So enjoy what does fall. Here's the NAM forecast total precipitation for those interested:
Looks like we're in the 0.10 inch liquid range which translates to a 1-1.5 inch snowfall.
Tuesday, January 11, 2011
18Z NAM Confirms Eastward Shift, Less Snow
The eastward shift continues bringing our area into the one-inch range for snow fall. Pretty disappointing but this was never going to be a large snowstorm.
So enjoy what does fall. Here's the NAM forecast total precipitation for those interested:
Looks like we're in the 0.10 inch liquid range which translates to a 1-1.5 inch snowfall.
So enjoy what does fall. Here's the NAM forecast total precipitation for those interested:
Looks like we're in the 0.10 inch liquid range which translates to a 1-1.5 inch snowfall.
Point to the Dryslot
Check out this radar. Can you see the dryslot?
Inland low is pretty strong but we're getting caught in the dryslot between it and the storm off the North Carolina coast.
Let's see if we can get more than an inch out of this thing.
The snow drought this season continues. Guess we are paying for last year!
Inland low is pretty strong but we're getting caught in the dryslot between it and the storm off the North Carolina coast.
Let's see if we can get more than an inch out of this thing.
The snow drought this season continues. Guess we are paying for last year!
Stupid Dryslot
It's going to be close around here. The shortest-term models we have are slightly east of the NAM positions yesterday which isn't that good for us. I think the gradient of snow shifts east a bit and we're talking lower totals of 1-3 inches here in the DC Metro with less to the west and more up around Baltimore. I felt pretty confident about 2 inches yesterday and I still think so.
Still reviewing data this morning. There's even less consensus as you get to small-scale modeling and a lot of it will have to be figured out once the snow starts.
You can see the eastward shift in the NAM in this precipitation total forecast map (compare to the one in the post just below this).
Still reviewing data this morning. There's even less consensus as you get to small-scale modeling and a lot of it will have to be figured out once the snow starts.
You can see the eastward shift in the NAM in this precipitation total forecast map (compare to the one in the post just below this).
Monday, January 10, 2011
18Z NAM Precipitation
Here's the NAM precipitation forecast. We'll be running right around 10-12:1 ratios so multiply by the liquid precipitation in the map above by 10 or so and you get forecast snowfall in inches.
As you can see, this fits nicely with the forecast so I'll leave things how they are for now.
More model guidance coming - especially later tonight. Those will be important runs.
As you can see, this fits nicely with the forecast so I'll leave things how they are for now.
More model guidance coming - especially later tonight. Those will be important runs.
18Z NAM
Here's the NAM from the 18Z run (earlier maps were from the 12Z run). Usually the 18Z models are taken with a grain of salt and usually just used to identify trends in the forecast or inconsistencies.
The NAM is pretty consistent with its slower, stronger philosophy although this low position is about the same but the precipitation shield is further west than that of the 12Z run at the same forecast time.
Here we see the phase occurring and some light to moderate snowfall over our area at about 1 AM on Wednesday.
No change to forecast at this point. No significant model trends to suggest that the forecast is off track.
The NAM is pretty consistent with its slower, stronger philosophy although this low position is about the same but the precipitation shield is further west than that of the 12Z run at the same forecast time.
Here we see the phase occurring and some light to moderate snowfall over our area at about 1 AM on Wednesday.
No change to forecast at this point. No significant model trends to suggest that the forecast is off track.
At This Point, a Light Event
Elements are coming together for our snow event tomorrow.
This will be a storm track and low pressure interaction pattern known as Miller B, the hallmark of which is two low pressure centers interacting - one from the west over land and one from the southeast.
Our first two big storms (December 19-20 and February 5-6) last season were both Miller B storms - so we know they can produce large amounts of snow. However, it is the timing and location of the interaction between the two lows and their original intensity (especially of the inland storm) that are crucial components.
Last year, in both cases, the low from the west passed close to our area before transferring its energy to the developing offshore low well to our southeast. As the storm moved up the coast this infusion of energy and time over the Gulf Stream caused the "bombing out effect" that buried our area in snowfall.
This time will be a bit different.
So let's look at some near-term model maps to see where this transfer might take place. This close to the storm, I prefer to look at the NAM which is usually a very good short term model . However, I should note that for this storm - at least so far - the NAM is an outlier. It is slower with timing overall and stronger with the inland low. Making it even more suspect is the fact that, generally, there IS consensus between the Euro, GFS and UKMET models and all of them are a little faster and a little weaker with the storm for this area.
However, the NAM essentially agrees with the other models for the time period leading up to early morning Wednesday. So, I'll show you the NAM because it represents the strong side of the guidance available - but my actual forecast rationale will be like a buffet: an entree of Euro/GFS/UKMET and a NAM dessert.
That darker-green area in Illinois is our inland storm moving off to the northeast. Our coastal storm is developing off the Carolinas. This would be about 15 UTC or around 10 AM tomorrow. As you can see, not much going on around here according to the NAM. I would agree. Maybe some flurries or light snow showers but I think most of the action will be later in the afternoon. NAM is a little slower than the other models but not significantly.
Six hours later and the NAM is forecasting the edge of the steadier light snow in our area. This is late afternoon tomorrow - around 4 PM. The low to the west is a bit better organized as is the one off the coast. These will continue to function semi-independently for another few hours before the energy transfer and phase begin to take place. I expect the light snow to begin falling steadily sometime tomorrow mid to late afternoon. This rationale is supported by the other models.
Here's where the models disagree - or really where most of the other models start looking at the NAM with a suspicious glare.
In the NAM map above (9 hours later - around 1 AM Wednesday morning) you can see the two low pressure areas characteristic of a classic Miller B presentation - one in western Pennsylvania and one just off the Delmarva coast. This is where the NAM forecasts the transfer and phase to really get going. The NAM is slower to make this transfer and, because it's also forecasting a stronger inland low, our region experiences a more extended period of moderate snow before the transfer is complete and the storm off the coast takes over.
Compare the NAM to the GFS forecast for the same hour:
The GFS/Euro/UKMET gang advocate a weaker inland low (it's not even rendered with an "L" in western PA unlike the NAM), a quicker energy transfer and a faster end to the snow in our area.
One other thing to worry about: when the 850 MB low passes to the north of our area in a Miller B scenario, there is a high likelihood of being dryslotted around here. We essentially end up in a "hole" between the storms as the phase takes place. Check out the precipitation totals by the time the storm is off the coast of Maine Wednesday night:
See the hole in the precipitation field over our region and to the west and south? Dryslotting.
It's worth noting that even the NAM - on the strong side of the guidance - forecasts the coastal storm takeover to be too far north (off the Delmarva coast) to affect us much.
Here's what I expect:
Based on the previously-discussed issues and scenarios and a few glances at upper level data which suggests that there will be a fair amount of precipitation ahead of that coastal low even before the Miller B transfer begins, I would say a chance of flurries or light snow beginning late morning Tuesday but especially by the afternoon. By the late afternoon/early evening I think the light snow becomes steadier and and more widespread and continues into the night. There's a good chance that the snow becomes moderate for a time Tuesday night (especially the further east you are) before it ends from west to east overnight on Tuesday into early Wednesday morning.
This is another storm where the lightest amounts will fall to the west (except once you get over the mountains) and the heaviest amounts to the east (except possibly along the coast and in southern Maryland where mixing of snow with sleet and rain could occur).
Generally in the DC Metro I'd expect 2-4 inches in a southwest (low end) to northeast (high end) gradient up to about Baltimore. North and east of Baltimore things could get interesting quickly with 5+ inch amounts. Similarly, towards the coast amounts could reach 5+ inches.
By Wednesday morning this is all out of here and in the wake we get more of the same so far this winter: cold and windy into the weekend.
What could change things?
- The NAM could be the start of a trend rather than an outlier. Slower and stronger means more snow for us.
- The phase could happen earlier and further to the south. The more time we spend under the influence of the coastal low, the better. If the phase has already happened at a time when precipitation is still falling from the coastal low, we'll get more snow.
- The UKMET model could be right. It represents the weakest edge of the guidance. The UKMET solution is for a storm further out to sea. We get even less of an impact from the coastal low than is already forecast by most of the other models and, as a result, less overall snow.
Why is forecasting winter weather so complicated and so often inaccurate?
I could write 6 posts about this and not be done but there are a number of reasons for it:
- The guidance is, for the most part, forecasting massive-scale events and we try to translate them to the small-scale. Take the storm after Christmas, for example. The distance between Ocean City and Washington, DC is about 100 miles (as the crow flies). You'll remember that we received a dusting of snow from that storm here in the DC area. Ocean City? 14 inches. We were around half that distance away from 6+ inches. 50 lousy miles in a storm hundreds of miles in diameter.
- The psychology of snow around here. Let's take last season's largest snowfall for example. Midweek forecasts that week were for between 24-32 inches of snow. Some had higher amounts up to three feet. By the time the snow started falling, forecasts were trimmed back by about 4 inches to allow for the reduced snowfall during the Miller B transfer. Do you think anybody cared that their snow forecast was now 20-27 inches instead of 24-32?
But now let's say I'm 4 inches off in my forecast for tomorrow. That's the difference between snow and no snow at all.
- Storm track is almost the whole ballgame. In our case for tomorrow it's actually TWO storms we have to track. What if the one from the west actually gets closer to us? What if the track of the storm from the south is too far east? Everything changes.
- Storm intensity is the rest of the ballgame. Our storms for tomorrow are just getting organized. How strong will they be? You can see the models already disagree about that with the NAM having a stronger inland storm than the rest of the models.
- Small-scale features. Things like "mesoscale banding" can cause big differences in snow totals. Check out this radar image from the December 26th storm in New Jersey/New York. You can see how a forecast of 12-14 inches ends up being 20+. Those yellow bands of snow are almost impossible to forecast in duration and extent. But if you get under one, it's like the snowpocalypse all over again.
Hey - I'm not making excuses - sometimes winter weather forecasts around here can be wretchedly bad. Sometimes meteorologists ignore their better judgement and their intuition in favor of computers. But it's a crazy science and even in this age of incredible technology advancement, there's still a lot of guessing to be done.
I'll be watching the afternoon model runs. If anything changes, I'll make a quick post to let you know.
This will be a storm track and low pressure interaction pattern known as Miller B, the hallmark of which is two low pressure centers interacting - one from the west over land and one from the southeast.
Our first two big storms (December 19-20 and February 5-6) last season were both Miller B storms - so we know they can produce large amounts of snow. However, it is the timing and location of the interaction between the two lows and their original intensity (especially of the inland storm) that are crucial components.
Last year, in both cases, the low from the west passed close to our area before transferring its energy to the developing offshore low well to our southeast. As the storm moved up the coast this infusion of energy and time over the Gulf Stream caused the "bombing out effect" that buried our area in snowfall.
This time will be a bit different.
So let's look at some near-term model maps to see where this transfer might take place. This close to the storm, I prefer to look at the NAM which is usually a very good short term model . However, I should note that for this storm - at least so far - the NAM is an outlier. It is slower with timing overall and stronger with the inland low. Making it even more suspect is the fact that, generally, there IS consensus between the Euro, GFS and UKMET models and all of them are a little faster and a little weaker with the storm for this area.
However, the NAM essentially agrees with the other models for the time period leading up to early morning Wednesday. So, I'll show you the NAM because it represents the strong side of the guidance available - but my actual forecast rationale will be like a buffet: an entree of Euro/GFS/UKMET and a NAM dessert.
That darker-green area in Illinois is our inland storm moving off to the northeast. Our coastal storm is developing off the Carolinas. This would be about 15 UTC or around 10 AM tomorrow. As you can see, not much going on around here according to the NAM. I would agree. Maybe some flurries or light snow showers but I think most of the action will be later in the afternoon. NAM is a little slower than the other models but not significantly.
Six hours later and the NAM is forecasting the edge of the steadier light snow in our area. This is late afternoon tomorrow - around 4 PM. The low to the west is a bit better organized as is the one off the coast. These will continue to function semi-independently for another few hours before the energy transfer and phase begin to take place. I expect the light snow to begin falling steadily sometime tomorrow mid to late afternoon. This rationale is supported by the other models.
Here's where the models disagree - or really where most of the other models start looking at the NAM with a suspicious glare.
In the NAM map above (9 hours later - around 1 AM Wednesday morning) you can see the two low pressure areas characteristic of a classic Miller B presentation - one in western Pennsylvania and one just off the Delmarva coast. This is where the NAM forecasts the transfer and phase to really get going. The NAM is slower to make this transfer and, because it's also forecasting a stronger inland low, our region experiences a more extended period of moderate snow before the transfer is complete and the storm off the coast takes over.
Compare the NAM to the GFS forecast for the same hour:
The GFS/Euro/UKMET gang advocate a weaker inland low (it's not even rendered with an "L" in western PA unlike the NAM), a quicker energy transfer and a faster end to the snow in our area.
One other thing to worry about: when the 850 MB low passes to the north of our area in a Miller B scenario, there is a high likelihood of being dryslotted around here. We essentially end up in a "hole" between the storms as the phase takes place. Check out the precipitation totals by the time the storm is off the coast of Maine Wednesday night:
See the hole in the precipitation field over our region and to the west and south? Dryslotting.
It's worth noting that even the NAM - on the strong side of the guidance - forecasts the coastal storm takeover to be too far north (off the Delmarva coast) to affect us much.
Here's what I expect:
Based on the previously-discussed issues and scenarios and a few glances at upper level data which suggests that there will be a fair amount of precipitation ahead of that coastal low even before the Miller B transfer begins, I would say a chance of flurries or light snow beginning late morning Tuesday but especially by the afternoon. By the late afternoon/early evening I think the light snow becomes steadier and and more widespread and continues into the night. There's a good chance that the snow becomes moderate for a time Tuesday night (especially the further east you are) before it ends from west to east overnight on Tuesday into early Wednesday morning.
This is another storm where the lightest amounts will fall to the west (except once you get over the mountains) and the heaviest amounts to the east (except possibly along the coast and in southern Maryland where mixing of snow with sleet and rain could occur).
Generally in the DC Metro I'd expect 2-4 inches in a southwest (low end) to northeast (high end) gradient up to about Baltimore. North and east of Baltimore things could get interesting quickly with 5+ inch amounts. Similarly, towards the coast amounts could reach 5+ inches.
By Wednesday morning this is all out of here and in the wake we get more of the same so far this winter: cold and windy into the weekend.
What could change things?
- The NAM could be the start of a trend rather than an outlier. Slower and stronger means more snow for us.
- The phase could happen earlier and further to the south. The more time we spend under the influence of the coastal low, the better. If the phase has already happened at a time when precipitation is still falling from the coastal low, we'll get more snow.
- The UKMET model could be right. It represents the weakest edge of the guidance. The UKMET solution is for a storm further out to sea. We get even less of an impact from the coastal low than is already forecast by most of the other models and, as a result, less overall snow.
Why is forecasting winter weather so complicated and so often inaccurate?
I could write 6 posts about this and not be done but there are a number of reasons for it:
- The guidance is, for the most part, forecasting massive-scale events and we try to translate them to the small-scale. Take the storm after Christmas, for example. The distance between Ocean City and Washington, DC is about 100 miles (as the crow flies). You'll remember that we received a dusting of snow from that storm here in the DC area. Ocean City? 14 inches. We were around half that distance away from 6+ inches. 50 lousy miles in a storm hundreds of miles in diameter.
- The psychology of snow around here. Let's take last season's largest snowfall for example. Midweek forecasts that week were for between 24-32 inches of snow. Some had higher amounts up to three feet. By the time the snow started falling, forecasts were trimmed back by about 4 inches to allow for the reduced snowfall during the Miller B transfer. Do you think anybody cared that their snow forecast was now 20-27 inches instead of 24-32?
But now let's say I'm 4 inches off in my forecast for tomorrow. That's the difference between snow and no snow at all.
- Storm track is almost the whole ballgame. In our case for tomorrow it's actually TWO storms we have to track. What if the one from the west actually gets closer to us? What if the track of the storm from the south is too far east? Everything changes.
- Storm intensity is the rest of the ballgame. Our storms for tomorrow are just getting organized. How strong will they be? You can see the models already disagree about that with the NAM having a stronger inland storm than the rest of the models.
- Small-scale features. Things like "mesoscale banding" can cause big differences in snow totals. Check out this radar image from the December 26th storm in New Jersey/New York. You can see how a forecast of 12-14 inches ends up being 20+. Those yellow bands of snow are almost impossible to forecast in duration and extent. But if you get under one, it's like the snowpocalypse all over again.
Hey - I'm not making excuses - sometimes winter weather forecasts around here can be wretchedly bad. Sometimes meteorologists ignore their better judgement and their intuition in favor of computers. But it's a crazy science and even in this age of incredible technology advancement, there's still a lot of guessing to be done.
I'll be watching the afternoon model runs. If anything changes, I'll make a quick post to let you know.
Saturday, January 8, 2011
Complex Pattern for Monday Night Through Tuesday
I didn't post for the last couple of snowfalls because they were expected to be minor and turned out to be. Just some snow showers on Thursday night into Friday morning and a few more on Saturday morning provided the dusting that the DC metro area received.
A much more complex situation is in the offing for a time period beginning Monday night through Tuesday night - possibly lingering a bit longer than that. Much is still undetermined but there's a likelihood of accumulating snow here in the DC metro. How much and where are both things that need a few more model runs to iron out.
As has become customary this winter, the models are all over the place and there's not much consensus beyond a good chance of snow on Tuesday. This begins with a low pressure area forming in southeastern Texas and moving along the Gulf coast on Sunday night. I expect this low to cross off the coast of the southeastern U.S. by Monday - probably around central Georgia.
From the north comes the polar jet and a disturbance rides along with it from the plains into the Ohio Valley. This is expected to be pretty strong and it, alone, would likely produce some snowfall around here as it tracks across our region Monday night.
Here's where the models start to diverge.
Scenario 1, at this time the more likely, has the northern energy missing phase with the southern energy until the storm is well northeast of us. This scenario means a weak coastal low, a more dominant inland low and light accumulation amounts depending on the strength of the inland disturbance and the proximity of the coastal low - weak as it might be - as it moves northeastward into the Atlantic.
Additionally, the models hint a secondary low forming in southwest Virginia and moving northeastward which would continue our snow chances through Tuesday afternoon and evening.
Even with the extended period of snow, it seems like a light event.
Scenario 2 would have the northern energy phasing with the southern energy which would strengthen the coastal storm and also having the effect of "holding" the storm closer to the coast as it moves northward. This scenario is a classic nor'easter scenario - but, as in previous situations, the timing of the phase is everything here.
Right now the model consensus is for a later phase but when and where is very uncertain.
An early call would be a chance of snow Monday night with better chances Tuesday and Tuesday afternoon possibly extending through Tuesday night. I can say that there's a good chance this will be "accumulating" snow which would mean, for me, "2 inches or more" but lots of details still have to be ironed out before we can have serious discussion about totals.
There's also the chance that storm track gives us mixing issues as temperatures will be borderline (low to mid 30's) during the balance of the day Tuesday.
I'll be watching the model runs pretty carefully for the next few days. I'll start posting some maps in my next post.
Hopefully this one decides what it wants to do sometime before the day of the event!
A much more complex situation is in the offing for a time period beginning Monday night through Tuesday night - possibly lingering a bit longer than that. Much is still undetermined but there's a likelihood of accumulating snow here in the DC metro. How much and where are both things that need a few more model runs to iron out.
As has become customary this winter, the models are all over the place and there's not much consensus beyond a good chance of snow on Tuesday. This begins with a low pressure area forming in southeastern Texas and moving along the Gulf coast on Sunday night. I expect this low to cross off the coast of the southeastern U.S. by Monday - probably around central Georgia.
From the north comes the polar jet and a disturbance rides along with it from the plains into the Ohio Valley. This is expected to be pretty strong and it, alone, would likely produce some snowfall around here as it tracks across our region Monday night.
Here's where the models start to diverge.
Scenario 1, at this time the more likely, has the northern energy missing phase with the southern energy until the storm is well northeast of us. This scenario means a weak coastal low, a more dominant inland low and light accumulation amounts depending on the strength of the inland disturbance and the proximity of the coastal low - weak as it might be - as it moves northeastward into the Atlantic.
Additionally, the models hint a secondary low forming in southwest Virginia and moving northeastward which would continue our snow chances through Tuesday afternoon and evening.
Even with the extended period of snow, it seems like a light event.
Scenario 2 would have the northern energy phasing with the southern energy which would strengthen the coastal storm and also having the effect of "holding" the storm closer to the coast as it moves northward. This scenario is a classic nor'easter scenario - but, as in previous situations, the timing of the phase is everything here.
Right now the model consensus is for a later phase but when and where is very uncertain.
An early call would be a chance of snow Monday night with better chances Tuesday and Tuesday afternoon possibly extending through Tuesday night. I can say that there's a good chance this will be "accumulating" snow which would mean, for me, "2 inches or more" but lots of details still have to be ironed out before we can have serious discussion about totals.
There's also the chance that storm track gives us mixing issues as temperatures will be borderline (low to mid 30's) during the balance of the day Tuesday.
I'll be watching the model runs pretty carefully for the next few days. I'll start posting some maps in my next post.
Hopefully this one decides what it wants to do sometime before the day of the event!
Subscribe to:
Posts (Atom)