Memorial Day 1995 in Great Barrington

May 29, 1995 was Memorial Day.

Brand new “Doppler” technology was just reaching many areas of the country. The 1988 Doppler Weather Surveillance Radar (WSR-88D) was the new radar being deployed across the nation. We still use the WSR-88D today, and even though the radars still bear the number “88,” all National Weather Service 88D’s have received substantial hardware and software upgrades since original installation. The most notable and recent upgrade, Dual Polarization, was discussed in my previous post about the SoCal wildfires earlier in May.

r1

KOKX Upton, NY reflectivity image from 7:01pm EST on May 29, 1995.

A major tornado hit western Massachusetts, particularly Great Barrington, shortly after the KOKX Upton, NY Doppler came online. Great Barrington is a quaint town nestled in the southern Berkshires. Every time I drive through town I think about the once-popular “Great Barrington Fair,” a now-defunct and heavily dilapidated horse racing track. The New York Times has a good article on Massachusetts horse racing, originally published way back in 2001. Anyways, long story short, horse racing was already struggling in Great Barrington long before the tornado hit. Then, on Memorial Day 1995, mother nature sucked any remaining life out of Great Barrington racing in a matter of minutes.

The National Weather Service storm survey team determined that an F4 tornado touched down at 7:06pm EST just west of the center of Great Barrington. It was on the ground for 11 miles, and had a maximum width of 300 yards (almost 2-tenths of a mile). 3 were killed and at least 24 were injured.

5-frame loop of base reflectivity from KOKX Upton, NY.

5-frame loop of reflectivity from KOKX Upton, NY. Loop begins at 6:56pm EST and ends at 7:16pm EST.

The animation above is from the then-newly installed radar on Long Island (more precisely, Upton, NY). No textbook “hook echo” was visible with this storm. So how did meteorologists identify a possible tornado? Before the 88D was installed, it would have been nearly impossible to the inexperienced (and maybe experienced?) eye. But the new 88D brought with it technology that allows meteorologists to analyze motion in the atmosphere in near-real-time. The next animation contains radial velocity data.

Remember that the Doppler is on Long Island. Green color indicates motion towards the radar (generally to the south) and red color indicates motion away from the radar (generally to the north). When the green is back-to-back with the red, that spells trouble – a strong indication of rotation. Remember play dough? If you can imaging rolling play dough back and forth in your palms, you’ll better understand the idea of rotation in this sense. In this case, the velocity indication was strong that low-level rotation was occurring within the thunderstorm. This kind of indication warrants tornado warnings and wall-to-wall TV coverage nowadays.

5-frame loop of radial velocity from KOKX Upton, NY. Loop begins at 6:56pm EST and ends at 7:16pm EST.

5-frame loop of radial velocity from KOKX Upton, NY. Loop begins at 6:56pm EST and ends at 7:16pm EST.

The Great Barrington Fair suffered major damage. While the damaged facilities at the race track were actually rebuilt, horse racing never returned in the form that is was during its glory days in Great Barrington (for reasons mostly unrelated to the tornado, as described in the NYT article above). Still, the tornado further impeded any forward progress in reviving the Great Barrington Fair to what it was in its heyday.

Back in late 2010, Meteorologist Ryan Hanrahan dug up and uploaded some aerial and ground footage from the aftermath. You can watch the WVIT video here.

Editor’s note: The KENX Albany, NY radar is much closer to western Massachussetts than the KOKX Upton, NY radar on Long Island, but the data from KENX is missing in the National Climatic Data Center archive that I used for this report. If KENX was in fact working at the time of the storm, forecasters would have relied on it more than any other radar. KENX provides a better idea of what is going on over Great Barrington because the beam scans at a much lower altitude than the beam from the Long Island radar.

Southern California Wildfires on Radar

Doppler radar is a tremendously powerful tool in modern-day weather analysis and forecasting. Recently, the National Weather Service completed the upgrade to dual-polarization on 122 of it’s radars across the country. This new technology allows meteorologists to get a better feel for what the radar is ‘seeing’ as it scans our skies.

Santa Ana winds. Source: UC San Diego.

Santa Ana winds. Source: UC San Diego.

Massive Santa Ana-fueled fires are burning in Ventura County, California. The defining characteristic of Santa Ana winds is dry, moisture-starved air. These winds blow into SoCal from the Great Basin and fuel the flames.

The SoCal wildfires are being monitored and fought by crews in the air and on the ground. But weather satellites and radars can even pick up on what is going on!

The plumes of smoke from the fires rise thousands of feet into the air.

The most interesting images to me are those from the KVTX Los Angeles Doppler radar. The actual radar is situated about 25 miles to the northwest of the fire site. The beam is scanning at about 1,500 feet above the ground over the fire – in other words, right through the smoke plume.

The first radar image is a reflectivity scan. This is simply a measure of how much energy is reflected back to the radar dish. The radar isn’t smart enough to know that it’s beam is hitting smoke particles and not rain. Don’t focus on the random blue specs – that is neither smoke nor rain. Rather, keep your eyes on the area just off the coast where the brighter greens and yellows are located.

KVTX Los Angeles Reflectivity scan from 11:50pm May 3, 2013.

KVTX Los Angeles reflectivity scan from 11:50pm EDT May 2, 2013.

The second radar image is the velocity scan. This is frequently used to assist in issuing Tornado Warnings because the colors indicate which way the winds are blowing – either towards or away from the radar site. In this case, however, the red color just off the coast indicates that the smoke plume was moving in a general direction away from the radar. By cross-examining the reflectivity image, I am able to determine that the smoke plume was moving to the southwest.

KVTX Los Angeles velocity scan from 11:51pm EDT May 2, 2013.

KVTX Los Angeles velocity scan from 11:51pm EDT May 2, 2013.

The third radar image is a new product called correlation coefficient – something that wasn’t available before the dual-polarization upgrade. This product allows meteorologists to gauge the similarity of the vertical and horizontal returned-energy in a small volume of air. Wait, what? Keep reading…

Essentially, pure rain shows up as being very similar because all the radar energy hits is rain drops. When the radar beam hits smoke, as in this case, the “cc” values are very low (indicated by the grey colors just off the coast). After all, a plume of smoke contains all sorts of particles. This provides solid confirmation that the reflectivity (greens and yellows) in the first radar image is smoke and not rain.

KVTX Los Angeles correlation coefficient scan from 11:51pm EDT May 2, 2013.

KVTX Los Angeles correlation coefficient scan from 11:50pm EDT May 2, 2013.

The image below is from KNBC-TV NewsChopper 4. We’re lucky that we rarely, if ever, see wildfires of this magnitude in New England.

Ventura County, CA wildfire. Source: KNBC-TV NewsChopper 4

Ventura County, CA wildfire. Source: KNBC-TV NewsChopper 4