Significant Snow Likely - But How Significant?

Things are still not as clear as I'd like them to be.

The Euro is still the westernmost track and the Canadian model (CMC) has come into line with that   Their tracks are quite close to the coast which create precipitation type issues in the forecast even west of DC and cuts down on the overall amounts.

The NAM, on the other hand, is a much colder solution and keeps us in the heavy precipitation but now it seems like the outlier.    The GFS is somewhere in between.

Here's the ECMWF deterministic output:

ECMWF Deterministic FORECAST FOR: DCA    LAT=  38.85 LON=  -77.03 ELE=    16

TUE 12Z 11-FEB  -7.6   -12.6    1030      60       9    0.00     546     523  
TUE 18Z 11-FEB  -2.0   -12.8    1030      43       9    0.00     547     524  
WED 00Z 12-FEB  -4.6   -11.8    1031      58      13    0.00     547     523  
WED 06Z 12-FEB  -6.3   -11.3    1033      51       4    0.00     549     523  
WED 12Z 12-FEB  -7.4    -9.5    1034      53       3    0.00     553     527  
WED 18Z 12-FEB  -1.8    -7.7    1032      39       3    0.00     555     530  
THU 00Z 13-FEB  -2.9    -6.7    1028      57      44    0.00     556     534  
THU 06Z 13-FEB  -3.2    -3.9    1021      88      98    0.18     554     537  
THU 12Z 13-FEB  -0.5     0.0    1010      93      99    0.55     552     544  
THU 18Z 13-FEB   0.9     2.8     998      91      76    0.21     546     547    
FRI 00Z 14-FEB   1.2    -1.2     992      89      53    0.15     533     540  
FRI 06Z 14-FEB   1.8    -2.7     996      87      90    0.25     532     536

 The same two highlighted times are still problematic - maybe even a bit more than they were last night.  The surface temperatures rise above freezing in the afternoon Thursday (0.9) and the thicknesses go well above the 540 benchmark (544, 547) meaning that mixing issues will almost certainly be present for areas east of I-95, DC included.

Looking at the GFS, I don't see as many issues.  This frame from the GFS shows heavy snow in our area Thursday afternoon:

The NAM shows a similar story but the precipitation is heavier.  However, this is a common bias with the NAM model so we can't take it too literally.

I'll be looking carefully at the upper level low as it moves into position over the next 12 hours.  If the storm "wraps up" too quickly it will make for a more potent storm but a more westward track.  The further west the track, the more mixing issues we have.  You really get a sense here of the precision necessary to produce big snows in the DC area.  The model data supports this.  There are SREF members forecasting everything from 2 inches to 20 inches for the DC area.  Clearly, that's not useful basis for a forecast.

I think we're in for some snow.  I think 4 or more inches is a good bet areawide with the first 3-5 accumulating overnight Wednesday into Thursday morning.  After that, it gets tricky.  Where will the storm track?  How close to the coast?  Does banding set up in our region?  Does the amount of lift in the atmosphere counter warming temperatures?

I think all of those questions are left to be answered still and will make the difference between a good storm, a big storm and a major storm.

UPDATE: Euro Puts Bullseye on DC

The early morning run of the ECMWF (Euro) model does nothing to dissuade snow forecasts and, in fact, paints a bullseye over DC.

Still some disagreement in the models - particularly the GFS and the Euro is still a bit too close to the coast for comfort (precip mixing issues).

I'll write more later this morning.

Let the Games Begin

First - a Winter Storm Watch has been put in place for our region for Wednesday night through Thursday.  By strict National Weather Service definition, a "Winter Storm" is one that has the potential to produce 5 or more inches of snow.

So what do we have here?

I'll start with what we don't have:  guidance model consensus.  The solutions range as follows:

GFS:  a smaller shot of snow - less impact

ECMWF (Euro): big hit, "warmer" solution - possible snow/sleet/rain mixing issues on Thursday

CMC, UKMET, NAM: some variation of "keep the dog inside lest it disappear beneath the snowfall"

I'd really like to have better model consensus about 60 hours out from the storm but that's why this is such a difficult area for which to make a forecast.

Right now, my instinct tells me that we're going to get snow - quite a bit of it - but that where you are might make a big difference as to how much you see.  I'm reasonably confident at this point that the western suburbs will see a lot but I worry about DC and points east.  If the Euro verifies as-is, it could be too warm to snow on Thursday.

If the GFS verifies as is, we're all out of the huge snow ballgame.

Here's a bit of the GFS run from this afternoon.  (Note:  the afternoon run of the GFS is considered less trustworthy because it is not initialized from upper air data - it uses physics to make a forecast based on the now-6-hours-old upper air data from the morning)

This would be Thursday afternoon.  The low is positioned off the coast of Norfolk and is throwing back precipitation into our region.  This is a "colder" solution which gives us no precipitation-type problems but gives us overall precipitation amount problems.

The storm is gone in the next frame working its way offshore and up the coast.  The GFS predicts the following as snowfall totals:

As you can see, this guidance suggests our region misses the jackpot zone and we find ourselves in a VERY tight precipitation gradient with almost nothing to the northwest and heavier precipitation to the southeast.

At this point, however, the GFS is an outlier.  It has, by far, the most eastward track for the surface low and the least impact to our region.

On the other hand (hide the dogs!), this is what the NAM model guidance suggests for the same storm period:

which is a monster hit for the DC area and really no precipitation gradient problems to worry about.

The ECMWF (Euro) forecast for our area is a little more problematic.  Here's the raw model output for National Airport:

ECMWF Deterministic FORECAST FOR: DCA    LAT=  38.85 LON=  -77.03 ELE=    16

MON 12Z 10-FEB  -4.0   -10.1    1023      65      64    0.00     543     526  
MON 18Z 10-FEB  -0.9   -11.3    1024      49       7    0.00     543     525  
TUE 00Z 11-FEB  -5.4   -11.5    1025      59       4    0.00     544     524  
TUE 06Z 11-FEB  -5.4   -12.5    1028      52       7    0.00     543     522  
TUE 12Z 11-FEB  -7.1   -13.4    1030      49       9    0.00     546     523  
TUE 18Z 11-FEB  -2.2   -12.6    1030      39       9    0.00     547     523  
WED 00Z 12-FEB  -5.0   -11.5    1031      51       9    0.00     548     524  
WED 06Z 12-FEB  -7.0   -10.8    1033      55       3    0.00     550     524  
WED 12Z 12-FEB  -7.9    -8.1    1034      57       3    0.00     553     527  
WED 18Z 12-FEB  -1.9    -6.6    1032      40       9    0.00     555     530  
THU 00Z 13-FEB  -2.9    -6.3    1027      62      85    0.00     555     534  
THU 06Z 13-FEB  -2.4    -2.3    1018      92      98    0.30     552     538  
THU 12Z 13-FEB   0.6    -0.6    1005      93      98    0.66     549     545  
THU 18Z 13-FEB   1.1    -3.8     995      92      42    0.45     537     541    
 FRI 00Z 14-FEB   1.6    -4.1     996      89      89    0.30     533     537  
FRI 06Z 14-FEB  -0.2    -2.3    1002      73      20    0.04     536     534  
FRI 12Z 14-FEB  -6.2    -1.2    1007      79      54    0.00     540     535

I have bolded the two problem times.  During the heaviest precipitation, two not-so-good things happen.  One, the 2M temperature goes positive (0.6 and 1.1).  That means surface temperatures are expected to be above freezing.  We've seen a couple of times recently where that made a big difference in snow acummulation.

The second thing is that the 1000-500 mb thickness rises above 540 (545, 541) which is the "benchmark" for snow vs sleet or rain.  The Euro is indicating that a little over an inch of liquid precipitation (normally around 10 inches of snow) is going to fall with marginal - at best - temperatures aloft and at the surface.

This type of thing can bust a forecast like a piñata.

The evening model runs will be very interesting to watch.  I'll let you know what things look like as we get closer.

For now, prepare for a significant winter storm from Wednesday night into Thursday night with snow and wind.  Amounts, timing and track to be determined as we get closer to the event.

Keeping an Eye on Wednesday/Thursday

This past weekend was a perfect example of what happens when two separate storms fail to phase.  The result is two rounds of snow showers and that's it.

This Wednesday-Thursday is starting to look much more interesting in the last few model runs.  There are still some crucial differences - particularly with storm track.  The GFS is currently the outlier showing a storm that phases too late and ends up out to sea, barely brushing the area.

The Euro and other models show a storm moving up the coast - maybe too close (which would draw warmer air from the Atlantic and give us precipitation type issues).  But trends in the models seem to be pointing towards a colder solution and more snow.

This could be the big one of the winter but it's still too early to tell.  Once I've had more time to look at the guidance, I'll post my thoughts but, for now, keep an eye on this time period.

Storm Potential for This Weekend

I'm keeping an eye on models and data suggesting the potential for a significant winter storm in our area Saturday into Sunday.  Details are a bit sketchy and the models don't agree so for now, the signal is there for a big storm and some models are trending colder - suggestive of all snow - while others keep us in mixed precipitation.

Storm track is everything and I suspect the models won't have a good grasp until after this ice storm gets through here tonight and tomorrow.

I'll post more once the situation clears a bit.

Significant Winter Storm Expected Tomorrow

All the ingredients are coming together for a significant snowfall in the DC area tomorrow.

For most of the day, I've been trying to figure out how this forecast could end up completely wrong.  We do live in the DC area, after all, and if there was ever a place where snowfall forecasts go to die, it's here.

But I can't think of one.

The cold air will be in place, the upper-level energy is where it should be right now and the front is heading this way which will knock a good 25-30 degrees off our high temperatures today by the time the snow starts.

I suppose the only thing that could go wrong is if the storm gets its act together later than forecast by the models which would mean a less intense storm as it passes by.  At the moment, this seems unlikely.

The GFS model shows the storm as it leaves the coast tomorrow and rapidly intensifies over the Atlantic.  The trough of low pressure is expected to switch from a positive to a negative tilt which almost always means a more intense and slower-moving storm.

As the storm starts to influence our region, it will encounter the sharply-colder air and that should improve the accumulation numbers relative to the amount of precipitation produced by the storm.  The rule-of-thumb ratio is 10 inches of snow per 1 inch of liquid precipitation.  I think we do better than that tomorrow.

This will produce fluffy snow that will blow around in the strong gusty winds and that will be easier to shovel.  It will, however, be complete crap for making snowballs or snowmen.

The snow could be quite heavy in the morning - especially if banding sets up (as it should).  Forecasting the precise location of the banding is next to impossible so that's the usual reason for the range of forecast numbers.  In this case, the National Weather Service is predicting 4-6" widespread throughout the DC area.

Another metric to consider is vertical velocity - a measure of the movement of the air.  In the case of storms, higher vertical velocities imply more intense storms and heavier precipitation. 

You can see from the graphic above that vertical velocities should be fairly high during the storm, suggestive of moderate to heavy precipitation.

Some snow amount considerations:

The NAM, which is a pretty good short-range model, shows the general thinking of most mets in the DC area - albeit with slightly higher values.  The map above shows 0.5-0.6 total liquid precipitation which, converted to snowfall, produces numbers in the 6-8 inch range. 

The GFS forecasts a similar range, maybe even a bit higher than the NAM - more in the 7-10 inch range for the DC area.

I feel that the NWS mets are being a bit conservative.  It seems to me that there's potential for more than 6 inches of snow with this storm.  Not crippling by any means but certainly more than we've seen around here in a while.

The evening models will be coming soon and I'll check those out as well.  The only impacts to this forecast that I can see are the time of storm intensification which is something the models usually handle pretty well.  My confidence in this forecast is, therefore, rather high - always considering, of course, that it's a snow forecast and it's for the DC area.

Increased Snow Chances / Amounts for Tomorrow

Well.  That was sudden.

As a frontal boundary approaches our area, it has been clear for a few days that the 21st would be cold, windy and that there was a chance that snow would overspread the area due to a small wave of low pressure riding along the front.

Our big snow producers typically come from the Gulf of Mexico and ride up the coast.  It takes quite a potent little low to produce a lot of precipitation if it has terrestrial origins. 

For the last couple of days, it looked like this would be a minor event.  An "inch-or-less" type of event.  Then came yesterday evening's model run.

The models run at 18z (1 PM locally) but since the upper air data from the day isn't available yet, they basically run from "inertia" or, more accurately, a "boatload of physics calculations" using the morning data.  Thus, they are used to get a general idea of the forecast.  They are rarely used for accuracy. 

In any case... the 12Z GFS which is based on upper air data had started to suggest the idea of a stronger and just as importantly, slower moving system.  The 18Z NAM then spit this clown map out and the meteorologists of the mid-Atlantic woke up from their nap and said, "hm!":

This shows a narrow area of significant snowfall in our area and especially points south and east.  Some of those totals are quite high but it's less about accuracy in the 18Z run and more about the overall suggestion:  a stronger, slower moving storm.

Overall, the storm has the look of a "clipper" system that then moves along the coast and intensifies rapidly. 

Overnight, the trend in the models continues to be stronger - except for the Euro.  Last night's 00Z Euro run showed this:

which still shows snow but in lighter amounts.

One this is certain:  it will be cold.  Precipitation type doesn't enter the equation like usual.  I expect high temperatures in the 20's which will result in high snow rations - perhaps as high as 15:1.  The map above shows 10:1 ratios so even taking 1.5 times the above snow produces a forecast for more snow than we've seen around here in 2-3 years.

This morning's data continues to come in and it all points to an increasingly significant storm for our area.  The RGEM, for example, calls for 7-10 inches:

The morning GFS is also in that range:

There is a Winter Storm Watch for our area and with the models trending stronger, this looks like it's trending towards a major event for our region.  For now, I'd say expect at least 4 inches of snow but the chances of more (maybe in the 8-10 inch range) are not out of the question.

There's unusually little variance with this storm as far as forecasting it goes.  I don't see any indications that it could end up producing nothing at all - but with the way winter forecasts go around here, I'm going to keep watching.

I'll update again later.