Saint Patrick's Day *POSSIBLE* Troll Storm

This will be my first ever attempt at tracking a storm that only exists on models at the moment. It will be updated quite irregulary, mainly depending on work load. It will be mainly a post about model forecasts and the evolution of the possible event March 17-20 period.

3/15/2014

7:30PM UPDATE:
As we draw nearer to the event, the model spread is still enormous. The 12Z Euro still brings it way to the south and way warmer than the other solutions. It is the southern outlier. The 00Z GFS has held the steadiest. It still simulates .6-.8 inches of liquid equivalent precip. The 18Z NAM has waffled North, putting us on the northern fringe of the bullseye. The 12Z CMC is similar to the NAM except without unrealistically high precip amounts. What I know is that any snowfall forecast over 8 inches for the area is unrealistic. Don't pay attention to those. Unless the NAM precip verifies, that will not happen.  The realistic amount would be 3-6 inches. borderline warning criteria, which is >5 inches of snow.

Original Text:
A new day dawns but it still seems that we don't have a clue about what's going to happen. Models have been flip flopping, shifting south and shifting north every run. There is little consistency. The 0z NAM gave us a crushing hit, but the 06Z NAM completely shuts us out to the south. This is considering that in the 18z run prior to the 0z run, the NAM had the storm going way north of us. The 00z CMC came in with a NAM-like hit. The 00z Euro came in way south, giving us around 1-3 inches although the WPC seems to want to discount it. The 06z GFS game in very similar to the 00Z GFS, maybe a tad north. It has been by far the most consistent model with this storm.

The WPC's thinking with this storm.
EVOLVING UPPER TROUGH OVER THE SRN PLAINS THIS WEEKEND
SURFACE CYCLONE IMPACTING THE DEEP SOUTH AND MID-ATLANTIC
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
PREFERENCE: BLEND OF 12Z ECMWF...00Z UKMET...00Z CMC THRU 12Z/17
BLEND OF 00Z UKMET/CMC FROM 12Z 17-18...F060-F084
CONFIDENCE: BELOW AVERAGE

A COMPLEX EVOLUTION OF SHORTWAVE ENERGY AND POOR RUN-TO RUN
CONSISTENCY IN THE MODELS REDUCES CONFIDENCE BELOW AVERAGE. THE
INTERACTION OF A SHORTWAVE CURRENTLY MOVING EWD THROUGH AZ/NM...A
SHORTWAVE DIVING SWD ACROSS THE NRN ROCKIES AND LESS DEFINED NRN
STREAM ENERGY FORECAST TO REACH THE DAKOTAS EARLY SUN MORNING HAS
MADE FOR A CHALLENGING FORECAST.

ENSEMBLE SPAGHETTI HEIGHT PLOTS SHOW MORE SUPPORT FOR A STRONGER
NRN STREAM SHORTWAVE OVER THE OHIO VALLEY BY MON MORNING AND FOR A
MORE ELONGATED TROUGH AXIS THAN DEPICTED BY THE 00Z NAM/GFS OVER
TEXAS AND OLD MEXICO...AT LEAST THROUGH MON MORNING. AFTER MON
MORNING...ENSEMBLE SPAGHETTI PLOTS SHOW FAR LESS CLUSTERING. THE
00Z ECMWF IS MORE AGGRESSIVE WITH THE SHORTWAVE AMPLIFYING SWD
THROUGH THE PLAINS SUN MORNING COMPARED TO ITS PREVIOUS RUN...AND
THEREFORE DOES NOT CARRY MUCH OF A NRN STREAM SHORTWAVE ACROSS THE
OHIO VALLEY BY MON MORNING. THIS CAUSES THE 00Z ECMWF TO BE
STRONGER/SLOWER WITH A SRN STREAM WAVE ACROSS THE GULF COAST ON
MON/TUE...WITH HARDLY ANY ENSEMBLE SUPPORT FOR THIS IDEA.

THERE IS DECENT SUPPORT FOR A 12Z ECMWF...00Z UKMET...00Z CMC
SOLUTION THROUGH 12Z/17...BUT BY 12Z/18...THE 12Z ECMWF HAS
TRENDED AWAY FROM ITS OWN CONTINUITY WITH A MORE SHEARED SHORTWAVE
ACROSS THE SOUTHEAST ALONG WITH THE UKMET. THE 00Z ECMWF DOES NOT
APPEAR TO BE ANY BETTER AS REFERENCED ABOVE. THEREFORE...BY MON
NIGHT...THE PREFERENCE IS FOR A 00Z UKMET/CMC BLEND...GIVEN THAT
THE GFS STILL APPEARS TOO SLOW/STRONG BY TUE MORNING.

The bottom line is that we may be too cold to snow- in mid march. The southern solution models show the high that is providing cold air to be so strong that is shoves the storm south.

3/14/2014
11:50PM UPDATE:
The GFS is much farther south with the storm than the NAM. Again... It is also much drier. .6 inch liquid equivalent as frozen precip as opposed to 1.2 inches on the NAM. The upside is that it seems to want to bring a stronger upper level low through than the NAM. That could be interesting. But otherwise, arghh... Don't want to look at it. There is little model consensus. Picking one over another is not wise. Good night. I'll post tomorrow when the GFS and Euro 00z runs  + the 06z GFS and NAM run are done. Hopefully we can get a better model consensus then.

10:45PM UPDATE:

The NAM is back to trolling...
It went from a cold heavy rainstorm to a massive snowstorm in 24 hours. The worrying thing is that it made its biggest jump south from the 18z to 00z run. If this doesn't stop, the storm will miss to the south. Sounds familiar doesn't it? The NAM buries central MD (including us) in 12+ inches of snow verbatim. Chop off a couple due to low ratios and initial melting, and you are still left with around 8 inches. Remember, this is just a model run, and not a forecast so don't get too excited (or depressed) yet. We have to either get support to confirm it or have the other models refute it. The good thing for snow lovers is that the NAM is no longer an outlier in terms of low track. However, it is far wetter than any other model. Now we wait for the GFS, which will come out next, around 12:00 AM.

ORIGINAL TEXT
Happy Pi day everyone! The models have mostly trended south and slower during the day with the first wave of energy. This is good to some extent because this keeps the cold air in place and let's the precip come in later at night when temperatures have dropped. There is a very large spread in low tracks and as a result, nothing is certain at the moment. The HPC thinks that we should be near the bullseye for at least 6-8 inches of snow with the bullseye to the usual north and west but that is subject to change.

HPC 50th percentile snowfall.

HPC low tracks. The bolded L is the official forecast track while the colored shapes are ensemble low positions at a given time. The greater the spread of points, the lower the confidence of the forecast.

We are awaiting the 00Z runs to see if the trends are going to continue or if we will stay in the bullseye. The 18z NAM is still the northern outlier, but has brought the storm a little farther south. The 18z GFS paints the bullseye of precip over us .8-.9 inch liquid equivalent of snow. Snow ratios will likely be quite low due to warm temperatures. The 12z Euro is slightly further south and drier with .6-.8 inch liquid equivalent. The 12z CMC has shifted way south from its previous 2 runs to be in line with the Euro. One interesting thing to note is that the models have trended further west and north with the second wave of precip with an upper level low. We will monitor this to see if we can score a second round of precip monday evening. The front end portion of the storm has the possibility of dropping 2-6 inches of snow- note the large range, this is a low confidence forecast.

Model maps below: Look at the snow swath positions, not how much snow falls.

GEM/CMC/Canadian Model backed off its absurd snowstorm.

The NAM is way far north and very heavy with precip where it does snow while we get dry slotted.

3/13/2014
First off, at 4 days out, the models are going to have a lot of trouble figuring out the storm. This is due to lack of sampling of the storms and because the calculations are not perfect. They are rounded and truncated all over the place. Compound that with possible observation errors and lack of data, models suddenly don't seem very reliable anymore. I did find a chart depicting the verification rates of of various model depiction of the 500 mb layer for a 5 day forecast.

According to the chart, the ECM(Euro) had the best verification rate, followed by the CMC (Canadian), which was a surprise. The GFS has not done too well and is third.

The 500 mb layer is up where large scale features such as troughs and ridges that drive our storms is in addition to mid-upper level lows. Large scale features are the most important at this range. You can't get a storm if the upper levels are crappy and many model runs can be discounted due to illogical presentations although sometimes they may be true so you need to be careful. The main problem confounding the models is whether or not to phase the 2 disturbances to form our storm. If the model phases it early, the storm goes farther north, bringing us precip. If the model phases it later, the storm goes south, leaving us dry. temperatures is also an issue. Temps will be well above freezing on sunday and will have to drop with the aid of high pressure to the north. This will be a cold air damming situation. The raw model outputs are usually too low so temps will be an issue as they have been all year. 

The 00z Model suite was very pessimistic for snow. The Euro a little snow and the GFS shunted the storm to our south completely. Only the Canadian (CMC/GEM/GGEM) gave us snow. I discount it because it is high on something and it had a very, very strange storm evolution. At least it shows hope for the storm.

00z GFS

The image below is a model output, not a forecast. This is extremely unlikely to happen. Do not look if you do not want to be disappointed.

00z CMC gives 15-25 inches across the area.