Over the past few days, hopefully you’ve been following along to see the progression of this storm forecast. If not, you’ve missed out. First we had a confluence zone along a frontal boundary set up near the region resulting in what looked like 2-3 days of precipitation. Then it was 2 rounds of precipitation, and now it is looking more like one round as some model guidance shows the frontal boundary pressing further south than expected and suppressing the main low pressure center that would give us the majority of our snow out to sea. However, differences among the models remain, and as we have seen it is quite unlikely that they have settled in on their final solution. To make this even more confusing, the main piece of energy associated with this storm system still has not even come onshore yet; it’s sitting off the California coast and will give them a serious rain storm tomorrow and Sunday morning. The upper level energy then proceeds to weaken and race across the country, and it is often these fast-tracking upper level pieces that can be hardest to forecast because models may miss connections and interactions when attempting to model the speed. This forecast is going to break down generally why I think the GFS model, that suppresses almost the entire storm out to sea on Monday, is likely overdone, but also why what many thought would be a very significant snow event will likely only be moderate at best, and why I never bought into the idea of too large a snow event (though I was worried about long duration for a period of time, and that is clearly incorrect). I’ll start by detailing the current forecast, impacts, and where models differ. I’ll then go into what I expect models to do over the net 48 hours and what solution they will likely converge in, so you can see both what we have on the table what right now and what I expect the end result to be. This is a tricky storm and the outcome I expect is no means the one that is guaranteed, but rather what experience has taught me to favor.
TIMING/PRECIPITATION TYPES: My goal here is to avoid being too technical (though I may finish with a technical big about how models are handling upper level patterns; it’ll be labeled) so this section is going to be straightforward expected impacts. The problem is, while yesterday models at least agreed about the onset of the storm, that disappeared today as they all went their own ways. Models generally agree that at some point Sunday afternoon precipitation breaks out across the region. They don’t agree on the precipitation type. Some show rain mixing with sleet/freezing rain before turning to snow when the cold front moves through. Some have all the precipitation behind the cold front so thus it should be all snow. Ironically, those of you that want more snow may support a solution where the precipitation doesn’t start as all snow, as that means the cold front is not pressing as far south and the wave of low pressure may slide further north and thus impact us more on Monday. Either way, all precipitation during the Sunday should be extremely light, if it’s snow it would be less than an inch through 7 PM and if it’s rain there would be no major impacts; I’m watching the freezing rain threat but confidence is not there for me to worry about serious icing at this time. Between 7 PM and 1 AM Sunday night the warmer models begin to show colder models move in, and by 1 AM just about every model supports snow across the region. That 6-hour region could still result in some sleet and especially freezing rain, though, which I’m watching closely still. Inland areas especially are susceptible to this threat, but I don’t think any icing will be a problem until the overnight hours if it is. Again, some models show the entire storm as snow. Between 1 AM and 7 AM we would likely see precipitation begin to pick up in intensity, and between 7 AM and 1 PM I think we see the heaviest precipitation across the region, assuming that the storm is not totally suppressed to our south. The GFS weather model, to put this in perspective, has most precipitation end by 7 AM because it shunts the main storm out to our south, and we would only see a couple inches overnight Sunday night. Most other weather models have at least some steady snow during the day on Monday, and I think that is the more likely scenario as I believe the GFS is overcorrecting, which I’ll explain in a little bit. Between 1 PM and 4 PM on Monday I expect the snow to begin to wind down then, and by Monday evening precipitation should have entirely moved through the area. Overall, I think worst conditions occur Monday morning unless there are major forecast changes and the GFS ends up being right.
ACCUMULATIONS: The going forecast is for 4-8 inches of snow across Southwestern Connecticut. This is a VERY tricky snowfall forecast for a number of reasons. For one, we don’t know yet if the precipitation Sunday afternoon, evening, and potentially overnight will be all snow. All short range models say no, all longer range global models say yes (except the CMC, which I don’t buy, even though I support its shorter range member here, which I’ll get into in a minute). Along with that, when snow starts the warm air will be near as the cold front will have just moved through, so snow may be heavy, dense, and wet. By the end of the storm, we will have some very cold air dominating the region, meaning that snowfall ratios will be higher and the snow will be much colder and powdery. What I’m seeing here is a HUGE discrepancy in snowfall ratios between the end of the storm and the beginning, again making the snowfall forecast harder. What may start as 9:1 or 10:1 ratios could end with up to 15:1 or 20:1, though again it is hard to say as the further north the low pressure moves the less absolutely frigid air sits over the region. Inland areas will likely see slightly higher ratios to even out lower amounts of moisture, though as the storm approaches it may become more clear that the coast will get snow and I may adjust the forecast accordingly. The 4-8 inches is just to give you a general idea of what to expect, it is by no means a final call for the storm and there is a good chance that I will need to change and detail that more tomorrow. What I generally expect is around .4-.5 inches of liquid to fall as snow across the area, and if ratios average around 10:1 then we see 4 inches of snow, and if we see ratios average closer to 14:1 or 15:1 (mainly only possible inland if it happens at all, which is still fairly unlikely) then we could hypothetically see up to 8 inches of snow. Again, a northern trend with less suppression would bring in more moisture which would make 8 inches a reasonable forecast. But going off most model trends, not significantly weighting the most recent run of the GFS, I expect 4-8 inches of snow across Southwestern Connecticut, by no means a very significant snow storm, but definitely a nuisance to borderline significance. I would say there is about a 15% chance that I will have to raise these amounts and about a 20-25% chance that I will have to lower them if the storm does end up being as suppressed as a few models show. This is actually very closely in line with the ECMWF weather model, which shows ratios a little over 10:1 but generally around .6 inches of QPF or so, so amounts around half a foot.
SCHOOL IMPACTS: Now I want to focus on school impacts, because I know many of you are checking in for that. Tuesday I believe schools will open on time now; the storm ending Monday afternoon means district have time to clean up and besides maybe some isolated delays Tuesday will likely be a normal day. Monday is a different story, and it all comes down to exactly what happens. If the main storm gets suppressed out to sea, then we just see light snow overnight Sunday night and schools may be able to open on delayed openings on Monday. The more likely scenario, with steadier snow during Monday morning and up to 6 or maybe 8 inches in areas would result in widespread school closures across Southwestern Connecticut, as the heaviest snow would likely fall during the morning rush under a regular or a delayed opening. If the ECMWF weather model or the majority of models showing a respectable hit from the storm are correct, then widespread school closures are likely on Monday. Latest data shows that this is the one day this week with the best chance of school closures, though a small chance remains for Friday too as another storm system may move through, so make sure to stay tuned as there are no guarantees with regards to school on Monday.
TECHNICAL: I wanted to briefly address a few technical points about this storm. First, I think the 18z GFS is wrong. About a third of its ensemble members support the suppression, but the GFS is likely under-doing the amount of precipitation if the storm is suppressed. I think the GFS overcorrected the polar vortex placement by moving it too far east thus suppressing the storm to the south, and models like the ECMWF seem to be handling this better. Yes, this is a progressive pattern, but when you have models as amplified as the NAM/SREFs and even the CMC I think the GFS is being unreasonable progressive, as is one of its major biases. I wish it was easier to display 500mb maps on here to display my point, but you can look for yourself on the Penn State E-wall (linked in the glossary). As for what models I am favoring, I don’t typically like to peg myself to any one model. The RGEM is out of range, but something tells me it will handle this storm well, as it does tend to over-amplify in a progressive pattern, but when you correct for that bias it handles the overall pattern well. So far, the RGEM looks VERY little like the GFS, which was a red flag. Another red flag is that the 12z NAVGEM weather model, normally one of the most progressive and the one that hypothetically should show the most suppression across SWCT actually had some of the most precipitation and was certainly more amplified than the GFS. Granted, this is no guarantee that it is right, in fact I doubt it is as the model has a horrible track record. This is a sign that the GFS is likely too progressive. As I have said above, the energy for the storm begins to move onshore tomorrow, and I expect models to begin to converge at least a little then when we get good sampling. I have heard reports that the National Weather Service is flying planes into the disturbance this evening too to get additional data, so overnight models will be important. I’ll be tracking the storm throughout the day tomorrow, but constant updates likely won’t be available until Sunday on Twitter. But each model run becomes progressively more and more important as some models show up to 8 inches here and some show as little as 2-3. I doubt we see widespread amounts higher than 6-8 inches unless something drastically changes, but we could see that suppression come into play. As I’ve said, I still don’t buy it, I think the GFS is under-doing the strength of the energy moving onshore aloft as it clearly strings it out more than other models have. Perhaps more interestingly, the high pressure center over the region is not quite as strong as expected per latest RAP guidance, which could result in slightly less suppression. That high pressure is getting kicked out either way, so likely not huge forecast implications, but it is important to watch these little trends in surface observations and short range weather models that can sometimes give hints about what the storm will do. I wish I had more time to go more in-depth about why the suppression is likely overdone and maybe show you, but as it is this post is already over 2,000 words and I am pressed for time so I’ll be wrapping it up.
CONCLUSION: Overall, I expect to see a light mix of rain/snow with maybe sleet/freezing rain mixing in Sunday afternoon and evening before a transition to all snow overnight Sunday into Monday morning. The heaviest snow is expected Monday late morning before snow winds down Monday afternoon and ends by evening. There is a chance that the entire storm ends as snow if the more suppressed models are correct, but I think there is a chance we see at least a little mixing across the area. General amounts are expected to be in the 4-8 inch range, and schools may need to close on Monday, though no large Tuesday impacts are expected. That’s generally it from here, I may try and post another blog earlier tomorrow (before 1 PM) that breaks down the overnight models and any forecast changes, but no guarantees. This is another very difficult forecast that I have very low confidence in, but that means that constant updates are all the more important so I will be doing my best to keep you ahead of the storm and as informed as you can be about what is changing in this very volatile forecast. Make sure to keep it here for the most local weather updates across Southwestern Connecticut!