Sunday, December 16, 2012

December 19-22 Potential Blizzard Event (Updated 5:30 PM)

**This post is dedicated to Charlotte Bacon, age 6, a victim of the Connecticut shootings.**

This is an update to this morning's original blizzard post, which can be found by clicking here.

This update concerns the new models from the 12z suite, and not so much for teleconnections, which were covered in the morning post. Also, I have now established the two 'camps' for the models. The 'North Camp' involves models that take the system into north Illinois at Hour 96, and the South Camp has the system in central/southern Illinois at Hour 96.

We start with the GFS model, known as one of the better forecasting models that is out there today. The GFS is forecasting the storm system to be in north Indiana on the morning of December 20th. The solid blue line is the rain/snow line- any precipitation to the north of this line is theoretically snow, and precipitation to the south is rain. The storm system has a central minimum pressure of 994 millibars, meaning it is fairly strong. The forecast calls for snow to be ongoing in Illinois, Wisconsin and Michigan, with a potential severe weather situation ongoing in Kentucky and Tennessee. I will not make an assumption on if the GFS is right of wrong until I review the other models below. Needless to say, the GFS is in the North Camp.

This is the European's ECMWF model forecast, once again valid for the morning of December 20th. The ECMWF is widely regarded as the best forecasting model used for daily weather forecasting, like this. The model predicts that the storm system will actually be in south Illinois at this time, thus being further south and slower than the GFS. The central minimum pressure is down to 995 millibars, roughly the same strength as the GFS. The ECMWF solution has shifted south from recent forecasts, not a surprising thing to happen in the winter. I am cautiously optimistic with this forecast, as I think the teleconnections fit in nicely here, but models are notorious for shifting south in this time period, only to shift back north later on. The ECMWF is in the South Camp at this time.

This is the forecast from the ECMWF Ensembles, also called the ECMWF EPS. The ECMWF EPS can be thought of 51 separate ECMWF model forecasts, each with slightly different starting parameters. This forecast, once again valid for the morning of December 20th, has the storm system in southern Illinois, actually a tad farther south than the ECMWF model itself. The minimum strength is at 1002 millibars, meaning it is fairly weak. However, there is something slightly more significant. The fact that an average of 51 different forecasts, all originally based off of the best forecasting model known to man, came up with such a solid solution is very supportive at this point for the South Camp.

The current overview of the models is this:

North Camp: GFS, GFS ENS (2 models)
South Camp: ECMWF, ECMWF EPS (2 models)

We now move to the Canadian's GGEM model. It shows the storm system in central Illinois with a central pressure of 1000 millibars. The blue rain/snow line remains in North Illinois, but far enough south for cities like Des Moines, Chicago and Milwaukee to get snow at this point. Considering the GGEM is taking a cold solution to the mentioned cities, I feel like it belongs in the South Camp.

This is the forecast for precipitation, 850mb temperatures and the freezing line from the GGEM Ensembles (GGEM ENS). Although the Canadian model is not the best, ensembles are always better than models. The GGEM ENS believe that a solution not all that different from the ECMWF will evolve, with the storm system taking a path through south Illinois and snow given to the previously mentioned cities. Heavy rain in TN/KY suggests a possible linear storm event could be ongoing. This is indeed a South Camp storm.

Another model overview at this point in time tells us the following:

North Camp: GFS, GFS ENS (2 models)
South Camp: ECMWF, ECMWF EPS, GGEM, GGEM ENS. (4 models)

This is the image of the GFS Ensembles. Yes, I know the ensembles were already listed in the North Camp, this is just for you all to see what it says. The GFS Ensembles (GFS ENS or GEFS) shows a major snow event in much of the Midwest after the storm system blows through. Rain is ongoing in southeast Michigan, and again that linear storm threat appears with the states of Kentucky and Tennessee.

Here's the final total after reviewing some other models:


North Camp: GFS, GFS ENS, JMA NOGAPS ENS,  (4 models)
South Camp: ECMWF, ECMWF EPS, GGEM, GGEM ENS, FIM. (5 models)

Analysis: This is not as difficult as you may think. Despite the input of 9 models and ensembles, only a select few are worth making note of. The GFS/GFS ENS should be watched carefully, as should the ECMWF/ECMWF EPS. The JMA and NOGAPS ENS are worthless; I threw them in there for some diversity. As for the FIM, it has historically verified slightly better than the GFS, so it can be taken note of. I would watch the Canadian model and ensemble set, but not as closely as the American and European sets.

Forecast Preference: ECMWF/ECMWF EPS
Confidence: 60%

Snow amounts, totals and eventual track will need to be determined at later dates, so please do not ask me for snow totals for your city, because I just don't know yet.

Andrew

December 19-22 Potential Blizzard Event- First Call

This is my first call on the potential December 20-22 blizzard event. Let's do a model rundown first, then review the teleconnections before we get into the forecast.

We start now with the American's GFS model on the evening of December 19th. Pictured above are 850mb temperatures in the dashed lines, sea level pressure values in solid black lines, and precipitation values as filled in colors. The rain/snow line is the farthest-south dashed blue line. Any precipitation above that southernmost dashed blue line is supposedly snow. The GFS predicts we see our storm system coming out of the Rockies on the evening of December 19th, with a snow event ongoing in the Plains. Accumulations could range from 1-3 inches at this time in the forecast. The presence of a ridge of high pressure in the Southeast tells me that this storm system will be forced more towards the north rather than along the Gulf Coast.

This forecast is for 12 hours later, and is valid on the morning of December 20th. We now see the storm system has fled northeast into Chicago, and the central minimum pressure is approaching 990mb. Precipitation shown above tells me that a light snow event would be ongoing in Wisconsin and possibly Iowa, probably accumulations of 1-4 inches. In the southeast quadrant of this storm system, heavy rain and thunderstorms are ongoing. The tight pressure gradient, tight temperature gradient and overall strength of this storm system tells me a linear severe weather threat certainly is possible for much of the southern Midwest, southern Ohio Valley and general Southeast US.

Snow accumulation forecasts favor Wisconsin as the big winner per the GFS:


So the GFS takes the system into Illinois and gives Wisconsin the big snows. Let's now turn to the European model.

The ECMWF model, regarded as a more trustworthy model than the GFS, shows the storm system ejecting from the Plains in Oklahoma with a central minimum pressure of 995 millibars. Again, we see that high pressure system in the Southeast which will support the prospect of this storm system staying into the Midwest and Ohio Valley. This forecast is valid for the evening of December 19th.

Moving into the evening of December 20th, we see that our storm system has now strengthened greatly since the 24 hours before, with the central minimum pressure now at 981 millibars. That is a change of 12 millibars in 24 hours- not a common sight away from the Northeast. The tight pressure gradient tells me that this will be a very windy event, and if cold air factors into the equation after the storm departs, lake effect snow could be on overdrive. Also, note that piece of lower pressure jutting out of the southern side of the storm system. This could be a push at another storm system developing, one which could go into the Northeast, but that remains to be seen.

A few images of the ECMWF's precipitation and temperature forecast centered over the Midwest:

Valid the morning of December 20th.

Valid the afternoon of December 20th.

Valid the evening of December 20th.
Now that we've covered the two big models, let's visit a couple other forecast models not as reliable, but still worth a glimpse.

This is the Canadian's GGEM model forecast for the afternoon of December 19th. Pictured above are sea level pressure (SLP) values for this timeframe, and the colored areas depict 10 meter wind speeds. We can see our storm system in north Texas, being squeezed towards the northeast as pressure from the ridge offshore the Southeast increases. The storm system has no choice but to shoot northeast- that's one thing we're confident in at the moment. The GGEM predicts the system will be much weaker when it exits the Rockies than other models are predicting.

Moving ahead to the morning of December 20th, we now see that the GGEM has taken the system into Indiana with a central minimum pressure of 994 millibars, indicating that some strengthening has occurred. Despite this, the system remains weaker than the GFS and ECMWF are predicting. Even though a strength difference may seem like nothing, I can assure you it can have significant implications. If the system is weaker than is being forecasted, the track can (and would) significantly change. Let's hope that doesn't happen, or this whole post would be moot. We still see that high pressure in the waters off the Southeast. Despite the presence of a tight pressure gradient, 10 meter winds are not showing up too strong, which I can understand, considering the GGEM is weaker than the GFS/ECMWF at this point in time. I am hesitant to disregard the GGEM, because it is in line with the GFS, meaning the two may be on to some sort of consensus.

Now, for the last model we look at, we see SLP values in solid black lines, 850mb temperatures in dashed lines (the same rain/snow rule as the GFS applies here), and precipitation values in colors. This is the NOGAPS forecast, made by the Navy but not regarded as a solid model that you can trust. I'm including it moreso to give some spread to the forecast, not so much to find a solution from the NOGAPS. This model starts out with the system ejecting from the Rockies at a minimum pressure of 993 millibars in the early morning hours of December 19th. Again, watch that timing difference! An already-tight pressure gradient in OK/AR/TX/MO tells me a severe weather situation may be trying to form in that area.

Moving ahead to the early morning hours of December 20th, we find the NOGAPS projecting this system to possibly have split in two. We see one storm system with a 1005mb minimum pressure centered in Arkansas, and another storm system centered in Wisconsin with a central minimum pressure of 1000 millibars. I will refer to them as System 1 and System 2, respectively. System 1 shows a more linear alignment of precipitation, telling me that if this solution verified, we would probably see a mesoscale convective system (MCS) or small squall line in the area as a result of tight temperature and pressure gradients in the region. As for System 2, this looks to be the main piece of energy. The precipitation looks to be a modest snowstorm, with accumulations around the 3-6 inch mark based solely on this image. If the two systems had stuck together, a solution similar to the other models may have evolved. Needless to say, I am discounting the NOGAPS model.

Teleconnections are also a big deal, in terms of the North Atlantic Oscillation (NAO). When the NAO is in its negative phase, you'll see a tendency for storm systems to move up the Northeast coast and strengthen rapidly into Nor'easters. This comes in response to the highly meridional flow set up by the high pressure system in Greenland and the storm system in the Northeast. There are two sub-categories of the NAO, and that is west-based and east-based. In a west-based negative NAO, you'll see the high pressure to the west of Greenland, hence the term west-based. Likewise, in an east-based negative NAO, you have the high pressure system to the east of Greenland. The difference is which one affects which areas. In a west-based negative NAO, you'll see the East US getting in on the most cold and storms, while the East-based negative NAO keeps the cold and storms in the Plains and Midwest. The forecasts for the west-based negative NAO are stronger than the east-based, so it is plausible to think that the East Coast has an advantage. However, because of that Southeast ridge, we'll see the track take a detour north and then go northeast through the Great Lakes. The whole situation is very complicated, but in summary, the NAO does favor a track being shown by the ECMWF or even the GFS, maybe a bit farther south if the negative PNA holds in place.

This image shows what the meteorologists at the Hydrometeorological Prediction Center (HPC) think in terms of best model choice for the next 7 days. Looking at the D4 and D5 areas, we see that the HPC generally leaned towards the ECMWF model, the ECMWF Ensembles (ECMWF EPS) and the Canadian model. Surprisingly, no mention of the GFS model is recorded.

Going off of all the models printed above, the teleconnection forecasts and the HPC forecast preferences, I have made my first call as printed below.


Forecast Preference: ECMWF/ECMWF EPS
Confidence: 55%

Andrew