Star Ledger/NJ.com article focused on "warning fatigue"

[NJHurricane]

http://www.nj.com/weather-guy/index.ssf/2011/08/after_multiple_warnings_and_no.html

food for thought.....let the discussion roll

[NJHurricane]

I wish the article were longer/more in depth. That said, its not often that a mainstream news outlet even obliquely deals with the subject...and I applaud Mt Holly NWS for allowing one of its employees to be quoted on the subject.

Its a tough nut to crack....there was a serious spin-up per radar over Wilimington recently....and yet I was not surprised that it was short lived and did not have serious affects on family as the storm moved into SW NJ. The argument that most "tornado" warnings in this part of the country don't really imply anything more damaging than straight line winds is legit IMO. The flip side of this is many times when straight line wind damage is significant casual observers compare the damage to a "tornado".

[SouthernNJ]

In this part of the country, the public tends to equate tornadoes with the amount of damage that happens

from a storm, or just odd cloud formations during severe weather. Based upon this, the public is

convinced we have had multiple tornadoes this year, even though that is not the case. Obviously, not

getting warning verifications for 13 TOR warnings hurts the official statistics, but the public's response

has not been to ignore warnings because they believe things have actually happened. The FAR only

hurts if the people believe that nothing happens during the time of issuance. I have talked to a lot of

people around my area and many of them continue to take protective action during warnings because

of the heavy damage (ie. Hackettstown, Denville, Jackson, Whiting, Brick) inflicted during them, even

though a tornado was never confirmed in any of these areas.

A lady that lives in Wharton State Forest took her kids to the basement last month when winds toppled

large trees near her home. The trees had to be removed for access to the roadway in and out. She still

says she is convinced it was a tornado, even though she did not even know there was a TOR warning at

the time. In any event, as long as the public continues to have its opinions like this, we do not have to

worry about Warning Fatigue just yet.

[phlwx]

I wish the article were longer/more in depth. That said, its not often that a mainstream news outlet even obliquely deals with the subject...and I applaud Mt Holly NWS for allowing one of its employees to be quoted on the subject.

Its a tough nut to crack....there was a serious spin-up per radar over Wilimington recently....and yet I was not surprised that it was short lived and did not have serious affects on family as the storm moved into SW NJ. The argument that most "tornado" warnings in this part of the country don't really imply anything more damaging than straight line winds is legit IMO. The flip side of this is many times when straight line wind damage is significant casual observers compare the damage to a "tornado".

People in general have no clue what the difference is between downburst/straightline and tornado and it's much harder in this part of the world to differentiate the two on a pure damage standpoint:

*Tornadoes here rarely get past EF-2 in terms of damage. There's been only a couple of EF-3 in the last 25 years in this part of the world and I don't believe there's been F-4 or higher in the Delaware Valley in the last 50 years. The weaker tornadoes do equivalent or a bit more damage compared to straight line winds. Once you get to EF-3 things get much nastier on scale.

*Tornadoes are rarely long track in this part of the world...again, mostly due to their weak nature and also in part to geography around here (especially west of 95).

I don't think the warnings themselves are a bad thing since it raises awareness to get your a-- inside and away from a storm. With that said, I think the media can do a better job of correcting "tornado" reports and calling them straight line when they really are that. Some (Glenn) are very good at calling it what it is...it's not universal though.

[Grothar]

The media must do a better job with informing and educating the public. For instance, in Memphis, when their is a tornado warning issued, the local tv stations use their fancy radar systems to track the storm instead of using the runner at the bottom of the screen with a on screen forecaster explaining to the public the difference between straight line winds and a tornado --see below. I would also ask that Mt. Holly office consider working with the local fire companies and or emergency management to utilize the volunteer fire companies dispatch sirens as tornado sirens using a wave length in the horn to differentiate the tornado siren from the normal siren.

[johnc]

meh

I think at some level people need to be responsible for themselves. And if people want to be morons about it, should the NWS really go out of its way to coddle their moronic behaviors? "Ohh yet another tornado warning... I'm just going to ignore it." "There was a tornado warning but my house is still standing... clearly they don't know what they're talking about." If you want to be an ignorant fool, then you deserve to get cracked on the head.

Actually the tornado warning part isn't even that concerning to me. They're not really that common. If you look at the square mileage of people under a warning where there was no actual tornado damage, the number of people affected is actually relatively small.

The one area where I think there needs to be greater skill is severe thunderstorm warnings. Those things get dropped like pepperoni on a pizza but more often than not people just get some lightning and heavy rain. Is there a verification rate available for SVR polygons? I'd be surprised if it's higher than 10%. Hmmm... that sounds like a project for me. Maybe I'll do one for Mt. Holly, then branch out to other WFOs.

[Grothar]

meh

I think at some level people need to be responsible for themselves. And if people want to be morons about it, should the NWS really go out of its way to coddle their moronic behaviors? "Ohh yet another tornado warning... I'm just going to ignore it." "There was a tornado warning but my house is still standing... clearly they don't know what they're talking about." If you want to be an ignorant fool, then you deserve to get cracked on the head.

Actually the tornado warning part isn't even that concerning to me. They're not really that common. If you look at the square mileage of people under a warning where there was no actual tornado damage, the number of people affected is actually relatively small.

The one area where I think there needs to be greater skill is severe thunderstorm warnings. Those things get dropped like pepperoni on a pizza but more often than not people just get some lightning and heavy rain. Is there a verification rate available for SVR polygons? I'd be surprised if it's higher than 10%. Hmmm... that sounds like a project for me. Maybe I'll do one for Mt. Holly, then branch out to other WFOs.

Go for the study. It needs to be done. I would also be nice if you could figure out how many of those warnings were issued too early, what type of damage derived from the storm and was there a watch in place. Good luck

[Lady Di]

Some of the public does not know the difference between a watch and a warning. I've been asked that question on more than one occasion believe it or not. I am not educated in weather other than "self" and having experienced weather for some 50 years.

I agree with the met in the article that a TOR warning should be issued on the criteria of an EF2 which would be a rare thing around here. Also, the servere t-storm warning includes the mention the storng winds -so possibly add something in there about rotation?

I don't think the general public cares to be educated on the different types of severe weather there is. A downburst, supercell, tornado, etc. Most people, imo, don't care to look at a radar shot. I feel you are better off with the beeping television and a firm voice reading the warning. Sirens too, if necessary. Get them used to "severe" weather warnings, rather than using the word tornado. They are rare here, and when folks hear that word, they may shrug off the warning.

[LocoAko]

meh

I think at some level people need to be responsible for themselves. And if people want to be morons about it, should the NWS really go out of its way to coddle their moronic behaviors? "Ohh yet another tornado warning... I'm just going to ignore it." "There was a tornado warning but my house is still standing... clearly they don't know what they're talking about." If you want to be an ignorant fool, then you deserve to get cracked on the head.

Actually the tornado warning part isn't even that concerning to me. They're not really that common. If you look at the square mileage of people under a warning where there was no actual tornado damage, the number of people affected is actually relatively small.

The one area where I think there needs to be greater skill is severe thunderstorm warnings. Those things get dropped like pepperoni on a pizza but more often than not people just get some lightning and heavy rain. Is there a verification rate available for SVR polygons? I'd be surprised if it's higher than 10%. Hmmm... that sounds like a project for me. Maybe I'll do one for Mt. Holly, then branch out to other WFOs.

They do exist - http://mesonet.agron.iastate.edu/cow/?syear=2009&smonth=8&sday=11&shour=12&eyear=2011&emonth=8&eday=11&ehour=12&wfo=PHI&wtype%5B%5D=SV&hail=1.00&lsrbuffer=15&wind=58

I ran the numbers here for 2009-2011 (2005-2011 was taking too long to load). The verification rate is 53.1% for WFO PHI, with a POD of 0.71 and a FAR of 0.47.

Sorry to disappoint you...

[johnc]

They do exist - http://mesonet.agron...ffer=15&wind=58

I ran the numbers here for 2009-2011 (2005-2011 was taking too long to load). The verification rate is 53.1% for WFO PHI, with a POD of 0.71 and a FAR of 0.47.

Sorry to disappoint you...

Disappointed? No. Surprised? A little bit. But the actual number is 18% when you require reports that verify severe criteria.

[Thunder Road]

Given the limitations of radar, wouldn't it be nearly impossible to distinguish between an EF1 and an EF2 by looking at BV scans. Besides, what a lot of people don't understand is that an EF2 tornado may only create EF2 damage for a very small area of an otherwise larger track. I'd say hold off on any nationwide changes until dual pol is installed at most or all WFOs. Right now it's just too hard to tell what is a damaging tornado and what isn't by looking only at the radar.

[HM]

Severe thunderstorm warnings are pretty successful I'd say and I think the problem comes from the lack of understanding how severe wind gusts work. It is very rare to have a large-scale macroburst (sorry for the overstating) that delivers a widespread area of wind over severe criteria. A thunderstorm capable of severe wind gusts will only produce small areas of that wind within the warned area.

I don't like the idea of issuing tornado warnings based on the estimated strength. That ignorance is grounds for a potential fiasco down the road.

[radarman]

There was a pretty good thread about this subject recently:

[RobbTC]

Article needs an update... 1 actual tornado confirmed.

PUBLIC INFORMATION STATEMENT

NATIONAL WEATHER SERVICE MOUNT HOLLY NJ

301 PM EDT THU AUG 11 2011

...TORNADO CONFIRMED NEAR BERGEN MILLS IN MONMOUTH COUNTY NEW

JERSEY...

LOCATION...BERGEN MILLS IN MONMOUTH COUNTY NEW JERSEY

DATE...08/09/11

ESTIMATED TIME...2:30 PM EDT

MAXIMUM EF-SCALE RATING...EF 0

ESTIMATED MAXIMUM WIND SPEED...70 MPH

MAXIMUM PATH WIDTH...50 YARDS

PATH LENGTH...1/2 MILE

BEGINNING LAT/LON...40.25 N 74.43 W

ENDING LAT/LON...40.27 N 74.41 W

* FATALITIES...0

* INJURIES...0

...SUMMARY...

THE NATIONAL WEATHER SERVICE IN MOUNT HOLLY NJ HAS CONFIRMED A

TORNADO NEAR BERGEN MILLS IN MONMOUTH COUNTY NEW JERSEY ON

08/09/11. THE TORNADO STARTED 1/2 MILE NORTH OF BUONO FARM AND

TRACKED NORTHEAST CROSSING HIGHWAY 33 IN BERGEN MILLS DAMAGING A

FLAGPOLE AND FENCE AT THE XTREME MACHINES SHOP. NUMEROUS TREE

BRANCHES WERE DOWN ALONG THE PATH. THE TORNADO MOVED ALONG

ARROWHEAD WAY BEFORE LIFTING.

THIS INFORMATION CAN ALSO BE FOUND ON OUR WEBSITE AT

WEATHER.GOV/PHI.

THE NATIONAL WEATHER SERVICE WOULD LIKE TO THANK THE XTREME

MACHINES SHOP FOR ITS ASSISTANCE IN THIS MATTER.

FOR REFERENCE...THE ENHANCED FUJITA SCALE CLASSIFIES TORNADOES

INTO THE FOLLOWING CATEGORIES:

EF0...WIND SPEEDS 65 TO 85 MPH.

EF1...WIND SPEEDS 86 TO 110 MPH.

EF2...WIND SPEEDS 111 TO 135 MPH.

EF3...WIND SPEEDS 136 TO 165 MPH.

EF4...WIND SPEEDS 166 TO 200 MPH.

EF5...WIND SPEEDS GREATER THAN 200 MPH

[phlwx]

Article needs an update... 1 actual tornado confirmed.

70 mph winds and tree branches knocked up...that's epic sh*t right there.

[Thunder Road]

70 mph winds and tree branches knocked up...that's epic sh*t right there.

looks like there will be plenty of baby saplings come next Spring

[SouthernNJ]

Funny, the Monmouth storm had less rotation and damage then the Wilmington storm, yet it was a brief

tornado as opposed to straight-line winds. That right there is the center of Joe Miketta's point of SVR vs.

TOR for our area. We only had a SVR in effect for the Monmouth storm, and it only did damage minor

damage, but yet it is a "missed tornado event" on the stats sheet. Mt. Holly actually does not issue

many TOR warnings compared to some of the surrounding offices:

Sterling Va.: 76 (only 1 EF-2 this year)

State College: 37 (only 1 warned EF-2 this year, 2 unwarned)

Binghamton: 30 (only 1 warned EF-2 this year, 1 unwarned)

I would be far more concerned with warning fatigue in these areas because of the shortage of

significant tornadoes with the extra amount warnings issued.

[Thunder Road]

Funny, the Monmouth storm had less rotation and damage then the Wilmington storm, yet it was a brief

tornado as opposed to straight-line winds. That right there is the center of Joe Miketta's point of SVR vs.

TOR for our area. We only had a SVR in effect for the Monmouth storm, and it only did damage minor

damage, but yet it is a "missed tornado event" on the stats sheet. Mt. Holly actually does not issue

many TOR warnings compared to some of the surrounding offices:

Sterling Va.: 76 (only 1 EF-2 this year)

State College: 37 (only 1 warned EF-2 this year, 2 unwarned)

Binghamton: 30 (only 1 warned EF-2 this year, 1 unwarned)

I would be far more concerned with warning fatigue in these areas because of the shortage of

significant tornadoes with the extra amount warnings issued.

I believe the 13 quoted in the article was just for NJ, although I'm surprised Sterling has issued 76 of 'em. I know the southern part of their CWA was part of the 4/27 outbreak, but still...

[SouthernNJ]

Polygon Era Statistics for SVR Warnings for Mt Holly...this is unofficial and based upon my own research using the NCDC data and matching it up to the Iowa Mesonet Warning Archive. Severe Events are Hail of severe criteria, Winds measured 58 MPH or higher, Wind Damage or Tornadoes:

Severe Events Events Warned Events Unwarned POD% SVR Warnings Verified FAR% Lead Time for First Event

2007: 238 206 32 86% 339 207 39% 19.8 mins

2008: 292 249 43 85% 284 169 41% 15.0 mins

2009: 297 241 56 81% 247 153 38% 16.7 mins

2010: 247 224 23 90% 195 128 34% 17.4 mins

Total: 1074 920 154 86% 1065 657 38% 17.2 mins

[MGorse]

I agree with the met in the article that a TOR warning should be issued on the criteria of an EF2 which would be a rare thing around here.

I would like to know how you can tell from radar data the strength of a tornado. Good luck with that! This would put even more stress on the warning forecasters. Joe Miketta knows how I and some others in the office feel about this. We typically have weak tornadoes that do not last long at all, which makes it rather difficult from a warning standpoint.

Regarding the severe weather back on August 9th, I am a bit surprised that my office did not send someone to look at the damage in the Wilmington, DE area especially since a tornado warning was issued. Yet, no tornado warning issued for Monmouth County and a weak tornado has now been confirmed. I know that many times we rely on the emergency managers to take a look at the damage first and if they feel it could be tornadic, then they ask us to do a damage survey. I get the sense that the Wilmington, DE area damage was ruled as straight-line wind with even some areal footage looked at.

[SouthernNJ]

I was watching KDIX on the Mt. Holly website when the SVR went out for Monmouth County. The storm was

embedded within an area of heavy rainfall. The SRV showed a brief but pronounced mesocyclone near a

larger area of downburst winds. I like to pretend to be in the warning forecaster seat, and I was about

50/50 on whether a TOR should have been issued, but of course by the next scan the signature was

gone. Just shows how quick things go around here. The Wilmington storm had much stronger

rotation and did a lot more damage despite not being a tornado.

I viewed many tornadoes this year on various radars at websites of NWS offices. It is impossible to

differentiate between an EF-1 signature and an EF-2. EF-3 plus is much easier to see and should have

stronger wording in the action statements in order to caution the public of a life-threatening situation.

Of course we never have to worry about that here.

[The Great Zo]

I viewed many tornadoes this year on various radars at websites of NWS offices. It is impossible to

differentiate between an EF-1 signature and an EF-2. EF-3 plus is much easier to see and should have

stronger wording in the action statements in order to caution the public of a life-threatening situation.

Of course we never have to worry about that here.

There is no such thing as an EF-1 signature, EF-2 signature, etc. For one, tornadoes are rarely directly sampled by radar, due to the resolution and height of the beam. Also, the variations in a radar signature based on beam height, resolution, viewing angle, and time-of-sampling, are orders of magnitude greater than the variations in wind speed of an actual tornadic vortex.

Radar science isn't going to improve much in the near term. Dual-pol's benefits do not include assistance in tornado detection, except in very rare cases when debris is observed and the tornado's presence is already patently obvious. Until additional radars are installed (good luck with that), there is little that can be done from a radar science standpoint.

[SouthernNJ]

TOR Warnings since 2002 for Mt. Holly:

2002: 14 Issued 3 Verified (4 hit 0 miss)

2003: 5 Issued 3 Verified (4 hit 4 miss)

2004: 11 Issued 4 Verified (5 hit 1 miss)

2005: 4 Issued 0 Verified (No Tornadoes in '05)

2006: 10 Issued 0 Verified (0 hit 2 miss)

2007: 7 Issued 0 Verified (No Tornadoes in '07)

2008: 10 Issued 1 Verified (1 hit 1 miss)

2009: 13 Issued 1 Verified (1 hit 1 miss)

2010: 12 Issued 2 Verified (2 hit 2 miss)

[famartin]

I find myself often thinking about "warning fatigue" out here, but for a somewhat different reason: no people! We have massive counties, and a lot of warning systems still activate by the county, but in every county the vast majority of the (still rather small) population is located in one or two areas which are extremely unlikely to be affected by any given warning. So you end up alerting many people who are going to be unaffected virtually every time, even if the storm in question produces a verifiable severe weather report.

[Rainshadow]

They do exist - http://mesonet.agron.iastate.edu/cow/?syear=2009&smonth=8&sday=11&shour=12&eyear=2011&emonth=8&eday=11&ehour=12&wfo=PHI&wtype%5B%5D=SV&hail=1.00&lsrbuffer=15&wind=58

I ran the numbers here for 2009-2011 (2005-2011 was taking too long to load). The verification rate is 53.1% for WFO PHI, with a POD of 0.71 and a FAR of 0.47.

Sorry to disappoint you...

This web site only contains preliminary info and should not really be used for verification purposes.

AS an FYI

Since we went to warning by polygon in Oct 2007 the official combined severe warning verification stats thru 5/31/11 for PHI POD is 0.854 FAR is 0.383 and the CSI is 0.558. The initial lead time average is 20.66 min. Tornado only POD is 0.321 FAR is 0.884 CSI is 0.093 and initial lead time is 4.73 min.

The 13 tornado warnings stated in the article are for the entire Mount Holly CWA. I know that a few of the tornado warnings this year that did not verify as tornadoes contained funnel clouds (Lehigh Valley, Colliers Management, Hammonton off the top of my head) its hard to argue that these warnings were not warranted, we're not good enough to absolutely know that the circulation would stop at x amount of feet above the ground.

Lastly we are not continuously scanning the atmosphere, there are 3 to 4 minute gaps at comparable heights with the present radar systems and our more typical type of tornadoes could sometimes slip through those slots. Conversely because of beam width resolution the criteria we use to issue a TOR for Berks County is not the same as for Burlington County. The added problem with places farther away from the radar, you could only go so low and we really can't tell how close to the ground the circulations are getting.

With all of the news coverage and pictures of damage occurring with these storms, even if it was not tornadic or severe imby I would think the warnings would be viewed as warranted. Then again maybe I'm wrong.

[johnc]

Well... I didn't mean to say that they weren't warranted; I tend to be the type that prefers to err on the side of caution anyway. Just that in the context of the article, I think that the limited number and areal coverage (in this region anyway) of TORs makes them much less susceptible to "warning fatigue". I think if the average person around here was in a TOR, they would take it seriously, and most people who have lived here even a long time may only have experienced one or two TORs in that period, if any. By contrast, the sheer number and areal size typically required for SVRs might make desensitization of them more likely amongst the average folk. And if a really nasty one comes through and people don't take the warning seriously -- e.g., by not suspending travel until the storm passes -- they could really be pushing up daisies.

And I apologize for the somewhat-facetious 10% comment. I wasn't so much criticizing the work done by mets as I was lamenting the fact that getting ripped off is so common. I mean... not that I want to see 70mph winds blowing through the woods here. Well... maybe 60-65. And some small hail. I don't care about my car anymore.

[Lady Di]

Not sure if any of you were around and remember this one:

Jul 27 ...1994...

An intense F3 tornado touched down in Limerick, Montgomery Co, PA, passing

through a development called "The Hamlet,"leveling 4 homes, damaging 16 others,

and killing 3 people, all of one family. An F3 tornado moved NE from Avondale to

2 miles SE of town, destroying 6 homes in London Britain Township's Hunter's Run

development. An F2 tornado touched down twice on the Srn end of Kent Island,

Queen Anne's Co, MD, damaging 40 homes. (Officially F0) A tornado struck near

Worton, Kent Co, MD, hurling trees into valuable homes.

I'm sure there were warnings out this night. Yes, this happened at night I think around 10:30-11pm (very rare)-I did not hear a siren and I'm pretty sure my TV was off at this hour. It came and went real fast. Intense is a good word to describe it.

Having experienced this once in my life, I don't have a problem with erring on the side of caution.

[SouthernNJ]

Mt. Holly was a new office in late 1993, and did not take on the current 34 county CWA until

1995 I believe. In 1994, Sterling Va. had warning responsibility for the Easter Shore of

Maryland. As far as the July 27, 1994 event is concerned, it is shrouded in mystery. I only

found out a few things about it from Sterling's Event History Page. Apparently, the Mt.

Holly PUP was down, and they were unable to access KDOX or KDIX on that night. As the

tornadic storms with a history of producing tornadoes in Maryland exited Cecil County and

entered Chester, Sterling used their system to access KDOX and KDIX in order to look at

the storms. They then relayed their warning info to Mt. Holly. By their own admission, they

were unable to find the Avondale tornado before it happened, hence the F3 there only had a

SVR in effect. As the mesocyclone closed in on Downingtown, it became more pronounced

and a TOR was issued at that time. The F1 south of Downingtown had a warning just as it

happened. The Limerick F3 was right on the Chester/Montgomery County Line. A TOR

was still in effect for Chester, but was not issued for Montgomery until after the Limerick

Tornado lifted. In any case, I still would have moved to a safe place if I saw a warning for

Chester and that the storm was heading right for my location. The bad news was that this

event happened just before Midnight. Most folks were asleep.

As far as Mt. Holly is concerned, we cannot accurately gage what really happened because

their equipment was down and unable to track the storms.

In order to find out what actually happened, we would need to ask a meteorologist who was

there at the time this occurred.

[Lady Di]

Mt. Holly was a new office in late 1993, and did not take on the current 34 county CWA until

1995 I believe. In 1994, Sterling Va. had warning responsibility for the Easter Shore of

Maryland. As far as the July 27, 1994 event is concerned, it is shrouded in mystery. I only

found out a few things about it from Sterling's Event History Page. Apparently, the Mt.

Holly PUP was down, and they were unable to access KDOX or KDIX on that night. As the

tornadic storms with a history of producing tornadoes in Maryland exited Cecil County and

entered Chester, Sterling used their system to access KDOX and KDIX in order to look at

the storms. They then relayed their warning info to Mt. Holly. By their own admission, they

were unable to find the Avondale tornado before it happened, hence the F3 there only had a

SVR in effect. As the mesocyclone closed in on Downingtown, it became more pronounced

and a TOR was issued at that time. The F1 south of Downingtown had a warning just as it

happened. The Limerick F3 was right on the Chester/Montgomery County Line. A TOR

was still in effect for Chester, but was not issued for Montgomery until after the Limerick

Tornado lifted. In any case, I still would have moved to a safe place if I saw a warning for

Chester and that the storm was heading right for my location. The bad news was that this

event happened just before Midnight. Most folks were asleep.

As far as Mt. Holly is concerned, we cannot accurately gage what really happened because

their equipment was down and unable to track the storms.

In order to find out what actually happened, we would need to ask a meteorologist who was

there at the time this occurred.

Thank you Southern. After I posted that I did wonder why Sterling had the report for Mt Holly, so now it makes sense to me.

As you said, most of us were in bed or getting ready to go to bed. I do remember there was a threat of bad weather that evening.

This particular tornado crossed Rte 41 about 60 to 70 yards away from my house. It was amazing that not 1 home on our complex property was damaged but we did lose some huge limbs off of every tree. Second amazing thing was that no cars were damaged either. These tree limbs fell and there were leaves blown everywhere but the two cars in front of our place, not one leaf, not a stick, nothing was on these 2 cars! I remember the power going out right before this big wind came. I do not recall hearing "the train" The rain came in the front windows (which were open at the time) sideways. I do recall having a hard time closing these two windows, so at this point I am a little scared. I run to the back part of our house into the bedroom. Couldn't wake the Mr. up at that moment. From this point I can see out the back window and the sky was black, green, and white, colors I doubt I will ever see again. After seeing the damage the next day I believe I was watching the funnel cloud out of my back window. You could tell it wasn't exactly on the ground as it sheered the tops of poles and tree just down the road from us. You could also tell where it did reach ground again going up Toughkenmaon hill where it did some damage at the local airport and it headed off toward U of PA New Bolton Center where I remember hearing someone getting knocked out of bed. He was injured I believe. We were up all night after this storm. My kitchen floor got soaked from the rain that came in sideways.

Luckily, we escaped with no property damage and our power was restored within 36 hours if I remember right.

[greg ralls]

That was probably the closest I've ever been to a tornado. I lived just a little northeast of West Chester at the time, and I'm guessing the meso went more or less over my apartment. It may have been cycling, because I don't remember much in the way of lightning, winds or rainfall. Anyway, I guess I went to bed not long after, and I was surprised - didn't even know there'd been a Tornado Warning for Chester County. I think I was just listening to music or perhaps reading that night, so I didn't have the TV on or anything.