2 major busts in the same storm.....did the first one cause the second?

[weathafella]

Human nature is funny even among scientists.  If we break down what happened during the past week with a major and anomalous storm we see why the initial error may in fact have led to the second one.  I'm talking initially the DC/BWI bust (rained or slop in the major metros vs snow further inland and elevated) and the SNE bust (overperforming snow in all areas exclusive of a large swath of RI).

 

Models were pretty bullish on a dynamically driven blue bomb for DC/BWI.  The night before temps were above freezing but not way above.  The thinking of course was the blocking would cause the storm to wind up off the coast of lower VA/NC and rather than move NE, get blocked and squeezed more easterly.  All of this happened!  NAM/GFS had a good snow signal.  Euro (to me) looked a little suspect in the BL but I figured it was an error which often happens in marginal events.  Unfortunately, I don't have access to Euro 925 temperatures but those doing this for a living probably do.  According to Will, 925s were very close ...0 to -1C.  BL was 2-3C.  Arguably, if rates were really strong, it would do the job.  

 

As the day kept going by with Cantore in DC and mix occurring with little or no accumulation, DC itself was in trouble.  Ironically, areas around RIC were getting snow as the dynamics appeared better there.  West of DC including the hills of VA/WV did quite well actually but the majority of the population lives in an area that was forecasted to have heavy snow and didn't get it.   It appears (to me) that the Euro was not weighted heavily enough)

 

Meanwhile, the Euro was very bullish on a big SNE hit despite a storm way SE of the BM.  The upper low to the NW was to phase and Fujiwara as the coastal was blocked from moving much.  The Euro was showing cold BL to the coast for the bulk of the storm especially Thursday night into Friday.  GFS was showing a drawn out period of modest qpf that wouldn't do the trick. NAM to a degree had the idea but they kept pushing H85s  way west but a thin layer.  The model never really resolved that anomaly and no other model had it.   Meanwhile CMC was quite bullish on a major snow dump close in.  Another model that shat the bed was crazy uncle UKMET.  The highly respected mets (some) like it but it seems pretty bad for big coastal storms close in.  

 

So you're a met and you busted in DC, are worried about the public blaming you for over-hyping (OPM closed government office effectively for a rainstorm and weenies couldn't enjoy it as they were so pixxed off!).   You also are mindful of what happened in DC.  You are sitting near Boston and noting that way up in Caribou and Quebec its above freezing.  The Euro insists that it is not an issue but you busted in DC and don't want to do it again.  You know it's into March and start buying into the sun angle argument (which doesn't work when 2m temps are below freezing and heavy snow is falling).   So while you believe the Euro, you remember it was too cold on an event 2 weeks earlier and adjust it warm.  You blend the models, leaning Euro.  You call for good snows elevations and well away from the coast.

 

Essentially you bust twice.  The first time for buying the more aggressive snow solutions in DC, the second time for NOT buying the aggressive snow solutions in Boston.  

 

Had the GFS been discarded in both cases, 2 busts may have been 2 correct forecasts.  Had the close in models getting more aggressive with snow (RPM, HRR, RAP) as it got closer been at least given credence, forecasters come out like stars.  Weenies in DC get mad at them for not calling for snow but grudgingly acknowledging them post mortem.  Weenies here think they're nuts but acknowledge the possibilities.  

 

So here in Boston, Harvey came off vacation and realized it would be more snow and was reasonably close.  Matt Noyes has a coastal mix fetish and that may have clouded his judgement, otherwise, close in her was ok for MA but bad for CT.  Pete Bouchard want balls to the wall negative and I'm wondering if he'll pay with his job.

 

NWS actually did ok.  They had warnings up and upped their maps early.

 

But everyone worrying about the DC  over bust under called Boston imho.  

 

Remember Al Roker being laughed at when he said 18-24 for the Boston area 2 days out?  He was pretty close!

 

I don't think any mets (well some of our red taggers knew the possibility) believed the firehouse would be so special. 

 

 

I'm not a met.  But in my years of following the models, I've learned that nearly all the time when the Euro doesn't budge inside 3 days, buy it.

[Ginx snewx]

yep said it yesterday, The DC bust scared the sh it out of most of them.

[SR Airglow]

It seems that once DC busted, the mets, at least in NYC, were terrified of having the same thing happen to them. Even at 6 thursday night, many only had 1-3” for the city, despite all the models showing an advisory event or better. Everyone in the NYC forum was basing Lonnie Quinn(Our resident hyper in NYC, knows little to nothing about the actual weather and models...) for going 6-12” and hyping it up, turns out, he was the only one who didn’t bust badly, even Upton busted pretty badly. I get being conservative, and at least NYC only got 6”, but calling for nothing and getting close to 2’? That’s inexcusable, especially since basically every model had an accumulating snow in boston. I’m fine with calling for something like 5-7” and saying potential for much more, but to not even highlight the potential leads to a total mess, exactly like what happened in BOS with the schools... The mets were scared that what happened in DC would happen to them, turns out that it did, just in a different way...(Harvey Leonard was really the only met in BOS that didn’t bust terribly, although he was too low, he mentioned potential for a big storm, and that redeems him in my eyes...)

 

-skisheep

[ORH_wxman]

I'm sure the DC bust was in the back of everyone's mind...it certainly was in the back of my mind as well as we got within 36 hours. I definitely think many did not believe the Euro QPF of 2"+ for a chunk of BOS and SE MA and 1"+ back to ORH. It almost felt like many were doing what we normally do with the NAM...take the qpf and slice it by maybe taking 60% of what it shows....draw it out over a long period...and voila! you have a less intense precip rate with marginal temps and a mainly nuisance event outside of the hills. Probably not a good move when it's the Euro inside of 48 hours...and not just one rogue run either, it did this for like 4 out of its final runs, only have one slightly less bullish run (still huge for SE MA) on the 00z Mar 7 run...but then the ensembles later that night said the OP had burped, and so did the synoptics in general.

 

The Euro had shown the fully saturated H7 firehose getting up into SE NH like every run inside of 48/60 hours. And never waivered. Sure the qpf oscillated slightly, but for the QPF-fetish people, they need to understand models handle it the most poorly of any parameter...certainly more than they handle somethink like an H7 firehose.

 

The temps were a bit of a bigger issue IMHO. Though they are not mutually exclusive from the QPF/precip rate. BOS was a tough call...even at 24 hours out. They always are being on the water. But the Euro was insistant on 925mb temps of like -4C to -5C even right there. Its almost impossible to rain in that setup unless it is early December with onshore flow when the water is still torching or if the precip is barely falling....neither scenario applicable. I thought the more egregious bust was in the suburbs and of course interior hills. It seemed very little respect was paid to interior SE MA from a public standpoint until Harvey got on the air and drew the 5-9" contour in there while explaining it could be much higher.

 

 

All in all, this storm will be a great learning experience for all regions it affected. In the end, nobody got it right. In the post mortem in the M.A. forum, I told them from everything I was looking at, if I had been professionally forecasting for DCA/BWI, I would have gone down in flames and busted horribly too. Up here, I stayed bullish...but I was still way too conservative in the end.

[CoastalWx]

I was more concerned with the QPF since that would eventually help govern temps in the city. When the euro has the bulleyes just se of the city..my concern was a slight tick SE and it's almost a non event outside of a few sloppy inches. The cumuative QPF was rather high, but over 48 hrs. You really need to get these 1-2" QPF bombs in 24 hrs and that's what ended up happening. By Wednesday Night and Thursday, this became a reality and it was time to be more bullish. I was not concerned about wind direction as all models had winds backing to more north. Take into account that the winds probably would be a bit more ageostrophic than models had...no problem there. We needed those heavy rates and needed a chunk of it to occur at night. That's exactly what we had, and it just ripped through early Friday aftn which helped sustain the temps near or below 32 during the day. The euro had the firehose signal as Will said at H7...one red flag no matter what the QPF output. I've preached that with the weenie 850-500 maps as a signal to saturation and heavy precip, but in the end..I should have recognized that potential as a red flag even for BOS. The atmosphere just aloft was cold enough to support snow if it came down hard, but it was still a bit stale of an airmass. We really needed some help along the coast and man did it happen. Same processes that destroyed us in 4/1/97. I actually didn't worry too much about the DC bust since the airmass here was colder and colder air advecting in from the northeast at 925mb! I know Will and I said that a few times. Of all things to help it snow, that is the best thing to do. Advect cold air in from the northeast. I guess if I had to think back and learn...I really should have practiced what I preached about the H7 RH signal and used that as a red flag to higher precip totals. My concern was that we were going to be on the edge of the real meaty stuff and it was tough for me to think real big totals if that happened. It basically was a huge bust potential. However, I think by Wednesday night and Thursday it was time to ramp up as everything started to hint at much better chances for a big snow in BOS..yet many outlets were still meh. I don't understand why many are using the excuse of wind direction and temps being colder than modeled. Guess what..temps will always be a tick or two colder then what models have. Now, you add the fact that nobody believed or looked at the euro QPF...no wonder why it was cold! Dynamics are a beautiful thing. Throw in the fact that nighttime in March might as well be mid January and boom. Did these OCMs just rip and read model temps and winds? I always tweak those in my mind almost every day when I forecast, especially for a snow event on the coast. I know at work, we told our clients that this thing is coming and they were prepared.

  I agreed with Bob, Will. Rollo, et all regarding the interior se ma areas and ORH. They looked great, especially interior SE MA regarding dynamics. I think people still think it can't snow in SE MA, but time and time again...it's not as bad as some make it out to be.

[OceanStWx]

I think we (as a field) may have created a mess of this one in an effort to be too deterministic. There has been this trend, especially this season it seems, of getting start/end times, amounts, snow maps, etc. out to the world many days in advance of a system. Some situations just don't lend themselves to long lead times. There were significant questions with this storm right into the 24 hour window. It was probably a situation where carrying a watch for the potential for 6+ inches longer than desired (remember a warning/advisory is still acceptable within 12 hours) with probabilistic forecasts of snow amounts (a la BOX's probability maps like the SREF does).

 

I think that need to get information to the public as fast as possible may have led to watches becoming advisories or OCMs making snowfall forecast maps too soon, when answers were still unknown and we were still asking questions. It is possible to still ramp up a forecast while riding a watch/uncertainty phase for a little longer than those locked in events.

 

I think we don't like to admit it sometimes, but we can be stumped on occasion. It does nobody any good to call blockbuster or whiff 4 days out when we still aren't sure of the most likely outcome. Once that first call is made to such an extreme, it is much harder to maneuver in the other direction.

[Ginx snewx]

I think we (as a field) may have created a mess of this one in an effort to be too deterministic. There has been this trend, especially this season it seems, of getting start/end times, amounts, snow maps, etc. out to the world many days in advance of a system. Some situations just don't lend themselves to long lead times. There were significant questions with this storm right into the 24 hour window. It was probably a situation where carrying a watch for the potential for 6+ inches longer than desired (remember a warning/advisory is still acceptable within 12 hours) with probabilistic forecasts of snow amounts (a la BOX's probability maps like the SREF does).

 

I think that need to get information to the public as fast as possible may have led to watches becoming advisories or OCMs making snowfall forecast maps too soon, when answers were still unknown and we were still asking questions. It is possible to still ramp up a forecast while riding a watch/uncertainty phase for a little longer than those locked in events.

 

I think we don't like to admit it sometimes, but we can be stumped on occasion. It does nobody any good to call blockbuster or whiff 4 days out when we still aren't sure of the most likely outcome. Once that first call is made to such an extreme, it is much harder to maneuver in the other direction.

I thought Box was doing a great job of ramping up until 12 hours before go time when they downgraded and did not have warnings up other than ORH and Prov county.Then they issued the warning after a foot had fallen.

[free_man]

One nuance I totally missed, was the far interior having above  (well above) 10:1 ratios.    Dendrite and others mentioned this,  iirc.   12:1 to 15:1 during overnight, despite a fairly ragged radar appearance.

 

 If someone told me a day or so advance, that the ratios would be that high  given the heaviest area of modeled qpf axis, orientation, "bullseye" etc  were well SE,  I would not have agreed.  

 

11.1" here from .75 - .8"?  or so.   That's a 5-8" call (conservatively).

[Damage In Tolland]

I thought Box was doing a great job of ramping up until 12 hours before go time when they downgraded and did not have warnings up other than ORH and Prov county.Then they issued the warning after a foot had fallen.

They seem to think there with the exception of a few..that Tolland County is just like Hartford County..I can think of numerous times when Windham Cty is under a watch or pegged to get much more snow..and it rarely happens that way. I couldn't believe they didn't have watches at least back to the river.. I think we had 6-8 inches OTG when they decided they'd better do something.

 

When I saw that last Euro run come NW and the Euro ens had 1 inch back to BDL Thursday afternoon..I knew we were good.

 

The 2 biggest busts were Rhode Island and Ct..one positive..one horribly negative

[weathafella]

Yeah the RI bust is a stumper. It's kind of like a patient coming in for a physical, hale and hearty, all systems working fine,no genetic or other clinical risks and a year later they are a 51 year old with greatly diminished life capacity after multiple strokes. Not everything can be explained before the fact and sometimes afterwards as well.

[Ginx snewx]

They seem to think there with the exception of a few..that Tolland County is just like Hartford County..I can think of numerous times when Windham Cty is under a watch or pegged to get much more snow..and it rarely happens that way. I couldn't believe they didn't have watches at least back to the river.. I think we had 6-8 inches OTG when they decided they'd better do something.

When I saw that last Euro run come NW and the Euro ens had 1 inch back to BDL Thursday afternoon..I knew we were good.

The 2 biggest busts were Rhode Island and Ct..one positive..one horribly negative

Actually the northern part of RI had 8-12

[nzucker]

I don't think temps were ever going to be a problem for SNE, so you can't use that as an excuse. It was early March and 850s were -7C or so, plenty cold for snow. I think some forecasters might have been conservative because March has failed to produce for so long it has almost stopped seeming like a winter month in recent years...in 2010 and 2012 we had temperatures in the 70s and 80s for a good part of the month. March 2012 finished +8.8F here.

 

But March is a great snow month from NYC north, especially for those of us with a bit of elevation. And early March is MUCH different from late March...the equinox is the time when the coastal plain's number of significant snowfalls declines precipitously. There have been SO many events around St. Patty's Day/Ides of March/Spring Equinox...but how many big snowstorms has NYC or BOS seen on 3/25? Or 3/28? After that it becomes more of a fluke rather than part of climo. The first week of the month, however, has seen many greats like 3/3/60, 3/6/1962, 3/3/1978, 3/6/2001, 3/2/2009, etc...

[Amped]

Conservative maybe a little better for the public. I think they are more forgiving when 4-8" is forecasted and they wake up with 15, than when 8-12" is forecasted and they're looking at wet grass and pavement.

 

And if we measured the storm in real inches instead of weenie snowboard inches, that 15" maybe more like 10-11". Actually a pretty decent call, and you can always tweak it by adding "Locally higher amounts"

[CoastalWx]

Temps were a factor and related to QPF Nate for this area, especially during daytime. Like I said, nighttime in March might as well be mid January. But the argument about it being colder than modeled, well that isn't necessarily true and just means you ripped off model temps verbatim and did not look at the euro right up until the snow was pounding your fanny. Even the GFS showed it was plenty cold so long as it would rip.

[CoastalWx]

I will say this, in a way it was nice to see the outlets not play up this storm. That meant no hype, not the top story for 48 hrs straight, no panicking from people, no grocery stores sold out of bread and milk as far as the eye can see and look what happened. We were just fine. As far as I'm concerned, let the outlets play it down.

[weathafella]

Conservative maybe a little better for the public. I think they are more forgiving when 4-8" is forecasted and they wake up with 15, than when 8-12" is forecasted and they're looking at wet grass and pavement.

And if we measured the storm in real inches instead of weenie snowboard inches, that 15" maybe more like 10-11". Actually a pretty decent call, and you can always tweak it by adding "Locally higher amounts"

But when 1-3 turns into 15 (which happened) and 12+ had school buses running t the storms height it's a big problem. Fortunately, NWS had warnings up so they fulfilled their public safety mission.

Most of the major tv markets expected minimal inside of 495. Disgraceful.

[CoastalWx]

But when 1-3 turns into 15 (which happened) and 12+ had school buses running t the storms height it's a big problem. Fortunately, NWS had warnings up so they fulfilled their public safety mission.

Most of the major tv markets expected minimal inside of 495. Disgraceful.

The writing on the wall was there too. Some had 3-6 on the Thursday 10pm news when I already had that lol.

It was a good learning experience. Certainly one to file in the back of my mind for this area.

[weathafella]

I don't think temps were ever going to be a problem for SNE, so you can't use that as an excuse. It was early March and 850s were -7C or so, plenty cold for snow. I think some forecasters might have been conservative because March has failed to produce for so long it has almost stopped seeming like a winter month in recent years...in 2010 and 2012 we had temperatures in the 70s and 80s for a good part of the month. March 2012 finished +8.8F here.

 

But March is a great snow month from NYC north, especially for those of us with a bit of elevation. And early March is MUCH different from late March...the equinox is the time when the coastal plain's number of significant snowfalls declines precipitously. There have been SO many events around St. Patty's Day/Ides of March/Spring Equinox...but how many big snowstorms has NYC or BOS seen on 3/25? Or 3/28? After that it becomes more of a fluke rather than part of climo. The first week of the month, however, has seen many greats like 3/3/60, 3/6/1962, 3/3/1978, 3/6/2001, 3/2/2009, etc...

Nate, since I returned to Boston 22 years ago, I can think of about 4-5 events >3 inches after the equinox. It's not as rare as you think here.

Another thing people get hung up on is any easterly wind component. Mar his one time when it will not nearly be the factor it is in December or January. Nevertheless, most of is do best when winds are 360-020.

[CoastalWx]

Nate, since I returned to Boston 22 years ago, I can think of about 4-5 events >3 inches after the equinox. It's not as rare as you think here.

Another thing people get hung up on is any easterly wind component. Mar his one time when it will not nearly be the factor it is in December or January. Nevertheless, most of is do best when winds are 360-020.

An east wind in December is almost a death wish unless it's very cold aloft like 12/16/07.

[nzucker]

Temps were a factor and related to QPF Nate for this area, especially during daytime. Like I said, nighttime in March might as well be mid January. But the argument about it being colder than modeled, well that isn't necessarily true and just means you ripped off model temps verbatim and did not look at the euro right up until the snow was pounding your fanny. Even the GFS showed it was plenty cold so long as it would rip.

I'm not so sure temperatures were ever the question; 850s were -7C to -8C and many had reached the teens and low 20s just a night or two before the storm, even down here in NYC metro. As you say, heavier QPF was important during the day to prevent a light, non-accumulating snow with the sun melting whatever tried to fall. That's not so much a temperature problem though, as the upper levels were wicked cold and I was down to 32.5F when snow was just starting in Dobbs Ferry with nothing on the ground. It's more a question of the snowfall rates overwhelming the natural melting during the day in March, or during the day in any part of winter when snow is falling extremely lightly. March exacerbates that, but I think marginal temps/sun angle were on the back burner in this storm compared to the main question of how much liquid we'd get with a storm 700 miles offshore and the firehose being aimed pretty far southeast at H7. Basically, I think you and I agree but are just saying it in different ways...

 

Nate, since I returned to Boston 22 years ago, I can think of about 4-5 events >3 inches after the equinox. It's not as rare as you think here.

Another thing people get hung up on is any easterly wind component. Mar his one time when it will not nearly be the factor it is in December or January. Nevertheless, most of is do best when winds are 360-020.

I'm not saying it's impossible for it to snow after the equinox. And it's certainly more likely in BOS metro than NYC metro. However, I think if you plotted all significant March snowfalls (say 3"+) on a graph for both locations, you'd see a fairly substantial drop-off right around the Equinox. It's no surprise that that's when the two locations are starting to see average highs in the 50s, meaning that it requires a -15F departure to snow, compared to a -8F departure in the beginning of March. I couldn't keep track of how many -8F departures we see most years, but I could probably count the -15F departures on one hand. Of course, if the event is perfectly timed at night, it is a little easier. I am going to try to make that graph of significant March snowfalls and get back to you about distribution.

That being said, I can remember some great snowfalls after the Equinox here in Westchester. The last one was 3/24/2011 which dropped 2.5" in an overrunning pattern. 4/7/2003 also had 3". Before that, we had 4/1/97 with 6" and 4/10/96 with 2.6". In the 1980s, three outstanding late season events were 6.2" on 3/30/1984, 1.9" on 4/19/1983, and 9.7" on 4/6/1982. I do feel it's been a while since a significant April event, and we may have a chance this year with the -NAO.

[CoastalWx]

Nate you are not getting it. Temps at the surface are helped driven by precip rates. Nobody argues it was cold aloft, but 1-2sm -SN during the day probably means temps in the low to mid 30s. 

[ORH_wxman]

Yeah BOS wa salwyas a lot tougher....-3C at 925mb meant it was going to snow, but if it stayed light, it would have been like 33F light snow that didn't stick well. That said, I think the 925mb temps actually verified colder than even the Euro said. Euro had them around that -3C to -4C mark...we actually got them down to like -6C, lol.

 

There wasn't much of an excuse for keeping totals really low though for interior SE MA and the ORH hills. Even some TV outlets had ORH at like 3-5" as recently as 5-6pm news Thursday afternoon when we already had close to 4".

[nzucker]

Nate you are not getting it. Temps at the surface are helped driven by precip rates. Nobody argues it was cold aloft, but 1-2sm -SN during the day probably means temps in the low to mid 30s. 

Right...of course more QPF means colder temps. But 2SM -SN isn't not accumulating because of warm temps, it's not accumulating because it's light and the sun's out for the most part. Snows that light rarely accumulate anyway. 

[CoastalWx]

Right...of course more QPF means colder temps. But 2SM -SN isn't not accumulating because of warm temps, it's not accumulating because it's light and the sun's out for the most part. Snows that light rarely accumulate anyway. 

Yes and because of that, it's 33-34F non accumulating snow.

[nzucker]

Yes and because of that, it's 33-34F non accumulating snow.

It could be...but it could be 30F and a 2SM -SN isn't going to accumulate much. Warmer temperatures in March exacerbates the problem but aren't the whole story. 

[CoastalWx]

It could be...but it could be 30F and a 2SM -SN isn't going to accumulate much. Warmer temperatures in March exacerbates the problem but aren't the whole story. 

 

That's true, but we did not have the airmass to keep snows that light and temps near 30F. This is why we needed it to rip.

[nzucker]

That's true, but we did not have the airmass to keep snows that light and temps near 30F. This is why we needed it to rip.

And by God we did!

[bch2014]

Was there a march event on like 3/24/05, I was 8 and I vaguely remember skiing in it.

[CoastalWx]

Was there a march event on like 3/24/05, I was 8 and I vaguely remember skiing in it.

 

http://www.erh.noaa.gov/box/climate/snowrpts/MA2505

[OceanStWx]

It could be...but it could be 30F and a 2SM -SN isn't going to accumulate much. Warmer temperatures in March exacerbates the problem but aren't the whole story. 

 

2SM -SN will accumulate as long as temperatures support the snow sticking around. It snowed for around 36 hours in some places. Say it accumulated a 0.1" snow every hour in that light stuff, you end up with 3.6" of weenie snow accumulation. Not much, but it makes a difference.

 

However, with a marginal air mass in place that snow melted on contact rather than slowly add up over time. Temperatures have to be an issue near the surface then if there was melting going on. Clearly it wasn't the only issue with this storm, but you can't just toss the temperature argument altogether.