New England Tornadoes

[vortex95]

I have completed my analysis of New England tornadoes.  What I have done is selected

the most significant tornadoes from the 19th century, 1900-1974 all killer tornadoes/F3

or greater (June 1953 outbreak is an exception), and all tornadoes 1975-present.

 

I used the Tornado Project (Tom Grazulis) compilation of all documented New England

tornadoes through 2000, cross referencing with the NCDC database that begins in 1950. 

Since 2000, I have compiled my own list using local PNSs and Storm Data.  In the vast

majority of cases, the Grazulis and the NWS F-ratings match.  One notable exception is

the Great Barrington killer tornado from May 1995.  The NWS rated it F4, but this was

only based on an auto thrown a few hundred feet.  It does not take F4 winds to do this. 

Grazulis feels F3 is more appropriate.  I have been told by people in the know that the

Grazulis database for F-ratings prior to the late 1990s is better to use than what shows

up in the NCDC database.

 

There are a few after 2000 that do not show up in the NCDC database, like the F2

tornado in Cavendish VT in July 2003.  I found about this tornado in the poster session

at the SLS conference in 2004, and it was quite detailed.  Sometimes oversights do

occur, or late reports come in that do not make the Storm Data publication deadline.

 

Why did I pick 1975 as a starting point for all tornadoes?  After the 1953 ORH tornado,

tornado reports in New England skyrocketed.  One year MA reported 12 and ME 11, and

as many as 26 tornadoes were logged in single year for the six state region.  Grazulis

strongly suspects that half of what is in the official database 1954-1974 were not

actual tornadoes, but downbursts.  For example, he found one entry for a tornado in

ME was logged only because there was wind damage under a hook echo on radar.  As

we know, this does not always mean a tornado has occurred, as RFD winds are just a

form of downburst and can exceed easily exceed 100 mph.  Fujita's studies made the

downburst recognized as a discrete phenomena, and as a result after the early 1970s,

tornado reports dropped considerably.  The annual counts are likely too low 1975 to

the early 1990s.  With the NWS Modernization in the early 1990s, detection and

documentation became much better, and with cameras everywhere and social media

in recent years, we are probably close to what actually occurs each year.  NCDC has

a map of average tornadoes per state per year 1991-2010, and it is 7 for New England.

 

http://www.ncdc.noaa.gov/extremeevents/images/tornado/ann-avg-torn1991-2010.gif

 

For 1995-2014, the 20 year average is 6.7, so that rounds to 7.  Wikipedia says 8 per

year, but that is the 1950-2010 average, and as noted above, totals were inflated

1954-1974.

 

For the few listed that crossed the border from NY, I only listed the statistics for what

occurred in New England.

 

What I did do for waterspouts?  I have seen the NWS count them over inland lakes,

even if they did not come ashore.  My opinion is that if a waterspout is spawned by a

full blown thunderstorm over an inland lake, count it.  Why?  From a climo perspective,

it shows the atmosphere was conducive to tornado formation on a given day.  This

isn't well-defined criteria, as thunderstorms just offshore over the ocean spawn

waterspouts, but it is what the NWS does.

 

One note I feel necessary to point out regarding the biggest tornado outbreak the

region has seen in terms of actual number of tornadoes in a single day (July 10, 1989). 

Four separate tornadoes are listed in Worcester County MA, all occurring within 10

minutes of each other in adjacent towns.  They are listed both by the NWS and

Grazulis.  I feel these were not 4 separate tornadoes.  Looking at radar loops for the

supercell, it was moving rapidly SE at 40+ mph, so this was probably the same

mesocyclone with intermittent weakening/strengthening (damage path gaps) of one

or two tornadoes, but I have listed them as they are in the official record here.

 

Attached is a text file that summarizes each tornado.  Main statistics are all there

(F/EF-scale, path length, path width, time, etc).  I kept the descriptions brief for the

most part.  Format is adopted from Grazulis.

 

Now Weatherwiz can rejoice in stats and CoastalWx can pride himself in the tornado

in Brockton in 1989. 

 

newenglandtor.txt
 

[CT Rain]

 

I have completed my analysis of New England tornadoes.  What I have done is selected

the most significant tornadoes from the 19th century, 1900-1974 all killer tornadoes/F3

or greater (June 1953 outbreak is an exception), and all tornadoes 1975-present.

 

I used the Tornado Project (Tom Grazulis) compilation of all documented New England

tornadoes through 2000, cross referencing with the NCDC database that begins in 1950. 

Since 2000, I have compiled my own list using local PNSs and Storm Data.  In the vast

majority of cases, the Grazulis and the NWS F-ratings match.  One notable exception is

the Great Barrington killer tornado from May 1995.  The NWS rated it F4, but this was

only based on an auto thrown a few hundred feet.  It does not take F4 winds to do this. 

Grazulis feels F3 is more appropriate.  I have been told by people in the know that the

Grazulis database for F-ratings prior to the late 1990s is better to use than what shows

up in the NCDC database.

 

There are a few after 2000 that do not show up in the NCDC database, like the F2

tornado in Cavendish VT in July 2003.  I found about this tornado in the poster session

at the SLS conference in 2004, and it was quite detailed.  Sometimes oversights do

occur, or late reports come in that do not make the Storm Data publication deadline.

 

Why did I pick 1975 as a starting point for all tornadoes?  After the 1953 ORH tornado,

tornado reports in New England skyrocketed.  One year MA reported 12 and ME 11, and

as many as 26 tornadoes were logged in single year for the six state region.  Grazulis

strongly suspects that half of what is in the official database 1954-1974 were not

actual tornadoes, but downbursts.  For example, he found one entry for a tornado in

ME was logged only because there was wind damage under a hook echo on radar.  As

we know, this does not always mean a tornado has occurred, as RFD winds are just a

form of downburst and can exceed easily exceed 100 mph.  Fujita's studies made the

downburst recognized as a discrete phenomena, and as a result after the early 1970s,

tornado reports dropped considerably.  The annual counts are likely too low 1975 to

the early 1990s.  With the NWS Modernization in the early 1990s, detection and

documentation became much better, and with cameras everywhere and social media

in recent years, we are probably close to what actually occurs each year.  NCDC has

a map of average tornadoes per state per year 1991-2010, and it is 7 for New England.

 

http://www.ncdc.noaa.gov/extremeevents/images/tornado/ann-avg-torn1991-2010.gif

 

For 1995-2014, the 20 year average is 6.7, so that rounds to 7.  Wikipedia says 8 per

year, but that is the 1950-2010 average, and as noted above, totals were inflated

1954-1974.

 

For the few listed that crossed the border from NY, I only listed the statistics for what

occurred in New England.

 

What I did do for waterspouts?  I have seen the NWS count them over inland lakes,

even if they did not come ashore.  My opinion is that if a waterspout is spawned by a

full blown thunderstorm over an inland lake, count it.  Why?  From a climo perspective,

it shows the atmosphere was conducive to tornado formation on a given day.  This

isn't well-defined criteria, as thunderstorms just offshore over the ocean spawn

waterspouts, but it is what the NWS does.

 

One note I feel necessary to point out regarding the biggest tornado outbreak the

region has seen in terms of actual number of tornadoes in a single day (July 10, 1989). 

Four separate tornadoes are listed in Worcester County MA, all occurring within 10

minutes of each other in adjacent towns.  They are listed both by the NWS and

Grazulis.  I feel these were not 4 separate tornadoes.  Looking at radar loops for the

supercell, it was moving rapidly SE at 40+ mph, so this was probably the same

mesocyclone with intermittent weakening/strengthening (damage path gaps) of one

or two tornadoes, but I have listed them as they are in the official record here.

 

Attached is a text file that summarizes each tornado.  Main statistics are all there

(F/EF-scale, path length, path width, time, etc).  I kept the descriptions brief for the

most part.  Format is adopted from Grazulis.

 

Now Weatherwiz can rejoice in stats and CoastalWx can pride himself in the tornado

in Brockton in 1989. 

 

newenglandtor.txt

 

 

 

Nice work! I did find the Cavendish storm in storm data but it was a super-late entry... you can find it a few months late. 

 

Also, I'm glad you included the July 21, 2010 Litchfield County tornadoes as 1 event. I have no idea why they are listed as 3 separate storms. You could really include the Hartford County/Bristol tornado as the same one since the towns are all adjacent and it was the same storm. 

 

The June 23, 2001 tornadoes in NW CT is a very interesting case. The Washington/Lake Waramaug tornado was a separate storm than the Torrington/Winsted/Hartland storm. The Storm Data entry is wrong and I believe the time is incorrect as well. I'm doing a study on CT tornadoes since 1994 and this case jumped out at me as an error in the climatology. 

[CoastalWx]

Nice work Sparky. I remember we talked about the Cavendish tornado a few months ago. Good stuff.

[Ginx snewx]

Great stuff thanks. May I contribute an undocumented uninvestigated possible F0.As the eyewall of Hurricane Bob passed just west of Ashaway RI on August 15th 1991, I witnessed about 100 yards from my back yard a possible tornado. Once the winds died down I investigated and found a distinct path about 1/3 of a mile wide, the path started at the top of Collins Rd elevation 210 ft, two telephone poles snapped with wires wrapped around each other, continued SE into a wooded area marked clearly by trees falling in opposite directions for about 3/4 mile, it then appeared to lift at elevation 150 with sheared trees for about 1/2 mile, as the elevation dropped to 100 a corn field and grass field showed obvious rotation signs. At the time I was a spotter for the NWS and called it in. Apprently no one investigated from the PVD office but notation did appear in the historical records of an unconfirmed tornado in SWRI. The severity of the telephone wire damage took 10 days to repair. It was obviously not straight line or macroburst winds but because of staffing issues it was never investigated. Just thought I would add this because to this day there are 0 documented tornados in Washington County RI

[weatherwiz]

This is really incredible stuff!  

 

Excellent work as well with some of the years with very high tornado counts and with the events which had multiple listings but probably were just from the same storm.  

 

This is just great though...definitely going to keep these notes and have them for reference in my notebooks and such.  

 

Fantastic work

[tamarack]

Nice work.  Couple of nitpicks on Maine tornados, and a question (or two):

--Nitpick 1:  7/23/2002 Aroostook - Knowles 'Corner' (not Center)

--Nitpick 2:  9/11/2013 'Piscataquis' - Actually Aroostook, and there were thousands of trees toppled in Oxbow and T9R5, though I don't know how much there was tornadic and how much microburst.  The small areas I viewed on Oxbow last week were from straight-line winds.

 

Question:  I've read - and seen some old evidence - of a significant tornado or other-cause blowdown from (I think) 1963, which crossed the Allagash River about 20 miles SW of Allagash Village.  It was late 1970s when I saw the blowdown patches east of the river, and the major damage was said to be on the west side.  I've never seen any official record of this event, which occurred entirely on unpopulated forest.  A search for it did, however, turn up an EF2 on 8/15/1958, in Winterville/Eagle Lake.  Did that (or they) fall out due to the 1954-74 over-reporting?

[radarman]

Good stuff.  Pretty amazing the only confirmed tornado in Eastern Hampshire County was an EF-0 embedded in a mostly non-convective heavy rain event in early May (2009).  You'd think if the valley channeling effect were real you'd see at least one out of an actual supercell.   In fact there are many more over the hill towns to the east and west.

[New England Storm]

This is cool, thank you!

[vortex95]

Great stuff thanks. May I contribute an undocumented uninvestigated possible F0.As the eyewall of Hurricane Bob passed just west of Ashaway RI on August 15th 1991, I witnessed about 100 yards from my back yard a possible tornado. Once the winds died down I investigated and found a distinct path about 1/3 of a mile wide, the path started at the top of Collins Rd elevation 210 ft, two telephone poles snapped with wires wrapped around each other, continued SE into a wooded area marked clearly by trees falling in opposite directions for about 3/4 mile, it then appeared to lift at elevation 150 with sheared trees for about 1/2 mile, as the elevation dropped to 100 a corn field and grass field showed obvious rotation signs. At the time I was a spotter for the NWS and called it in. Apprently no one investigated from the PVD office but notation did appear in the historical records of an unconfirmed tornado in SWRI. The severity of the telephone wire damage took 10 days to repair. It was obviously not straight line or macroburst winds but because of staffing issues it was never investigated. Just thought I would add this because to this day there are 0 documented tornados in Washington County RI

 

Awhile back, a NWS forcester told me there were a number of reports of tornadoes on the south coast, including Cape Cod, during Hurricane Bob, but there was no easy way to confirm them due to so much straight line wind damage and the huge task of

surveying the hurricane itself.  This was pre-NEXRAD days and prior to the consolidation of the local officies to Taunton.  Given the

very meticulous surveys to count the swarms of tornadoes for the big Gulf of Mexico and FL landfalls last decade, I bet we'd be

we'd see more surveying and be able to confirm such reports today.

There has been a tornado in Washington County, just not on the mainland (Block Island).

[vortex95]

Nice work.  Couple of nitpicks on Maine tornados, and a question (or two):

--Nitpick 1:  7/23/2002 Aroostook - Knowles 'Corner' (not Center)

--Nitpick 2:  9/11/2013 'Piscataquis' - Actually Aroostook, and there were thousands of trees toppled in Oxbow and T9R5, though I don't know how much there was tornadic and how much microburst.  The small areas I viewed on Oxbow last week were from straight-line winds.

 

Question:  I've read - and seen some old evidence - of a significant tornado or other-cause blowdown from (I think) 1963, which crossed the Allagash River about 20 miles SW of Allagash Village.  It was late 1970s when I saw the blowdown patches east of the river, and the major damage was said to be on the west side.  I've never seen any official record of this event, which occurred entirely on unpopulated forest.  A search for it did, however, turn up an EF2 on 8/15/1958, in Winterville/Eagle Lake.  Did that (or they) fall out due to the 1954-74 over-reporting?

 

This is part of the PNS from NWS CAR for the 9/11/13 event.

 

LOCATION...13 NW MOOSEHORN CROSSING IN PISCATAQUIS COUNTY MAINE

DATE...SEP 11 2013

ESTIMATED TIME...348 PM EDT

MAXIMUM EF-SCALE RATING...EF0

ESTIMATED MAXIMUM WIND SPEED...80 MPH

MAXIMUM PATH WIDTH...125 YARDS

PATH LENGTH...450 YARDS

BEGINNING LAT/LON...46.4726N / -68.9674

ENDING LAT/LON...46.4718N / -68.9623

* FATALITIES...0

* INJURIES...0

 

AN AREAL SURVEY SHOWED THE TORNADO OCCURRED IN THE NORTH MAINE

WOODS PRODUCING A DAMAGE PATH APPROXIMATELY 450 YARDS LONG BY 125

YARDS WIDE. THE STORM WHICH PRODUCED THE TORNADO QUICKLY BECAME

OUTFLOW DOMINATED AND PRODUCED WIDESPREAD STRAIGHT LINE WIND

DAMAGE AS IT CONTINUED TO MOVE EAST-SOUTHEAST. THE AREA HIT

HARDEST BY THE STRAIGHT LINE WINDS IS APPROXIMATELY 18 MILES

SOUTHEAST OF ASHLAND OR ABOUT 12 MILES EAST-SOUTHEAST OF OXBOW.

THE DAMAGE WAS MORE SPORADIC THEREAFTER.

 

-----

 

So what I did is plot the initial lat/lon of the tornado position given above, and that puts it just under a mile NNE of Smith Brook Pond.  Checking maine.gov, Smith Brook Pond is in Piscataquis County.  I fixed the entry to say this:

 

ME    SEP 11, 2013    1448    0k     0inj    125y  0.3m  EF0

PISCATAQUIS - Damage to trees in the North Main Woods

near Smith Brook Pond.

 

Regarding the 8/15/58 Allagash forest blowndown, Grazulis said this might have been a major tornadic event, but there is nothing

he could find that conclusively proves it.  He did list it as a possible F2 tornado 300 yards wide and 20 miles long.

 

For 1900-1974, I only listed killer tornadoes or those of F3 or greater, with the exception of 6/9/53.

[tamarack]

Thanks for the added info, and explanation of the pre-1975 data.  That report from CAR is still confusing, though with your post it can be discerned.  A search for "Moosehorn Crossing", of which I'd never heard despite riding over and canoeing under it, reveals it to be where the Oxbow Road crosses the Aroostook River, on T8R8 WELS, in Penobscot County.  That spot lies about 13 miles from the lat-lon provided in the report, and your accurate plot of +/- one mile NE from Smith Brook Pond, on T9R10 WELS, which is indeed in Piscataquis.  Not sure whether there are two separate Moosehorn Crossings, or what, but the damage description settles the extent of the tornado, for me at least.

[Ginx snewx]

Awhile back, a NWS forcester told me there were a number of reports of tornadoes on the south coast, including Cape Cod, during Hurricane Bob, but there was no easy way to confirm them due to so much straight line wind damage and the huge task of

surveying the hurricane itself. This was pre-NEXRAD days and prior to the consolidation of the local officies to Taunton. Given the

very meticulous surveys to count the swarms of tornadoes for the big Gulf of Mexico and FL landfalls last decade, I bet we'd be

we'd see more surveying and be able to confirm such reports today.

There has been a tornado in Washington County, just not on the mainland (Block Island).

Yea I worked ground truthing Nexrad back then too. Lol ask any SW RIer and they will insist BI is not a part of WC, but of course it is.

[CT Rain]

Yea I worked ground truthing Nexrad back then too. Lol ask any SW RIer and they will insist BI is not a part of WC, but of course it is.

Didn't it become part of south county in the mid 20th century from Newport county ? There's a marker near Yellow Kittens about it.

[Ginx snewx]

Didn't it become part of south county in the mid 20th century from Newport county ? There's a marker near Yellow Kittens about it.

Many a stupor night at the kittens, yea," A series of conflicts involving the Niantic's offshore sub-tribe the Manisseans on Block Island gave that island to the Bay Colony for a number of years, then to the Rhode Island Colony under Newport County, and then to Washington County in 1959."

OT the archaeological team from work recently won a contract from the State of RI to investigate artifacts unearthed by Sandy on BI. They have been out there all summer and have been bringing stuff back to the labs for microanylsis and xray

[vortex95]

So Storm Data through August is now in, and I updated a few things, like total damage

for the Revere MA tornado in July.

 

Turns out a few waterspouts in NH I overlooked.  Thing is a LSR/PNS is not always

issued when one of these occur, so I have to wait until Storm Data is published.

So it is now 9 tornadoes for 2014 instead of 7.  Like I have said, small inland lake

waterspouts pose somewhat of a quandary as to classification, but the line has to

be drawn somewhere.  Given the risk is higher than the open ocean and most of

these are spawned by full blown tstms, they are included.  NWS does this.  In

Storm Data however, there is a separate entry for "waterspout"!  Ok, but what if the

waterspout never gets to land, but flips a boat and causes fatalities?  Do the fatalities

show up in the annual list of killer tornadoes in the U.S.?  It should, but then the

definition is violated!  So back to just calling them tornadoes when over small, inland

lakes!  I know this sounds pedantic, but there are real world applications and issues

that can arise from such.  Like insurance claims as one example.  We know this all

too well from Sandy.

 

Here's what I added.  Full text file attached for everything.

 

 

NH    JUN 25, 2012    1805    0k   0inj   10y  0.1m  EF0
BELKNAP - Waterspout on Lake Winnipesaukee just off
Chase Island.

NH    JUL 17, 2012    1715    0k   0inj   30y  0.5m  EF0
GRAFTON - Waterspout on Newfound Lake.  

 

NH    JUL  4, 2014    1905    0k   0inj   10y  0.4m  EF0
BELKNAP - Waterspout three miles northeast of Gilford
on Lake Winnipesaukee.

NH    JUL 24, 2014    0730    0k   0inj   10y   35y  EF0
BELKNAP - Thin rope-like waterspout in Melvin Bay on
Lake Winnipesaukee.

newenglandtor.txt

[eekuasepinniW]

lol thanks for the reminder of all the waterspouts I missed.