The Hillsdale, NY and Great Barrington, MA Tornadoes
May 29, 1995

The Northeast is not especially prone to large scale destructive severe weather outbreaks due in large part to geography.  However, on occasion, large violent thunderstorms do develop and go on to produce damaging straight line winds, large damaging hail, and in rare instances, large damaging tornadoes. For example, on June 9, 1953 and extremely violent tornado leveled parts of Worcester, MA killing 94, injuring over 1000, and leaving 10,000 people homeless.  On August 28, 1973, a large tornado hit the town of West Stockbridge, MA killing four and injuring forty people.  More recently, on May 12, 1984 a moderate tornado injured eleven people on the Altamont fairgrounds and one in Schenectady, NY as well as damaging property.  And on November 16, 1989, a tornado, which had dissipated into a straight line wind gust, crumbled a wall at the East Coldenham Elementary School in Montgomery, NY killing seven children. Even with the historical occurrences of damaging tornadoes in the Northeast, many residents of the southeast Columbia county, NY, and in Egremont, Great Barrington, and Monterey, Massachusetts in Berkshire county were stunned, when shortly after 6:30pm on the evening of May 29, 1995 two powerful tornadoes, spawned from one rotating thunderstorm, struck.

Meteorological Aspects of the May 29, 1995 Tornado Event

Synoptic weather charts showed a warm front slowly moving northward across New York and southern New England. The warm front produced a cloudy, cool, and showery start to the Memorial day holiday across the region. Sunshine, however, began burning through the low overcast during the late morning setting the stage for later events. A powerful cold front, extending from low pressure located over Ontario, was plowing eastward into the increasingly warm and humid air over western New York. Radars displayed a broken line of showers and thunderstorms developing in the moist air mass over central New York. These thunderstorms were the beginnings of what would develop into an evening weather disaster further east.

Dewpoint temperatures steadily edged up through the sixties and temperatures climbed into the seventies through the morning and early afternoon providing the necessary moisture and heat that thunderstorms need in order to grow. Using temperature, moisture, and wind information from the National Weather Service balloon launches at 8:00 am, regional meteorologists, including myself, analyzed soundings of upper level atmospheric conditions for the Northeast to determine if weather parameters aloft were as favorable for severe thunderstorms as the surface parameters appeared to be.

The upper air profiles revealed several very interesting facts. A surface air temperature of 73 degrees F and dewpoint temperature of 68 degrees F would be sufficient to trigger large thunderstorms with no aid from any other atmospheric parameters. That fact alone was very impressive. The profiles also indicated that a tremendous amount of wind shear, especially along the advancing warm front, was present. For example, wind velocities increased from 12 knots at the surface to 36 knots at 7000 feet and the directional change in the winds measured greater than 60 degrees. Those numbers indicate strong directional and speed shear, parameters quite favorable for supporting severe weather.

It was apparent to me that there would be sufficient heat and humidity to spark thunderstorms as sunshine continued to increase over New York and New England sending temperatures soaring beyond the needed 73 degree F critical temperature. Southeasterly surface winds established themselves, transporting additional moisture into the atmospheric brew guaranteeing dewpoint temperatures would be in the upper 60's later in the day. Simultaneously, increasing westerly winds aloft were increasing the wind shear as each hour passed. The initial parameters were coming together for a major severe weather event.

On close inspection of the smaller atmospheric details present during the early afternoon of May 29, 1995 it became very clear to me that not only some of the parameters necessary for severe weather were coming together but almost every parameter in the proverbial book seemed to be in place or would be in place later in the day to support a large severe weather event. For example, a cold air pocket aloft at around 20,000 feet above the ground was due to arrive over eastern New York at the time of maximum surface heating. The arrival of the spinning cold pool aloft would cause a further and rapid destabilization of the atmosphere supporting severe thunderstorms. The warm sector of air at the ground was in place with thunderstorms already developing along the strong cold front. Upper winds in the atmosphere were diffluent. In other words, winds aloft were blowing away from each other allowing the thunderstorms to vent, potentially increasing their intensity. Lastly, and perhaps most importantly, the wind shear over New York and New England was unusually strong and increasing promoting the development of rotating supercell thunderstorms and tornadoes.

It would later be determined through research by meteorologists at the Albany National Weather Service forecast office that local terrain effects were responsible for increasing the low level wind shear environment in southeast Columbia county, NY and in Berkshire county, MA which supported the intense supercell thunderstorm that produced the tornadoes.

The National Severe Storms Forecast Center in Kansas City, Missouri (now called the Storm Prediction Center) issued the first of several severe thunderstorm watches for the day. The initial watch included all of upstate eastern New York and was issued to cover the developing thunderstorms over central New York during the morning. Damaging winds, hail, frequent lightning, and heavy rain were all possible and likely in this watch area. At this point there was no "Official" mention of the possibility of tornadoes.

Binghamton, NY National Weather Service Doppler weather radar began showing rotation in a thunderstorm moving into Otsego county, NY at 4:30pm. The first tornado warning of the day was issued at 4:30 pm to cover this storm. By 5:15 pm radar showed the storm weakening. A few reports of scattered straight line wind damage were relayed to the National Weather Service. Apparently, no tornado had formed from the rotating thunderstorm, which is not altogether unusual. Through 6:30 pm, scattered heavy thunderstorms moved through the greater Capital District, producing isolated pockets of wind damage near Rotterdam, NY, and large hail in Poestenkill, NY. The Albany Doppler radar, located in Berne, NY, did not indicate any significant development to the storms through 6pm. The relative calm did not last long, however.

At 6:30 pm, the thunderstorm cell that had been associated with the tornado warning for Otsego county two hours earlier had held together and moved into eastern Greene county, NY. The Albany Doppler radar detected an intensifying circulation of wind inside the storm. The storm grew rapidly, almost exploding as it encountered increase low level wind shear and high dewpoint air in the Hudson valley.  A tornado warning for Columbia county, NY was issued at about 6:40 pm. The storm's trajectory had it aimed straight for the Egremont, Great Barrington, MA area in southern Berkshire county. Warnings were broadcast first on WRGB giving residents of southeast Columbia county NY and Great Barrington, MA an unprecedented 21 minutes lead time.

"Like a bomb went off..." That was a quote from a caller in the Great Barrington area to Channel 6. Radar had pegged it, an apparently devastating storm had ripped through southern Berkshire county. Reports from Columbia county, NY also indicated a serious storm had moved through producing extensive damage to buildings and trees. Home video arrived at Channel 6 by 9:15 pm from a family traveling north on New York's Taconic parkway. The video showed a fifty foot wide path of destruction across the highway and into the woods. The tell tale twisting of debris was convincing evidence that a tornado had been on the ground and was responsible for the damage.

The official report from the teams of meteorologists dispatched from the Albany, NY National Weather Service forecast office to survey the damage concluded the following:

Multiple homes and wooded areas sustained extensive damage in the towns of Greenport and Hillsdale in Columbia county, NY from a tornado which began at 6:40 pm and continued through 7:00 pm. The path length of the tornado was estimated to be thirteen miles long. The storm packed winds estimated between 73 mph to 135 mph.

In Berkshire county, MA, the same parent thunderstorm responsible for the Columbia county tornado produced another, and more violent tornado. Damage began at 7:06 pm at the Great Barrington airport and ended in Monterey, MA at 7:16 pm. Over a length of seven miles and maximum width of 300 yards, the Great Barrington tornado produced a swath of devastation through the town, destroying buildings, tossing cars, and shearing off tree tops. Three people were killed, 27 injured, over 100 homes and businesses were either damaged or destroyed, and many thousands of trees were felled by the storm. Winds in the Great Barrington tornado were estimated to range up to 165 mph making it one of the strongest tornadoes on record in the Northeast.

Berkshire County Damage Photographs (Photographs by Steve LaPointe)

Damage Along Route 23

Photo #1 of 5: May 30, 1995....Decimated forest along route 23 approaching
     Great Barrington, MA

The trees in this photograph are merely a handful of the thousands of trees felled by the tornado. The tree tops were sheared off by winds estimated over 100mph. Debris from the forest shelled the roof of the home in the background of the photograph, producing extensive damage. Had the tornado been level with the ground at this point, the home would likely have suffered damage even more severe than the large holes in the roof. Within minutes after obliterating these trees the tornado went on to devastate the Great Barrington fairgrounds. The next three photographs illustrate the power of the tornado as it tore through the fairgrounds.

Great Barrington Fairgrounds

Photo #2 of 5: May 30, 1995...Destruction of the Great Barrington fairgrounds

The full force of the tornado bore down on the fairgrounds reducing the pavilion to a pile of twisted debris. Winds at this point in the tornado's life were close to 150mph splintering the roof and gutting the interior. Fortunately the buildings were vacant at the time of the storm, so there were no injuries or fatalities on the fairgrounds.

Great Barrington Fairgrounds

Photo #3 of 5: May 30, 1995...Destruction of the Great Barrington fairgrounds

This section of the pavilion crumbled under the force of the tornado. Typically, once the roof blows off a structure the walls collapse very easily, which was the case in this instance. Notice the debris behind the front wall...This is mostly the remains of the roof indicating it blew off first allowing the front, side, and interior walls to collapse, obliterating the entire section of the building.

Great Barrington Fairgrounds

Photo #4 of 5: May 30, 1995...Destruction of the Great Barrington fairgrounds

This is another view of the ruined pavilion. Some of the surrounding debris actually belonged to buildings that were not part of the fairgrounds. As the tornado moved through town damaging trees and buildings it picked up pieces of debris and deposited them hundreds of feet away. Notice in the background of the photograph, the impact point on East Mountain where the tornado slammed into the hill. The next photograph is a close up examination of the impact point.

Tornado Impact on East Mountain

Photo #5 of 5: May 30, 1995...Tornado Impact on East Mountain

Click Here For the Full Gallery of Tornado Damage Photographs

After demolishing the fairgrounds, the tornado slammed into the side of East Mountain, leveling a section of the forest covering the hill. From the photograph it's apparent that the tornado literally hit the mountain and then quickly lifted off the ground, since trees along the top and bottom of the mountain show no damage. As the storm hopped over the mountain it touched down again in Monterey where it attained its maximum intensity with winds estimated briefly in excess of 150 mph.

The intensification of the storm on the other side of East Mountain is not a surprise. As the tornado column descended over the lower elevation adjacent to the mountain, the column stretched vertically causing the winds to increase. The physics involved are known as conservation of angular momentum. It's the same process that allows a figure skater to spin faster when pulling in the arms and extending them upwards.

Also, notice the rotary damage pattern in the trees. The trees are blown down in a pattern showing the counterclockwise spinning motion of the air around the tornado. This type of damage pattern is what meteorologist look for when trying to determine whether or not storm damage was caused by a tornado or severe thunderstorm straight line winds. If the damage in this photograph had been caused by straight line winds the trees along the entire slope of the mountain would have been blown down in one direction, not in the rotary pattern shown.

Doppler Radar (NEXRAD) Images of the Tornado Event

Base Relectivity Display of the Great Barrington Supercell
Albany National Weather Service Doppler Radar Reflectivity Display
on the Channel 6 Weather Spectrum 9000 Computer

1.5 degree Elevation Tilt
7:01 pm May 29, 1995

The radar image above is called a reflectivity display. The radar emits a beam of radiation which bounces off precipitation and returns to the emitter. From the returned pulse, the radar generates an image which shows areas of precipitation and precipitation intensity, a line of intense thunderstorm in this case. The various colors give an estimate of the intensity and even the type of precipitation occurring. The blue and green shades indicate light rain in this case and the yellow, orange, and red shades indicate heavy rain and hail.

There are two storms of note on the display. The storm circled over western Massachusetts was producing the tornado that struck Great Barrington at the time this image was taken. A storm over northern Fairfield county, CT (due south of the Massachusetts storm on the map) was also suspected at this time of producing a tornado. The reflectivity on the Connecticut storm is quite impressive given the distance that storm was from the radar site which is located in Berne, NY in Albany county. The next image is a close look at the Great Barrington, MA supercell thunderstorm.

Close Up Reflectivity Display of the Great Barrington Supercell
Albany National Weather Service Doppler Radar Reflectivity Display

1.5 degree Elevation Tilt, Close-up
7:01 pm May 29, 1995

This is a close-up look at the parent thunderstorm which produced the devastating Great Barrington, MA tornado. This type of thunderstorm is called a supercell. Supercell thunderstorms rotate, most in a counterclockwise direction. These types of storms actually form small centers of low pressure, called mesocyclones. The mesocyclone in this case is located at the head of the arrow in an area of very light precipitation. From the mesocyclone the smaller tornado circulation sometimes develops. The radar signature of the mesocyclone, illustrated in this image, is called a hook echo.

Close up Radial Velocity Display of the Great Barrington Supercell
Albany National Weather Service Doppler Radar Radial Velocity Display
on the Channel 6 Weather Spectrum 9000 Computer

.5 degree Elevation Tilt
7:01pm, May 29, 1995

The colors on this display are not indicating precipitation but instead showing wind speeds and directions relative to the Albany radar site. This image is called a radial velocity display and is the heart of what makes Doppler radar such an important tool in forecasting local severe storms.

The physics are known as the Doppler effect. Essentially, the radar emits a beam of radiation at a certain frequency. When the beam bounces off a particle in the atmosphere which is in motion such as precipitation, dust, or even bugs, the frequency of the reflected beam of radiation returning to the radar is shifted. The shift in frequency is called the Doppler shift. From the frequency shift the radar is able to determine whether the wind is blowing away from or towards the radar site and is able to determine a velocity that is a close approximation to the actual speed of the particle.

The yellow, red, and orange colors on the picture above indicate winds blowing away from the radar site at various velocities and elevations and the greens and blues indicate winds blowing towards the radar site at various velocities and elevations.

The term for the radar indication in the image above is called a mesocyclone vortex signature. The thunderstorm's circulation is evident by there being an area of bright red shading, which shows strong winds blowing away from the radar, directly next to an area of green shading, which shows an area of strong winds blowing towards the radar site. This radar signature shows the parent circulation from which the smaller tornado vortex evolves.

By being able to see the small scale wind currents within severe thunderstorms meteorologists can now give tremendous tornado warning lead times. In fact, this signature on the Doppler allowed a 21 minute warning lead time to the arrival of the tornado in Great Barrington, saving peoples lives.