Forecast Updated 10:50 PM — Tonight’s GFS just available continues with the high snow totals, even in Philadephia, similar to those earlier today.     I  guess we’ll have to ignore the NAM group of models low snow totals.

Forecast Updated Sun 9:12 PM — Tonight’s NAM model has become available.  The NAM model continues to show a significant mixing with sleet after midnight as the warm front event continues.  Total snowfall at at 7AM is unimpressive around the city—

NAM Snow totals 7 AM  (Click on image for a larger view.)


The NAM has most of the accumulation resuming after 11 AM Monday as everything changes back to snow.

By 7 PM —

NAM Snow totals 7PM  (Click on image for a larger view.)

Some snow continues into the evening hours.

Even less impressive is the higher resolution NAM-NEST—

NAM NEST 7 PM Monday (Click on image for a larger view.)


Be advised that most of the models today have continued with the much higher amounts that are highly advertised in the news.   And those models may very well be correct.  But one has to take into account that the NAM group has much more sleet and lower snow totals, especially near the city.   

The GFS group and other models become available about 10:50 PM.  It might be too late to do graphics, but if things change significantly, I’ll update after 11 PM.  Otherwise, look for an update tomorrow morning.

Forecast Updated Sun 3:56 PM — Despite the impressive snow totals predicted by most of the models, the NAM continues to forecast much lower snow totals.  The afternoon NAM just became available.  The NAM is insistent on showing temperatures at 6000 feet (800 mb) to be above freezing, mixing the snow with sleet and reducing snow totals by at least 1/2  compared to the GFS model.   Below is the current forecast position of the NAM 32º (~6000 ft)  isotherm at 10 AM Monday.  Areas south and east of this line may have significantly reduced snow totals if the NAM is correct—

NAM 32º isotherm (pink) at 6000 feet.  (Click on image for a larger view.)


The NAM-NEST and Canadian High Resolution are similar. Despite the impressive high snow total headlines of the other models,it’s never good to ignore the NAM model”.

NAM Snow totals—

NAM snow totals by Tuesday morning (Click on image for a larger view.)


In past storms, the average between the GFS and NAM has been a better forecast when there’s been so much divergence.  


Updated Sunday 2 PM— more graphics added at the end of this post…

The morning models have come out.  The models are in fairly good agreement forecasting a significant snowfall for our region. There is still some issue with a warm layer moving in at 6000 feet (800 mb)  which would reduce the accumulation in Philadelphia, south and east.   The latest GFS is colder and has more QPF than previous runs.

This will be a prolonged event with a warm front (warm air advection snowfall) followed by a blocked, slow-moving coastal secondary low. The snowfall totals quoted below are through Tuesday evening.

The latest GFS model data provides a reasonably good estimate of the currently expected snowfall totals. It’s inline with many of the other models, perhaps on the high side.  But it’s the latest —

GFS snowfall by Tuesday evening.  (Click on image for a larger view.)


Additional snow total graphics added below at 2 PM:

The latest ECMWF snow totals—

ECMWF snowfall totals  (Click on image for a larger view.)


German ICON model

German ICON model snow totals  (Click on image for a larger view.)


High Resolution Window FV3 (experimental) model—

Experimental HIRESW (High Resolution Window) Snow-Liquid equivalent 1:10 ratio without compaction  (Click on image for a larger view.)


Canadian Regional GEM model

Canadian Regional GEM model  (Click on image for a larger view.)


For the record, the NAM has warm air moving in at 6000 feet, reducing the above snow totals by about half near Philadelphia.  With so many models at the higher end, I’m ignoring the NAM snow totals.  (The NAM QPF values are consistent with higher snow totals.) 

Here’s the NAM model snow totals (without the fancy graphics).  It forecasts a fraction of the other models snowfall near Philadelphia—

NAM snow totals  (Click on image for a larger view.)


More updates later today.  Stay tuned…



Forecast Update Sun 8AM —  Quick update.   Snow starts between 2PM and 4PM this afternoon from south to north.  Accumulations 9-12 inches from Philadelphia, including much of the city and northward.   Higher (12+) accumulations northern Montgomery and Bucks and Lehigh  counties.  I’ll have more info based on the morning’s models later .   Stay tuned. 

Next Update about 1PM Sunday

Forecast Update Sat 6 PM— This remains a high uncertainty, low confidence forecast. 

The NAM group of models has a significant mix with sleet and rain, but maintains surface temperatures below freezing in most of Philadelphia. More freezing rain.  The NAM was wrong in the same direction this past week and its snow totals are running 2 inches or more less than the GFS model (shown just below).

NAM 7AM Monday with 32º isotherm at about 6000 ft  Warm air aloft will create precip mix and reduce snow totals. (Surface temps, white line is 32º.)   (Click on image for a larger view.)


The GFS group of models is warmer at the surface, but also has a rain sleet snow mix for almost 8 hours from midnight Sunday through part of Monday morning.  Despite a QPF of 1.5 inches of water, it cranks out only about 6 inches of snow (Blue Bell).

GFS PTYPE forecast 7 AM Monday. (Click on image for a larger view.)

 GFS snow totals based on latest model run this afternoon—

GFS snow totals Tuesday by Tuesday morning. (Click on image for a larger view.)


The Canadian models have 2-3 inches more snow than the above.  The NAM 2-3 inches less!


Look for an new update Sunday morning!

from Sat morning…

This remains a complicated forecast.  There are uncertainties about the low pressure system track, timing and degree of intensification.    All of these uncertainties affect the amount of moisture falling as snow, the temperature profile (meaning 3-D profile) and the duration of the storm.

The general sequence has not changed.  A “warm air advection type snow (warm  front)  will occur Sunday afternoon where warm air rides above a cold layer at the surface.  About 1 ” of snow is possible by early Sunday evening in most areas. 

Meanwhile, a secondary low pressure system will develop along the coastline and will move slowly northeastward, being blocked by a high pressure ridge in the mid Atlantic.  This will result in additional snow, lasting into late Monday night!

The main complication is the amount of warm air at 3000-6000 feet that moves in with the initial warm front and the subsequent development of the coastal low bringing those same 3000-6000 ft temperatures below freezing.

Further complicating the picture is the development of the secondary low circulation that is multilobular with at least two circulation centers—

GFS surface forecast showing two centers at 9 AM Monday.  The GFS has shown this for at least a day. This further complicates the forecast for the precipitation field and the thermal profile.  (Click on image for a larger view.)

Switching to the model blend (NBM), which is designed to take most available models and statistically weight/blend them based on history and hourly changes, it shows mixed precipitation Monday morning with surface temperatures above freezing—

NBM Precip TYPE at 10 AM Monday  (Click on image for a larger view.)

This mixed precip will further complicate accuracy of snowfall prediction.

All models have the mix changing back to all snow during the afternoon and evening Monday. 

The current European (ECMWF) model has the following snow totals—

ECMWF snow totals early Tuesday morning  (Click on image for a larger view.)


My own snow totals, based on variation of the NBM 50 percentile snow accumulation—

NBM snow totals based on modification of snow accumulation 50 percentiles  (Click on image for a larger view.)

I expect changes as the storm time horizon falls into the range of more short range models.  Stay tuned.

Just added Sat 10 AM: NAM PTYPE with 32º isotherms at critical 3 D levels

Latest NAM with precipitation type and freezing isotherms 10 AM Monday forecast  (Click on image for a larger view.)



Fri 11 PM Update— Tonight’s models just becoming available suggest that widespread accumulations greater than 12 inches are possible with this storm.  I’ll update in the morning.

The coming snowfall for the Philadelphia area is now looking to be significant.  All models have snowfall starting Sunday afternoon and lasting into Monday evening as a warm front approaches, followed by a secondary coastal low pressure system.  Most models have all snow, but the GFS has a mix with some rain late Sunday night into Monday morning before changing back to all snow. 

The majority of the accumulation is expected during the day Monday.

The latest GFS model snowfall amounts are—

GFS snowfall forecast by Monday evening  (Click on image for a larger view.)

I’m posting the GFS because it probably best captures the snowfall range of all the models.

However, I think that these model-calculated snowfall totals may be the low end of the spectrum and it assumes that some of the precip will be mixed with sleet/rain for a period of time increasing compaction.    If there’s no mix to rain, the snow amounts could almost double in some areas.  Indeed, a straight QPF to snow calculation gives many areas over 12″   Many models are forecasting QPF values of over 1-1.5 inches of water.

The ECMWF, ICON, NAM, SREF and other models are similar, although they each have their snowfall maxima at somewhat different locations. 

I’ll be updating through the weekend.  Stay tuned.


Usually, I update this blog with the ‘the latest model runs’ and the forecast trends.  

As the time frame of the Sunday-Monday storm comes into the range of the higher resolution models, this  potential snow storm is presenting a higher level of variation and  uncertainty with each model addition rather than a clarification of the likely storm snowfall.   Usually, I’d supply some snowfall maps, but it’s a waste of time right now.  Here’s why:

  • First, the European model has backed off of its extreme snowfall amounts for our area. 
  • However, the German ICON model has gotten extreme with its totals. 
  • The GFS has remained somewhat consistent with a mix with rain on Monday. 
  • The Canadian Global and Regional GEMS remains colder with only snow, no rain mix
  • The NAM has come into forecast range and it has a mix with rain/sleet.
  • The SREF forecast has come into range and it’s somewhat higher with snow totals.

Rather than present snowfall maps, I’m going to present maps that capture the current uncertainty—

ICON Statistical Ensemble model showing large variations in the possible position of the coastal low. (Click on image for a larger view.)


Here’s the Short Range Ensemble Model (SREF) showing high spread in standard deviation of the surface pressure. High spread suggests high uncertainty in the low pressure position .

SREF ensemble Pressure/standard deviation Monday 10 AM  showing high spread in pressure values in the NE quadrant of the storm.  This could suggest faster movement or uncertainty in position/configuration of the surface low.  Equally uncertain is the high pressure in NY State.   (Click on image for a larger view.) 


All this said, here’s the large scale “Snow-Liquid” forecast from the SREF suggesting the statistical areas where highest snowfall is expected—

SREF snow-liquid forecast (correlates with snow totals)  This suggests where the highest snow accumulations will be.   I’m intentionally not putting numbers on the graphic.  (Click on image for a larger view.)


So, the takeaway is that we’re getting several inches of snow, starting Sunday afternoon and running through Monday; high uncertainty about specific snow totals in any one specific area.   Sometimes, the best forecast information is to indicate what we don’t know.

I’ll take another attempt at specific snow forecast this evening, when we have today’s model runs to review.   Stay tuned.




Thu 05:58 PM Update — The Sunday time frame is just coming into range of the NAM and Canadian Regional models (84 hours). 

The general trend with today’s models has been for more snow.  The ECMWF model has as much as  8-12 inches for our area, but it is on the very high end of the scale right now. I believe the ECMWF, despite its stellar reputation, is over-doing the snow.

ECMWF (European model) snow totals by early Tues morning.  (Click on image for a larger view.)


The latest GFS forecast shows significantly  less snow, the difference being a further east development and less intensification of the coastal low.  The GFS brings warm air in during the storm, with a mix of snow/sleet at times reducing accumulations.

GFS snow totals by Tuesday morning  (Click on image for a larger view.)


The NAM, Canadian regional GEM and the ECMWF have the snow starting late afternoon/early evening Sunday, while the GFS and model blend (NBM) have some light snow starting earlier in the afternoon Sunday with the warm front.  

As mentioned, the storm consists initially of a “warm air advection type snow” on Sunday afternoon from the original low pressure in the Midwest. Then, the bulk of the storm’s precip comes on Monday with the expected secondary coastal low.  The Monday time frame still beyond the forecast range of all of the higher resolution models.

Stay tuned. 

I’ll update tomorrow morning when we have some newer model data.

from earlier this morning…

Last night’s models have become available.  Here’s the breakdown on current model forecasts— The GFS is more disorganized with the secondary low formation off the coast.  As a result, the primary low’s circulation continues to bring in milder air in our area through a low level jet circulation instead of having a colder northeast flow from the secondary—

GFS low level 925 mb jet flow 1 AM Monday shows SE flow (Click on image for a larger view.)

As a result, the GFS has a changeover to rain for some part of the storm in the immediate PHL area and suburbs, keeping the snow accumulations to 1-4 inches.

GFS snow totals by Tuesday morning. It’s visible where that low level warm air ‘ate away’ at the snow accumulation.  (Click on image for a larger view.)

The GFS sequence is snow (Sunday)–> rain mix (Monday morning)  –> Snow (Monday afternoon.) 

On the other hand, the Canadian CMC-GDPS and European ECMWF have it colder but develop the storm further east and north.  The have somewhat different snow totals and for different reasons— the secondary storm is stronger and colder, but is slower to develop and is further east.  They don’t show the changeover to rain, but they have less total precip for our immediate area.

The model blend (NBM) in recent years has been quite good (a recent exception being this past Monday).   Here’s its snow totals (average)—

NBM mean snow total Forecast by Monday eve (Click on image for a larger view.)


The current takeaway from all of this early prognostication is that the initial warm front snow starts as early as noontime on Sunday. About 1 inch of snow likely by Sunday evening. 

The coastal storm, depending upon track and development, affects us on midnight Sunday into Monday with additional accumulations. 

Look for more updates later today.