Jump to content
  • Member Statistics

    17,586
    Total Members
    7,904
    Most Online
    23Yankee
    Newest Member
    23Yankee
    Joined

Late warning for Springfield


bobbutts

Recommended Posts

  • Replies 245
  • Created
  • Last Reply

TORNADO WARNING

MAC011-013-015-012000-

/O.NEW.KBOX.TO.W.0002.110601T1928Z-110601T2000Z/

BULLETIN - EAS ACTIVATION REQUESTED

TORNADO WARNING

NATIONAL WEATHER SERVICE TAUNTON MA

328 PM EDT WED JUN 1 2011

THE NATIONAL WEATHER SERVICE IN TAUNTON HAS ISSUED A

* TORNADO WARNING FOR...

SOUTH CENTRAL FRANKLIN COUNTY IN WESTERN MASSACHUSETTS...

NORTHWESTERN HAMPDEN COUNTY IN WESTERN MASSACHUSETTS...

WESTERN HAMPSHIRE COUNTY IN WESTERN MASSACHUSETTS...

THIS INCLUDES THE CITIES OF...NORTHAMPTON...AMHERST...

* AT 327 PM EDT...NATIONAL WEATHER SERVICE DOPPLER RADAR INDICATED A

SEVERE THUNDERSTORM CAPABLE OF PRODUCING A TORNADO NEAR CHESTER...

OR 11 MILES WEST OF NORTHAMPTON...MOVING EAST AT 40 MPH.

* AT 327 PM EDT...NATIONAL WEATHER SERVICE DOPPLER RADAR INDICATED A

SEVERE THUNDERSTORM CAPABLE OF PRODUCING A TORNADO NEAR CHESTER...

OR 11 MILES WEST OF NORTHAMPTON...MOVING EAST AT 40 MPH.

Typhoon Tip @ 03:48PM: That cell NNW of Springfield just seriously hooked on the last scan!

tornadotony @ 03:57PM: Where's the warning?

BULLETIN - EAS ACTIVATION REQUESTED

TORNADO WARNING

NATIONAL WEATHER SERVICE TAUNTON MA

405 PM EDT WED JUN 1 2011

THE NATIONAL WEATHER SERVICE IN TAUNTON HAS ISSUED A

* TORNADO WARNING FOR...

WEST CENTRAL WORCESTER COUNTY IN CENTRAL MASSACHUSETTS...

EXTREME SOUTHEASTERN FRANKLIN COUNTY IN WESTERN MASSACHUSETTS...

EAST CENTRAL HAMPDEN COUNTY IN WESTERN MASSACHUSETTS...

SOUTHEASTERN HAMPSHIRE COUNTY IN WESTERN MASSACHUSETTS...

* UNTIL 445 PM EDT

* AT 359 PM EDT...NATIONAL WEATHER SERVICE DOPPLER RADAR INDICATED A

SEVERE THUNDERSTORM CAPABLE OF PRODUCING A TORNADO NEAR GRANBY...OR

6 MILES SOUTH OF AMHERST...MOVING EAST AT 35 MPH.

CT Rain @ 4:03PM: very impressive couplet now west of springfield

CT Rain @ 4:10PM: hat storm near Southwick/Westfield looks really good. Tight rotation too

tornadotony @ 4:12PM: Major couplet W of Springfield. Over 100 knots gate to gate.

brettjrob @ 4:16PM: Wow... intense couplet evident on both ENX and BOX... hope they torpedo it soon.

CT Rain @ 4:20PM: whoA WESTFIELD/BARNES,MA (BAF) ASOS reports Tornado KBAF 012024Z 30005KT 1 1/4SM R20/4500VP6000FT FC +TSRA BKN024 BKN030 OVC065 26/22 A2984 RMK FUNNEL CLOUD B23 E24 AO2 LTG DSNT NW-E TSB1958RAB00 P0023

brettjrob @ 4:23 PM I really try to shy away from criticizing others' forecasts and warnings, but I'm at a loss on this one... very impressive couplet on multiple 88D's for the past 15 minutes, and it's about to enter a major town.

TORNADO WARNING

CTC003-013-015-MAC013-015-027-012115-

/O.NEW.KBOX.TO.W.0004.110601T2030Z-110601T2115Z/

BULLETIN - EAS ACTIVATION REQUESTED

TORNADO WARNING

NATIONAL WEATHER SERVICE TAUNTON MA

430 PM EDT WED JUN 1 2011

THE NATIONAL WEATHER SERVICE IN TAUNTON HAS ISSUED A

* TORNADO WARNING FOR...

SOUTHWESTERN WORCESTER COUNTY IN CENTRAL MASSACHUSETTS...

EXTREME NORTHEASTERN HARTFORD COUNTY IN NORTHERN CONNECTICUT...

NORTHERN TOLLAND COUNTY IN NORTHERN CONNECTICUT...

NORTH CENTRAL WINDHAM COUNTY IN NORTHERN CONNECTICUT...

EASTERN HAMPDEN COUNTY IN WESTERN MASSACHUSETTS...

THIS INCLUDES THE CITIES OF...SPRINGFIELD...CHICOPEE...

EXTREME SOUTH CENTRAL HAMPSHIRE COUNTY IN WESTERN MASSACHUSETTS...

* UNTIL 515 PM EDT

* AT 426 PM EDT...NATIONAL WEATHER SERVICE DOPPLER RADAR INDICATED A

SEVERE THUNDERSTORM CAPABLE OF PRODUCING A TORNADO NEAR WEST

SPRINGFIELD...OR 6 MILES WEST OF SPRINGFIELD...MOVING EAST AT 40

MPH.

twright86 @4:27PM: tornado on ground! http://www.wwlp.com/...deo/player_page

stebo48858 @ 4:29PM: (reference to live video) Holy crap….
Link to comment
Share on other sites

One vid I saw in SPFD , a guy is surveying his neighborhood helping while his girl films, he tells his neighbor my girl came home and said it looks like a tornado so I turned on the wxchannel and just then the beeping warning came on the same time the tornado hit.

Link to comment
Share on other sites

I think anytime there is an ASOS augmented with FC +TSRA, a computer generated warning should fly which can only be canceled by a person at the helm if there's clear evidence AGAINST a tornado...kinda like the NFL instant replay, you don't overturn unless the call is clearly incorrect.

This is really not a bad idea. No way an augmenter just puts in a tornado, unless he/she is very sure of it happening. If there is clearly NO tornado and they put it in, they can kiss their job goodbye.

Link to comment
Share on other sites

I just don't get it.. was the guy who issued warnings in the bathroom or something? It's not like this popped up suddenly on a clear day, there was a tornado watch, and many of us were already focusing on this cell and it's impressive velocity plots. Add in the ground report and it's a slam dunk imo.

What went wrong?

It's an interesting question... I don't know what went wrong. Looking back at the radar starting at about 4:10 there was strong rotation showing up west of Westfield. By about 4:20 we had nearly 120 knots of gate to gate shear in the lowest slice. At 4:24 BAF reported a FC and the warning went out at 4:30.

These things are never easy and the NWS does a fantastic job. There's always things that can go wrong and it's easy to be sitting here on my couch looking at radar saying holy crap look at that thing. There are all sorts of demands and tasks put on forecasters issuing warnings that could complicate the process especially since they already had a warning out on the Northampton cell which looked very nasty at the time.

OK people...you were NOT sitting in the chair, OK?? Guys, we look at ALL slices, not just the lowest slice to see correlated rotation. Just because it's in the lowest layer doesn't mean it is going to drop a long lived tornado.

FYI, our radar was switched over to VCP212 to give us as many slices as quickly as it can. However, there is a drawback. It's called range folding. And, guess what??? That storm was in the range fold. We didn't have the luxury of time to fix the range fold area and, from what the radar operators (two of my colleagues told me) it may not have helped. I was working the short term desk, not warnings, but I did hear that AS SOON AS IT CAME OUT OF RANGE FOLD and saw how much rotation the storm had, we issued.

For another thing, we are NOT staring at the OBS every minute of our shift, we have MANY other responsibilities to handle. Luckily, I saw messages on NWSchat we now have at the office, and I did relay the FUNNEL CLOUD report from BAF to the office as soon as I saw it!!! Plus, we had communications with WWLP directly on NWSchat and they sent the message over about the tornado forming right near their studios on the Connecticut River on their webcam. Once we could get the streaming video to go (our Internet connection is terrible most of the time) happened to coincide to when the storm came out of range fold. This was the FIRST warning issued for this storm, and, even though other TORs were issued prior to this one, it can still be a bit hard to pull the trigger especially for such a large city as Springfield. But, considering everything that was going on 6/1, we did a pretty good job, IMHO. We had lead time of up to 54 minutes on one warning downstream of Springfield, which probably saved MANY LIVES.

I'm going to be starting another thread on my project for today (6/2). I was a part of the tornado survey team that toured through Monson, Brimfield, Sturbridge and Southbridge. Once the final determination is issued, I think you will be very surprised with our findings. Two of my colleagues and myself (including one of the people that was operating radar and warnings yesterday) arrived in Monson at about 1030 AM, and didn't arrive back at the office until around 930 PM. A very long and unbelieveable day. I can tell you this...and my 26 years in the NWS, I NEVER thought that I would see this type of damage in my home state. It was simply AMAZING. More tomorrow...

I hope this explains things. For those that will still yap and complain, all I can say is...if you could have the opportunity to sit in the chair and issues these type of warnings, I think you would take a MUCH different perspective on the ramifications of these decisions. We DO NOT take issuing tornado warnings lightly. We know the importance of getting these as best we can given our abilities, capabilities and experience. I know that I think about the lives I can possibly save by my actions whenever I issue warnings. It's heartbreaking to hear of 4 people that lost their lives, and the stories I heard today from people that lost their homes. Our #1 job in the NWS is "the protection of life and property," and we take that VERY SERIOUSLY!!!

'nuff said.

--Turtle ;)

Link to comment
Share on other sites

Disagree with Tip, agree with most everyone else. The same issues came up with the Rogers, MN event five years ago - a tornado fatally injured a young woman as the NWS forecaster was writing the warning. From that several recommendations were made - including adding "severe thunderstorms can produce tornadoes without warning" (something like that). I think what may have transpired was a) with only a 5% tor risk out, forecasters perhaps believed rotation was too far aloft or just a temporal feature, so they waited a scan or two to decide further; b ) after a couple of scans to confirm the persistence of the couplet, they began writing the warning but c) if more forecasters were manning the northern storms, switching focus and trying to add the specific warning text and towns may have taken a few minutes to do. By then, the storm would have been nearly over Springfield. (This was, in fact, nearly the sequence of events described in the Rogers service assessment that led to the late warning there.)

NWS in Maine was issuing tor warnings for "severe thunderstorms with potential to produce tornados" (which I assume means no ground truthing was available), with both GYX and CAR issuing such warnings for storms 70-80 miles from their respective radars. At least one (F1 in Embden, NW of Skowhegan) has been confirmed.

...and all the other quotes...

Well pardon me and my colleagues for existing! A lot of the posters here seem to have a strong perception of: a) What happened in an office that none of them were in, and B) How NONE of it would have happened if only the NWS had had the wisdom to give THEM the keys to the AWIPS warning workstation.

This is America, and you all have the right to vent as you please. But from the view of someone who was there at the time (although not on the radar desk), your negative comments come across as the grumblings of a group of backseat drivers.

Here are a few more items to consider in addition to the indignation.....

252pm Tornado Warning issued for Cheshire County ... subsequent ground truth ... several reports of large hail

309 pm Severe Tstm Warning issued for several segements of NW MA ... subsequent ground truth ... two reports of straight line wind, more reports of large hail

324 pm Severe Tstm Warning issued for parts of Cheshire and Hillsborough .. subsquent ground truth ... reports of large hail

328 pm Tornado Warning issued for parts of Franklin-Hampshire-Hampden ... subsequent ground truth ... reports of large hail

332 pm Severe Tstm Warning issued for parts of Worcester-Middlesex-Cheshire ... subsequent ground truth ... Nothing!

405 pm Tornado Warning issued for parts of 4 counties around Quabbin ... subsequent ground truth ... Large hail, one report of wind damage in Hadley at 415 pm.

And then there were...

418 pm Severe Tstm Warning issued for parts of several counties including Springfield-Holyoke-Westfield. This warning included a statement to the effect that a tornado was possible.

430 pm Tornado Warning issued for Springfield, et al. We all now know what happened.

We know how to read couplet signatures at Taunton. We also know that proper Warning Decision making goes beyond just looking at the SRM. For example, incorporating ground truth. Three tornado warnings with no tornadoes to show for the effort.

The site surveys that occurred today were just a part of the review process on this storm. We will learn what we can from it and move on. Without righteousness.

Link to comment
Share on other sites

On the topic of late warnings, there was a severely late tornado warning on June 23, 2010 for Simcoe County, Ontario.

At 630pm a stovepipe F2 tornado touched down west of Midland tracked through the city injuring 20 and causing $15million in damage. It then dissipated after 25km of being on the ground. A second tornado rated F1 developed near Washago at 7pm and tracked for 10km and dissipated just a hair west of the town.

That evening:

At 642pm a severe thunderstorm warning was issued for northern Simcoe County. This is 12 minutes after the tornado initially touched down and was likely in the city at that point.

At 705pm a tornado warning was issued for northern Simcoe County, including Midland, 35 minutes after the tornado hit the city, 5 minutes after the second tornado touched down, and after the supercell tracked for 30 miles.

Strangely enough, a group of us were commenting on how there looked to be a hook feature on King City radar as the tornado was happening, and on Buffalo radar there was rotation high in the storm from 130 miles away.

Link to comment
Share on other sites

...and all the other quotes...

Well pardon me and my colleagues for existing! A lot of the posters here seem to have a strong perception of: a) What happened in an office that none of them were in, and B) How NONE of it would have happened if only the NWS had had the wisdom to give THEM the keys to the AWIPS warning workstation.

This is America, and you all have the right to vent as you please. But from the view of someone who was there at the time (although not on the radar desk), your negative comments come across as the grumblings of a group of backseat drivers.

Here are a few more items to consider in addition to the indignation.....

252pm Tornado Warning issued for Cheshire County ... subsequent ground truth ... several reports of large hail

309 pm Severe Tstm Warning issued for several segements of NW MA ... subsequent ground truth ... two reports of straight line wind, more reports of large hail

324 pm Severe Tstm Warning issued for parts of Cheshire and Hillsborough .. subsquent ground truth ... reports of large hail

328 pm Tornado Warning issued for parts of Franklin-Hampshire-Hampden ... subsequent ground truth ... reports of large hail

332 pm Severe Tstm Warning issued for parts of Worcester-Middlesex-Cheshire ... subsequent ground truth ... Nothing!

405 pm Tornado Warning issued for parts of 4 counties around Quabbin ... subsequent ground truth ... Large hail, one report of wind damage in Hadley at 415 pm.

And then there were...

418 pm Severe Tstm Warning issued for parts of several counties including Springfield-Holyoke-Westfield. This warning included a statement to the effect that a tornado was possible.

430 pm Tornado Warning issued for Springfield, et al. We all now know what happened.

We know how to read couplet signatures at Taunton. We also know that proper Warning Decision making goes beyond just looking at the SRM. For example, incorporating ground truth. Three tornado warnings with no tornadoes to show for the effort.

The site surveys that occurred today were just a part of the review process on this storm. We will learn what we can from it and move on. Without righteousness.

Totally understand and thanks for the posts Bill and Eleanor. None of us were sitting in the hot seat or seats at the time and should not Monday morning quarterback. There's something to be learned from every event. No met is perfect and the science is not perfect.

I certainly don't fault anyone and think everyone did as best as they could in the situation. The storm being so far from the radar site certainly didn't help things and neither did the fact there were other multiple supercells in the CWA at the time. It was unfortunate that the one storm with tight rotation that wasn't warned as fast wasn't warned and others were... really not much the forecaster can do there.

For the most part I think the posts in this thread have been respectful and have been just asking questions about why a warning wasn't issued sooner (which Eleanor explained very well) and what all of us (mets and hobbyists) can learn about this entire event.

Link to comment
Share on other sites

Remember also the beam at that distance (on the base scan) is ~6500 feet off the ground. Combine that with RF issues and a logical issuer would wait for at least like a funnel cloud report before shooting off the warning while still in RF. Once out of RF you could probally gauge with velocity's and other conditional parameters (LCL, Shear etc.) whether a warning was warranted. Fortunately, dual-polarization should fix the problem with waiting for velocity scans and guessing if there is a tornado, the Correlation Coefficient would point out a tornado on the ground.

Link to comment
Share on other sites

Thank you Eleanor and Bill for the explanations. I think we are all in the same boat when we are concerned for the well being of the public and you have an important responsibilty that many of us do not have. I think as difficult as these discussions and debates may be, I think they are posotive and a good learning experience. Tuesday was a rare and uncommon event.

Link to comment
Share on other sites

To clear one thing up -- an ASOS cannot report a tornado or a funnel cloud. Those are items that must be manually added by an observer at the weather station in question; they get tacked on to the RMK section of the METAR for transmission. Most airport observers are highly trained, but even then it would be hard to consider such a report to be completely infallible.

sorry, I knew that, and should have been more careful with the post. I believe the report was from the airport in Westfield.

Yes, the observer does augment the OB to put in that remark. When that goes into the remark, TORNADO shows up in the present weather!! I saw the TORNADO and funnel cloud (FC) in the OB.

For those of you that didn't read the remarks further, the FC began at 23 after the hour, and ENDED at 24 after. It lasted a grand total ONE MINUTE!!! It did remain in the OB much longer, likely because ASOS didn't pick up the ending time for some reason. I noticed it didn't clear the FC, so I called the observer at BAF as soon as I noticed it. They sent a special to take it out nearly immediately.

Another FYI.

--Turtle ;)

Link to comment
Share on other sites

Remember also the beam at that distance (on the base scan) is ~6500 feet off the ground. Combine that with RF issues and a logical issuer would wait for at least like a funnel cloud report before shooting off the warning while still in RF. Once out of RF you could probally gauge with velocity's and other conditional parameters (LCL, Shear etc.) whether a warning was warranted. Fortunately, dual-polarization should fix the problem with waiting for velocity scans and guessing if there is a tornado, the Correlation Coefficient would point out a tornado on the ground.

True, if there is debris being sampled by the radar.... By the time the TDS shows up, the tornado is on the ground and doing significant damage, which clearly would still upset the people in this thread... I mean if it only scans every 5 or 6 minutes, it might not improve on the lead time that much in a situation like this, really.

Link to comment
Share on other sites

For another thing, we are NOT staring at the OBS every minute of our shift, we have MANY other responsibilities to handle. Luckily, I saw messages on NWSchat we now have at the office, and I did relay the FUNNEL CLOUD report from BAF to the office as soon as I saw it!!! Plus, we had communications with WWLP directly on NWSchat and they sent the message over about the tornado forming right near their studios on the Connecticut River on their webcam. Once we could get the streaming video to go (our Internet connection is terrible most of the time) happened to coincide to when the storm came out of range fold. This was the FIRST warning issued for this storm, and, even though other TORs were issued prior to this one, it can still be a bit hard to pull the trigger especially for such a large city as Springfield. But, considering everything that was going on 6/1, we did a pretty good job, IMHO. We had lead time of up to 54 minutes on one warning downstream of Springfield, which probably saved MANY LIVES.

I'm not trying to make this personal, but your post reveals several huge problems with the NWS:

1) Why is your Internet connection terrible when you're in charge of assimilating various radars, videos, and ground observations to protect lives and property from dangerous weather? Shouldn't the NWS have the fastest access possible in order to ensure that all pertinent information is viewed and relayed to the appropriate officials and the general public as soon as possible? I just see something a little bit wrong with a hobbyist being able to see a video of a tornado on the ground while a NWS officer waits for it to load...

2) When there are discrete supercells in an area, why aren't ground observations given the utmost priority? I mean who cares what the radar shows if someone from the NWS is telling you there's a massive funnel cloud with +TSRA when you know you're tracking a threat with high shear, high lapse rates, and extreme instability? What is more important than communicating this ob to the public quickly?

...and all the other quotes...

We know how to read couplet signatures at Taunton. We also know that proper Warning Decision making goes beyond just looking at the SRM. For example, incorporating ground truth. Three tornado warnings with no tornadoes to show for the effort.

So what? I'd rather sit in my basement watching TV for 10 minutes even if it's just heavy rain than be killed by an unwarned tornado. Many tornado warnings are not going to verify with massive F3+ twisters hitting populated areas, obviously. It's a warning, not a forecast that needs to be verified. You don't think it's time to pull the trigger when BAF is reporting FC +TSRA in an area with like 3000J/kg CAPE, huge shear values, a strong low to the north with associated height falls in early June, and a history of supercells developing that afternoon? I can understand that you don't want to give constant false alarms, but it seems like erring on the side of caution is best. Most members of the general public aren't going to blame you if you issue a tornado warning and they get funnel clouds with 2" hail stones and straight-line winds...they're going to be grateful NWS alerted them of danger.

And if you know how to read couplet signatures, then why did most people on the forum, many of whom are just hobbyists, recognize that this was surely a tornado? The radar was showing 120kts g2g, reflectivity over 50dbz, hook/bow echoes, etc. At the very least, given the radar presentation, shouldn't BAF's ground ob automatically trigger a warning, whether it's by computer or by hand?

Link to comment
Share on other sites

Thank you Eleanor and Bill for coming and explaining what happened. Just a question for future reference.... I watched this unfold on ALB radar. Since the warned area was in the range fold....would it be against protocol to take a look at ALB radar and try and make a determination if there were ground reports?Thank you guys in Taunton for all that you do. Looking forward to seeing the survey.

Link to comment
Share on other sites

I thought the public data was a different level than what the NWS offices see. Don't they have the same kind of data that these pay services offer which is better resolution and that show the shear etc?

AFAIK, NO, we do not send out different levels, just not as many as we have access to. I believe it's mainly because of a bandwidth issue. Not terribly sure of this, however.

Why would a fine discussion about the lack of advance warning be perceived as dubiously self serving? Improving performance of advanced warning and getting the message out without false alarms should be priority one. I have a right to expect the best from a taxpayer point of view. I also can express my dissatisfaction, which I have not or have any. USA not CHINA

Absolutely, Ginxy. And I hope we do provide this to the taxpayers.

--Turtle ;)

Link to comment
Share on other sites

Thank you Eleanor and Bill for coming and explaining what happened. Just a question for future reference.... I watched this unfold on ALB radar. Since the warned area was in the range fold....would it be against protocol to take a look at ALB radar and try and make a determination if there were ground reports?Thank you guys in Taunton for all that you do. Looking forward to seeing the survey.

I saw this mentioned earlier, and yes, KENX radar is certainly an option. Since I was not working the radar desk, I don't know if they were looking at KENX radar. I can tell you things were happening fast and furious at the office around the time of the touchdown in Springfield/Westfield, and kept up until I left at 8 PM (and I heard longer). However, as someone else mentioned in this thread, both KBOX and KENX radars to cut off the lower 5-6Kft of the storm. This does make a big difference in getting the whole structure of the storm, but we have a bit more of a problem with this in low topped convection. Also, I don't know what VCP they were operating in. OK, yes, they saw the rotation. And yes, we can use KENX, KGYX and KOKX to help us. But again, things were happening quickly, and we did the best we could to keep up. Bill was on the long term desk and shift supervisor, so he was monitoring things as best he could too.

As many other said here, this review process, the storm surveys (there were two teams out there yesterday) along with a event review at the office will only help us get better for the next time.

--Turtle ;)

Link to comment
Share on other sites

I saw this mentioned earlier, and yes, KENX radar is certainly an option. Since I was not working the radar desk, I don't know if they were looking at KENX radar. I can tell you things were happening fast and furious at the office around the time of the touchdown in Springfield/Westfield, and kept up until I left at 8 PM (and I heard longer). However, as someone else mentioned in this thread, both KBOX and KENX radars to cut off the lower 5-6Kft of the storm. This does make a big difference in getting the whole structure of the storm, but we have a bit more of a problem with this in low topped convection. Also, I don't know what VCP they were operating in. OK, yes, they saw the rotation. And yes, we can use KENX, KGYX and KOKX to help us. But again, things were happening quickly, and we did the best we could to keep up. Bill was on the long term desk and shift supervisor, so he was monitoring things as best he could too.

As many other said here, this review process, the storm surveys (there were two teams out there yesterday) along with a event review at the office will only help us get better for the next time.

--Turtle ;)

Thank you...have definitely learned a lot past couple days about a lot of stuff that I had no clue about.

Link to comment
Share on other sites

I'm not trying to make this personal, but your post reveals several huge problems with the NWS:

1) Why is your Internet connection terrible when you're in charge of assimilating various radars, videos, and ground observations to protect lives and property from dangerous weather? Shouldn't the NWS have the fastest access possible in order to ensure that all pertinent information is viewed and relayed to the appropriate officials and the general public as soon as possible? I just see something a little bit wrong with a hobbyist being able to see a video of a tornado on the ground while a NWS officer waits for it to load...

2) When there are discrete supercells in an area, why aren't ground observations given the utmost priority? I mean who cares what the radar shows if someone from the NWS is telling you there's a massive funnel cloud with +TSRA when you know you're tracking a threat with high shear, high lapse rates, and extreme instability? What is more important than communicating this ob to the public quickly?

1.) Our big problem with bandwidth is because we operate on what's called NOAAnet. We are hooked up with our entire Eastern Region, I believe that's a total of 27 offices through our regional headquarters in Bohemia, NY. This includes NYC, us, Philly and Washington, DC. This can get very bogged down (remember this past winter when there were major server problems??), and makes things very slow. OK, maybe not yesterday, but we do know that our website (weather.gov/boston) is nearly at the top or actually the top in Internet "hits," especially for a winter storm. It is VERY frustrating for us, but we do have AWIPS to look at most things. And yes, it is somewhat embarrassing (IMHO) that you guys at home can get stuff quickly over the 'net than we can at KBOX. Also, there's the old issue of $$$. Increasing bandwidth ain't cheap, and I'm sure you know about the gov't budget situation. It isn't going to get better anytime soon. Can't come up with a solution for this one.

2.) Yes, ground truth is of utmost importance. BUT, when the FC only lasts ONE MINUTE?? And, I wasn't working the radar to see if there was continued rotation. It IS of utmost importance to get this out quickly, of course. Yes, we still do the warnings "by hand" so we can read over the warning before sending them. But, if we issued TORs with every one minute FC seen, and automatically to boot, we'd have a HUGE FAR. (And yes, we are being scrutinzed for our FARs and PODs.) Granted, the atmospheric environment was totally different than a "run of the mill" New England thunderstorm threat, OK. This was a totally different animal for us yesterday. I do have to give SPC a HUGE :thumbsup: :thumbsup: . They had this event NAILED up to 5 days out on their experimental convective outlooks on their website, along with the EMLs that were mentioned. Maybe we didn't give this enough credence, I don't know. Again, something to look at during the event review.

On another note, I wasn't taking this thread and some comments personally (well, maybe I was a bit). I'm just really sick and tired of the "Monday morning quarterbacking," as Ryan said in his post. It's really VERY FRUSTRATING to read this stuff after working this huge event, knowing that we did the best job we could to get those warnings out as quickly as possible, with the perverbial you-know-what hitting the fan. I do apologize if I did come across that way in my original post.

--Turtle ;)

Link to comment
Share on other sites

Thank you both, Eleanor and Bill, for taking the time to explain some of the process that goes into issuing warnings. I appreciate not only your knowledge and experience, but your passion and determination to be as timely and as accurate as possible. I feel that pretty much all the posts in here echo those sentiments.

To a layperson, it would seem that one of the "issues", for lack of a better term, might be insufficient hi-res radar coverage in certain areas. To the average John Q. Public, it's generally believed that the NWS has a radar dish pretty much on every city block, and that every tiny change in the atmosphere is picked up almost instantaneously. Terms like range-folding are foreign to the average person.

Do you feel that this is a fair assessment? Would more extensive radar coverage have led to a warning being issued more quickly?

Link to comment
Share on other sites

There are probably slight differences from office to office, but in terms of most places, here's how radar data gets into AWIPS:

For on-site radars, with a dedicated Radar Product Generator (RPG) running at the office, all of the data flows into AWIPS in real-time. For example, Taunton would get the radar data from KBOX literally as the scans are occurring. That's 8-bit data, super-res at the lowest few slices, with as high a quality and resolution as an external user would see in Level II data (such as with GR2Analyst).

For a handful of nearby off-site radars, a limited amount of data is streamed in over the Satellite Broadcast Network (SBN) -- basically, think of the SBN as a permanent data feed coming into the office. The SBN is also used for the majority of other products coming into AWIPS (models, satellite, etc.). It's fairly quick for radar data, but not quite in real time. This data is also 8-bit, but is limited in other ways. For example, no super-res data is sent through, and only a few of the lowest slices are available.

This data can be augmented by running a Radar Multiple Request (RMR) -- an AWIPS procedure that connects to the remote radar site and requests in additional products. These requests are totally customizable, but for example, one might ask the RMR to acquire super-res data in the lowest few slices where it's available, and a few additional slices going up too. The amount of data one can request through an RMR is, as one might expect, also limited.

The amount of data sent over the SBN has actually improved quite a bit over the last few years. In an ideal world, the entire suite of radar products from all surrounding radar sites would be automatically delivered in real time. However, the bandwidth and funding issues have already been mentioned in this thread.

Link to comment
Share on other sites

I'm curious about the number of trained spotters who were available in the Springfield area (and the area to the west) at the time. Might be worth actively recruiting more spotters in areas that are problematic for low level radar coverage. Probably not ever going to have a better time to recruit if the NWS wants to expand its trained spotter network.

Just sent this link to a friend that works at NECN: http://www.weather.gov/skywarn/

Link to comment
Share on other sites

Yes, the observer does augment the OB to put in that remark. When that goes into the remark, TORNADO shows up in the present weather!! I saw the TORNADO and funnel cloud (FC) in the OB.

For those of you that didn't read the remarks further, the FC began at 23 after the hour, and ENDED at 24 after. It lasted a grand total ONE MINUTE!!! It did remain in the OB much longer, likely because ASOS didn't pick up the ending time for some reason. I noticed it didn't clear the FC, so I called the observer at BAF as soon as I noticed it. They sent a special to take it out nearly immediately.

Another FYI.

--Turtle ;)

Exactly. This is an important point. The field offices are evaluated on both detection and false alarms. So we can't just throw up a bunch of warnings and hope one or two will stick. We need to be able to justify, if only to ourselves, that there is a fair likelihood that warning criteria will be met.

I'm not trying to make this personal, but your post reveals several huge problems with the NWS:

1) Why is your Internet connection terrible when you're in charge of assimilating various radars, videos, and ground observations to protect lives and property from dangerous weather? Shouldn't the NWS have the fastest access possible in order to ensure that all pertinent information is viewed and relayed to the appropriate officials and the general public as soon as possible? I just see something a little bit wrong with a hobbyist being able to see a video of a tornado on the ground while a NWS officer waits for it to load...

2) When there are discrete supercells in an area, why aren't ground observations given the utmost priority? I mean who cares what the radar shows if someone from the NWS is telling you there's a massive funnel cloud with +TSRA when you know you're tracking a threat with high shear, high lapse rates, and extreme instability? What is more important than communicating this ob to the public quickly?

(regarding the warning decision process)

So what? I'd rather sit in my basement watching TV for 10 minutes even if it's just heavy rain than be killed by an unwarned tornado. Many tornado warnings are not going to verify with massive F3+ twisters hitting populated areas, obviously. It's a warning, not a forecast that needs to be verified. You don't think it's time to pull the trigger when BAF is reporting FC +TSRA in an area with like 3000J/kg CAPE, huge shear values, a strong low to the north with associated height falls in early June, and a history of supercells developing that afternoon? I can understand that you don't want to give constant false alarms, but it seems like erring on the side of caution is best. Most members of the general public aren't going to blame you if you issue a tornado warning and they get funnel clouds with 2" hail stones and straight-line winds...they're going to be grateful NWS alerted them of danger.

And if you know how to read couplet signatures, then why did most people on the forum, many of whom are just hobbyists, recognize that this was surely a tornado? The radar was showing 120kts g2g, reflectivity over 50dbz, hook/bow echoes, etc. At the very least, given the radar presentation, shouldn't BAF's ground ob automatically trigger a warning, whether it's by computer or by hand?

1. In addition to what Eleanor has said on this subject, the data flow into our office is actually for TWO offices...BOX and the Northeast River Forecast Center. We could use more bandwidth, but we're not holding our breath given the current fiscal situation.

2. The BAF observation is mostly automated, with augmentation by FAA controllers in the tower. There are no NWS people at the various observation sites, and have not been for many years. The observation that come through is funnel cloud, not "massive funnel cloud". And as Eleanor noted, the Funnel Cloud was gone as fast as the observation came out (1 minute duration).

3. "high shear, high lapse rates, extreme instability" We had had that all afternoon. And as noted in my previous post, we had previously issued 3 Tornado Warnings...and had no reports of tornadoes or funnel clouds or wall clouds. You had 120 kts gate to gate. Well, bully for you. We had purple haze. (I had forgotten that, thanks to Eleanor for bringing that up.) Based on the information we had at the time, our radar/warning operator went with a Severe Tstm Warning that mentioned the possibility of a tornado. That went out at 418 pm. Once the couplet moved out of the haze, and the couplet became viewable, the warning was upgraded to a Tornado Warning.

4. A supercell in New England carries a high likelihood of ANY severe weather, but does not guarantee a tornado. Please refer to the Northern MA supercell of last summer, or the CT-Nrn RI-SE MA supercell of August 10 2000; both showed couplets and incredibly frequent lightning, but neither produced a tornado. Couplets are important, but they are not infallible as some posters in this thread seem to be implying.

5. "rather sit in my basement" Congratulations. Others are not so forgiving. They will complain about our unnecessarily frightening them, and the population of a major city, for a phantom tornado. And then next time they will be less likely to react. We will warn for what we feel is justified based on the information that we have.

6. "It's a warning, not a forecast that needs to be verified." Warnings are verified, watches are verified, forecasts are verified.

7. Earlier quote: "CT Rain @ 4:20 pm: whoA Westfield/Barnes MA (BAF) ASOS reports Tornado KBAF 012024Z [etc] FUNNEL CLOUD B23 E24"

Think this is a typo. How can a 4:20 pm entry be quoting an observation from 2024Z...which is 4:24 pm...and citing a funnel cloud that began at 423 pm and ended at 424 pm?

Link to comment
Share on other sites

^^^The quote from CT Rain bears the timestamp for his post on this board, and the board's timestamp is likely off by several minutes.

__________________________________

In the performance metrics for the NWS, its clear to me that greater weight is given to the lead time metric than to the false alarm metric.

The lead time circa 1996-1997 was about 10 minutes, and this has slowly increased over the past 15 years. The false alarm rate has generally remained in the upper 70th percentile over the period, and peaked (over the last 15 years) at 80 percent in 1998. This suggests the earlier the warning, the greater the likelihood for a false alarm, and the converse holding true.

Since 1996, the accuracy performance metric has seen the greatest gains: from 60 percent or less to close to 80 percent.

The deficiency in the public metrics for lead time and accuracy is that these are not sub-aggregated by tornado intensity: if I were guiding the NWS on its metrics, I would suggest subcategories for tornadoes rated EF0-EF1, EF2-EF3. and EF4-EF5. It may be that the NWS already sub-aggregates internally, but aggregates to a single number publicly.

Link to comment
Share on other sites

Hopefully, events like these prove that we can't cut money from the NWS budget.

This x 1000

NWS does a great job for the public. I would imagine there have been studies done showing what the value of their work is versus the amount of taxpayer dollar that go into it. I would imagine something like a 10x or higher return on investment.

Anyway, I appreciate what they do and what the NWS posters (Turtle and Bill) have to say on this. I am a newbie hobbiest at this and have had very little background and training (1 grad level course in Met). I don't know what the criteria they have to meet for giving a warning is. I hope they don't take offense to any of this (looks like that has already happened). Everyone can view this as a learning experiece. Everyone. Even Blizz.

Thanks for all you do.

Link to comment
Share on other sites

This x 1000

NWS does a great job for the public. I would imagine there have been studies done showing what the value of their work is versus the amount of taxpayer dollar that go into it. I would imagine something like a 10x or higher return on investment.

Anyway, I appreciate what they do and what the NWS posters (Turtle and Bill) have to say on this. I am a newbie hobbiest at this and have had very little background and training (1 grad level course in Met). I don't know what the criteria they have to meet for giving a warning is. I hope they don't take offense to any of this (looks like that has already happened). Everyone can view this as a learning experiece. Everyone. Even Blizz.

Thanks for all you do.

The criteria for verifying a tornado warning is a verified tornado. I would THINK a storm that produces a funnel cloud or a waterspout SHOULD be warned, but unless it produces a tornado, it doesn't count, as far as verification processes.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.

×
×
  • Create New...