Underperforming

No one likes it when a forecast does not work out well.  The public feels misled. The meteorologists feel frustrated.  Only misanthropes are happy when this type of thing happens.

I take only two comforts from this particular storm.

First, my forecast totals were generally lower than many other outlets.  Not all… but many. I carried 3-6″ in Lynchburg. We got two.

Second, most of the meteorological community figured the heaviest amounts would be a short drive north of Lynchburg, and that was the case… with much of Amherst, Bath, Nelson, and Rockbridge Counties getting 3-6 inches.  Unfortunately, I was going with a 4-8″ idea in those places (with a few areas approaching 10″).

The late afternoon burst of snow got much of Southside into the 1″ territory, only edging into the 1-3″ range that was forecast.

Roanoke was probably the biggest bust.  Only an inch there, where I was carrying 2-5 inches.

A colleague in Richmond, Dave Tolleris, runs a weather site called wxrisk.com.  Dave has built quite an internet reputation for himself.  While his communication style may not fit everyone, he knows what he is doing.  Dave has an excellent post-mortem, which he has given me permission to repost below.  Enjoy… while I go compile snow reports.  -ss

* * *

Over central and southern portions of the Piedmont… as well as southwest Virginia and in Central Virginia in and around the Richmond metro area… the snow accumulations were killed by the appearance of the dry slot. For those of you who are not familiar with what this is, large areas of LOW pressure will pull in dry air from a nearby location which gets pulled into the center of the low. In doing so, this dry air can break apart clouds (and moderate or heavy precipitation), so that a dry area (or one of little precipitation) develops.

Anytime you have a winter storm… or even a regular low pressure area coming from Tennessee and Kentucky and southwest Virginia… you always have to be very conscious of the dry slot appearing… and coming up from western and Central North Carolina into Central Virginia. The dry slot effect can also be enhanced by the mountains over Tennessee, West Virginia, and southwestern Virginia which can break apart precipitation whether it’s rain, thunderstorms, or even snow. The point here is that I, along with many other forecasters, were looking to see if there was going to be a dry slot with the system… because we all know that this can happen when you have LOW pressure areas coming in from Tennessee and tracking into North Carolina / Southern Virginia.

The short-range RAP model snow forecast from the from 8 am MONDAY RUN for the next 12 hrs showed not even the slightest hint of the dry slot. There was a weakness in the heavy snow shields over Lynchburg and the southern Piedmont area, but certainly nothing like a massive dry slot which appeared at midday over the Piedmont and Central Virginia. The RAP is specifically designed for a very short-range mesoscale features, and it should have been for more accurate. Another one of these short-term high-resolution small-scale weather models… known as the HRRR… at 7am had 7-8″ over RICHMOND.

The truth is that this was an extremely difficult system to forecast at all levels. From several days out when this Low / system appeared to a Pennsylvania New England snow… then a Maryland Pennsylvania NYC snow… then a good Virginia Maryland snow… and now it has turned out to be a very ordinary system. In some ways this is symptomatic of the seasonal pattern which has seen weather systems several days out come across as a big East Coast winter storms only to end up as a more moderate system when the low pressure area finally develops bring in the precipitation.

So, from the extended range… through the medium range… and to the past weekend up until early Monday morning all the models at every single stage really failed quite badly at handling the system. This is also true with the European model which also over-forecast the snow amounts in Kentucky and Tennessee as well as Central West Virginia.

I suppose one could make the argument that given how badly the models have handled this particular system… how they were always playing catch-up to reality… how the cold air kept coming in faster and faster forcing the system to shift farther and farther south…. that this should have been a warning sign that something else was going on. On the other hand, weather models should be able to detect the dry slot within 24 hours before the event begins. It’s not 1964…. it’s 2014 and it really should not be that much of an unreasonable demand.

One can make the case that because the system was crossing the mountains of eastern TN eastern KY, WVA that there would this gaps in the snow shield. But again, the other side of the coin is that because this system was moving so fast one could make the argument that the impact of the mountains on this sort of low and its precipitation would be significantly less.

A difficult, challenging, and sobering forecast.

Advertisements

About seansublette

Meteorologist at Climate Central. Broadcast meteorologist in Virginia from 1995 to 2015. Born and raised in Richmond, VA. Penn State alumnus. Loves baseball and the rock band Rush. Views are independent of my employer. Past performance does not guarantee future results.
This entry was posted in Forecast Discussions, Weather Communications and tagged , , , , , , . Bookmark the permalink.