished using a pairing of Z DR (a logarithmic ratio of the returned horizontal and vertical power) and a correlation coefficient (a statistical number that indicates the degree of similarity among particles within a cloud).

WHAT'S NEXT: Detailed microphysical information (Brandes, Schuur, Ryzhkov, Zhang, Ikeda) Research by NSSL and NCAR scientists showed that polarimetric radar provides detailed microphysical information that could be helpful in cloud models

</snip>

 

[rdale]

Bobby - any insight on what products will be available outside the NWS office? Still getting mixed messages from HQ...

 

[Bobby Prentice]

Bobby - any insight on what products will be available outside the NWS office? Still getting mixed messages from HQ...

No. Hopefully Greg Stumpf will reply since he's involved with the process.