BIS: 3D-VAR Struggling at Long-Range from Radars

As would be expected, 3D-VAR struggles at greater distances from the radar.  A fairly large supercell thunderstorm that had a history of producing brief tornadoes was at 12-15kft AGL on BIS 88-D’s lowest slice.  It could be implied from base radar products that this storm had a powerful updraft, modest to strong mid-level rotation and strong storm-top divergence.  However, since the storm was too far from the radar, 3D-VAR products such as max updraft composite, updraft helicity, updraft vorticity and max convergence above 8km appeared much weaker than what was likely reality.

KBIS 5/22 2320 UTC 0.5 deg base reflectivity (upper left), max divergence above 8km (upper right), max updraft composite (lower right), 5km updraft helicity (lower left)

WFO-ABR Max Updraft and Vorticity Composites

Several severe hail-producing supercells formed early this evening across northern South Dakota.  Reports of up to golfball size hail were received.  Here, we will focus on some 3DVAR products.

-10C Isothermal Reflectivity (top left), -20C (top right), Composite Updraft Maximum (lower left) and Composite Maximum Vorticity (lower right).

We want to draw your attention to the lower panels (the image was at 00Z).  The panel on the left clearly shows three distinct updrafts associated with each of the supercells.  The panel on the right shows the magnitude of the vorticity associated with each updraft very distinctly.  Let’s look at the next image, 0015Z:

Same as above image, for 0015Z.

Look at the bottom panels again.  You can clearly see the middle and southern storms have merged into one homogenous cell, especially in the updraft image on the left.–Gordon Strassberg for WFO ABR.

3DVAR Looking Interesting in JAX & MLB Areas

An outflow boundary from earlier coastal convection is intersecting instability in north-central Florida, leading to new development in the far southern portion of the JAX CWA and far northern portion of the MLB CWA. Updraft strength, updraft helicity, and even 10km divergence (not in this image) have all become very interesting across Marion, Lake, and Volusia Counties.

3DVAR Simulated Reflectivity, Composite Updraft Maximum, Updraft Helicity, and 10km Divergence valid 2012-05-17 2210 UTC
3DVAR Simulated Reflectivity, Composite Updraft Maximum, Updraft Helicity, and 10km Divergence valid 2012-05-17 2210 UTC

These data have prompted the issuance of SVRs for parts of Marion and Volusia county, and special marine warnings have been issued off the coast.  The JAX team has split into half-JAX, half-MLB to address the evolving threats.

3DVAR Captures JAX-area Outflow

Another severe thunderstorm warning was issued near the Florida-Georgia border based on strong dual-pol radar and MRMS signatures.  3DVAR hasn’t been as interesting (perhaps the storms are too close to the KJAX radar), but the simulated reflectivity product is picking up on an impressive outflow boundary crossing the state line.

3DVAR Simulated Reflectivity valid 2012-05-17 2110 UTC and KJAX 0.5-Degree Reflectivity valid 2012-05-17 2108 UTC
3DVAR Simulated Reflectivity valid 2012-05-17 2110 UTC and KJAX 0.5-Degree Reflectivity valid 2012-05-17 2108 UTC

First Warning of the Day for JAX

The first severe thunderstorm warning for the day for WFO JAX comes along the JAX-CHS CWA line in Glynn County, GA.  The 3DVAR analysis indicates some of the strongest indicators seen this week.

3DVAR Analysis of Simulated Reflectivity and 1km U/V Wind, Updraft Composite, Updraft Helicity, and 10km Divergence valid 2012-05-17 1905 UTC
3DVAR Analysis of Simulated Reflectivity and 1km U/V Wind, Updraft Composite, Updraft Helicity, and 10km Divergence valid 2012-05-17 1905 UTC

The 1-km U/V winds indicate 30-40 knot winds (perhaps a rear-inflow jet) nearing the coast, 19 m/s composite updraft strength, 159 m/s^2, and 10km divergence of 10.72 s^1.  These values (particularly the updraft) are among the stronger storms we’ve seen this week.  After some AWIPS 2 problems, a severe thunderstorm warning was issued after confirming these data with the KJAX radar.

Max Updraft preceding Max Reflecitity

Late in the afternoon, as convection continued along the southern fringe of BTV’s county warning area, a storm cell began pulsing up and down.  This series of images shows how the max reflectivity on radar and the maximum updraft output from 3D-VAR began oscillating.  Maximum updraft strength from 3D-VAR would precede a peak in radar reflectivity by about 20 minutes.  We issued a warning for a southern BTV county, based on the Max Updraft data.  The storm produced 3/4 inch hail.

2205z.  Reflectivity is maximizing from a previous pulse, and current Max updraft values are less than 10 m/s.

2235z. Maximum Updraft values around 14 m/s. Max Reflectivity around 56 dbz.

2300z. Max Reflectivity back up to 60 dbz, while Max Updraft Strength continues around 18 m/s.

Dankers/Kearney

Miscellaneous musings from ALY

Had an interesting day working with convection in Albany’s CWA today.  Several storms grew big enough to make severe hail, strong winds and show nice organized rotation..also a bow echo looking storm.  In general the 3D Var and MRMS products proved useful for detecting hail concerns and also evaluating whether or not updrafts were strengthening or weakening…how high they were and whether or not there was good updraft helicity/vorticity.  Also the azimuthal shear product from MRMS was useful/interesting.  The nearcast thetha e product had its scale and units reversed since yesterday which caused some initial confusion but did seem to depict the more unstable regions well…though instability was fairly marginal.

Some product limitations we ran into:  With all the new fancy products we had, there really wasn’t anything to aid with max wind gusts.  With the bow echo storm the 3d var fields were barely perturbed and did not at all seem representative of the probable wind speeds (images below show reflectivity and 3d var quad at same time with 1km (lowest available) winds…higher levels didn’t show much either)…old school use of radar velocities worked much better.

Some other products that would really be nice to have would be a 3-d or cross section way to look at updraft and downdraft strength, the horizontal plots of this could be overlayed at 1 km spacing in 3 d-var and was useful to watch through the life cycle of the storms but was very clunky to step through and hard to mentally visualize. In a 3d view this could be a very powerful way for forecasters to monitor and diagnose storm structure.  Another thing we ran into today were training storms and flash flood concerns…dual pol was handy for this among other things but wonder about what applications of the fancy new products there might be to hydro concerns.

AWIPS 2/Warngen/CAVE issues continued again today with warnings that wouldn’t go out, multiple Cave crashes, lots of random error messages and slow downs.  This system is not at all ready for operational use as they slowed/prevented warning issuances and having Cave crash mid warning ops when all the work station is running are 2 Caves not at all confidence inspiring when you think about everything else AWIPS in WFOs runs and does in addition.  Also found an interesting bug with the map background in the quad panels where they don’t always stay linked and you can unpredictably end up with map scales and alignments in different sections of the quads.

Another SVR in eastern NY

Another storm is affecting eastern Saratoga County and southern Washington County, NY (prompting a severe thunderstorm warning).  3DVAR storm-top divergence has been a pretty solid predictor with other storms in the region today, and this storm was no exception.  However, this was the first storm to show more than marginal updraft helicity, maxing out at 117 m/s^2 at 2045 UTC.

3DVAR Reflectivity, Updraft Composite, Updraft Helicity, and 10km Divergence valid 2012-05-16 2050 UTC
3DVAR Reflectivity, Updraft Composite, Updraft Helicity, and 10km Divergence valid 2012-05-16 2050 UTC

About 15 minutes later, MRMS MESH reached 1.25″ with a 62% POSH; that has since increased to 72% POSH and 1.5″ MESH.

BTV: 3DVAR/MRMS usefulness

A short line of storms developed rapidly over Essex County.  The MRMS data (lower right pane) indicated a MESH of nearly 2 inches!  At the same time, the 3DVAR vertical velocity data (upper right pane) was 14 m/s for the same cell.  Using this data against already high reflectivity (upper left) increased confidence for a warning, especially given the environment.  Unfortunately, the 1KM wind field (lower left) gave a mixed signal at the same time frame, which was possibly an effect of the terrain.

Monitoring More ALY-area Storms

Two more storms in extreme northern Saratoga County, NY are being monitored for warnings (update: an SVR was just issued for the northern storm).  3DVAR has shown the same signals as the Herkimer County storm with strong storm-top divergence:

3DVAR 10km Divergence valid 2012-05-16 1930 UTC
3DVAR 10km Divergence valid 2012-05-16 1930 UTC
3DVAR 10km Divergence valid 2012-05-16 1935 UTC
3DVAR 10km Divergence valid 2012-05-16 1935 UTC

MESH was peaking just under 1″ with near-50% POSH.  (Update: it has increased above 1″ and POSH is above 50% now, prompting the warning.)

The UW CTC was also strongly negative in eastern Hamilton County at 1925 UTC, indicating a CTC of less than -30 C/15 min, so northern Warren County will need to be monitored.

Visible Satellite, UW CTC, and UAH CI valid 2012-05-16
Visible Satellite, UW CTC, and UAH CI valid 2012-05-16 1925 UTC