6:00 PM

Now that the winter storm is over, it is time to see how I did with this storm.  The forecast analysis appeared to be a picture of two worlds.  Over the interior, I feel I did poorly which I will discuss what went wrong.  Meanwhile, the forecast along the coast could not have been better.

Here is how I broke down the grading.  The snow map (seen below) is broken down into state sections based on climatology and typical mesoscale impacts.  I also added two grades based on long range threat analysis (catching the threat of the storm 3 days out) and short range threat analysis (3 days through the event).  The grades are broken down as the following:

A+= 100      A= 95       A-= 90     B+= 87     B= 85     B- = 80     C+= 77   C= 75    C-= 70

D+= 67        D= 65       D-= 60      F=55     EP= Epic Fail  0

The grades basically range from complete verification of the snowfall forecast to an F which is where snow occurred but no where near verification.  An EP is where snow is forecasted and none accumulates OR where no snow is forecasted and snow ends up accumulating.

As we can see with the map to the left, I did not do well at all over the interior.  I looked over the forecast and how I went about making this forecast.  The problem with the interior was that I was too aggressive with the moisture advection in these locations.  The dry air at the start of this event killed what easily would have been a six inch snowfall.  I should have picked up on this in the morning when dew points were around -1 degree F.  The best forecast in hind site would have been the map I issued on Thursday, which is the next map to the left.  This map would have covered the entire forecast area much better.  However, I was pulled into

Thursday's Snow Map

believing the higher precipitation potential due to the very impressive radar signatures observed on Friday afternoon and the convergence of almost all guidance on a heavy precipitation event from the GFS to the ECMWF to even MM5 guidance.

The forecast map to the left would have been a much better verified forecast and would have lead to a much higher verification score.

Long Range Threat Analysis: A+

I was able to pick up this threat almost a week in advance based on a review of past posts.  I also remained fairly consistent with illustrating the threat that was present and describing why this storm had such a high potential for becoming a significant if not historic winter storm.

Short Range Threat Analysis: B-

I feel I did well with picking up the mesoscale threats and impacts of this storm and did not waver in the threat analysis while the model guidance was anything but consistent.  However, I lose points in jumping on the high precipitation idea over northern sections and failed to pick up the degree of impact the dry air in place on Saturday morning would have on the storm in northern sections.  Good lesson on how virga can reek havoc with a forecast!

The other issue was the snow banding over eastern Pennsylvania.  The problem with this forecast is not illustrating the threat correcting.  Although I did warn that some will miss out on the snow due to banding to the east or west of their location, I should have been more forceful in describing the impact.  The lighter snow over parts of Pennsylvania just to the northwest of Philadelphia was directly due to the intense banding over Philadelphia.

However, I am very happy with the forecast from Philadelphia through central and southern New Jersey, New York City, Long Island, and Connecticut.  The threat for a foot to two feet of snow was illustrated and warned of three days ahead of the storm and those locations were properly warned of the threat.

Overall score:

Map Verification Score= (Total combined score divided by number of sections) 87.5 = B+

Total Score:  89.2  B+

Tags: analysis 3, c 70, dew points, forecast area, inch snowfall, map, moisture advection, NY NJ PA Weather verification, radar signatures, state sections, threat analysis, two worlds, winter storm