2019-11-11 Veteran’s Day Overachieving Snowstorm (SE Michigan)

Aug 18, 2018
Novi, MI
This event unfolded mainly because of intense 850-700 mb frontogenesis along a pseudo-stationary front that had stalled behind an intense arctic cold front. In addition, low pressure began spinning up across southern Indiana/Ohio during the day which aided in strengthening the fgen.

Light snow started to break out across southern Lower Michigan during the late evening of November 10. By 4:00 A.M. on November 11, snow had become moderate at times and was finally really sticking to the grass. It was probably accumulating around 0.5-0.7’ per hour, respectable for early November standards.

All my classmates had been asking me the night before if we were going to have a snow day, and I told them that we would definitely have one based on my forecast (it usually takes 6+ inches of snow for superintendents to really consider a snowday in the Detroit suburbs). But I was worried the superintendents would only look at the local news, which never seems to be very accurate with the weather here. I was very disappointed to wake seeing that NOT ONE SCHOOL DISTRICT in the entire area had closed for the day (which is very rare when there’s an advisory out). I went to school, sat in my precalc class, and looked out the window seeing the snow get heavier and heavier. I knew many superintendents were going to regret their decision. A lot of my friends were frustrated, along with me, that we were going to have to drive home in the height of the event, but I told them just to pray and they’d be fine ;D

Back to meteorology...By late morning an intense band of snow had materialized across SE Michigan with light to moderate snow across the rest of the southern half of the state. It was centered about halfway between Flint and Ann Arbor. Snowfall rates reached at least an inch per hour, possibly nearing 2 inches per hour in the center of the band. At the same time, bands of intense lake-effect snow were being observed in the Thumb coming off of Lake Huron producing snowfall rates of 2+ inches per hour.

The main band began weakening a bit by late afternoon, but snowfall rates still around an inch per hour were common. The forcing continued into southern Ontario, and the widespread snow ended in most areas by 8:00 P.M. Some weak lake-effect snow showers moved in from the north and dropped an additional inch of snow overnight.

By the time it was all over, most areas in SE Michigan had seen 8-12 inches of snow with a couple reports in Thumb up around 14 inches because of the lake-enhancement. I’m convinced there must’ve been some areas in eastern Huron county (tip of the Thumb) that saw 18+ inches because the most intense lake-effect band sat over there for the entire afternoon and much of the evening. No reports came out of that area.

How well was this storm’s magnitude forecasted? The answer: Horribly.

The day before the event the local news was only saying 2-4 inches, and the NWS had a winter weather advisory for 3-5 inches. I had been anticipating the storm for several days: My first forecast 3 days out was 3-6 inches, but I kept increasing amounts slightly each day. The night before the storm, I forecasted 5-8, and that next morning I raised it to 6-10. I was astonished that the NWS had only increased to 5-7 inches overnight. A winter storm warning was more than justified.

I kept increasing snow amounts in tandem with the synoptic and high-res models. At 7:00 AM. the morning of the storm, I checked all reliable models and not ONE of them had anything below 6 inches with a 10:1 SLR. I was thinking this storm would generally have a 12-14:1 SLR, so I thought 8 inches would be a good medium, hence my forecast of 6-10 inches.

Also, I know the Kuchera ratio is available for some models, but I don’t understand what it means, so I resorted to mental math with the 10:1 SLR instead.

I have to applaud the NWS for really picking up on the lake-enhancement in the eastern Thumb. Days out they knew this would occur and called for localized amounts up to 20 inches there.

Everybody’s forecast ended up having a pretty good percent error (very high in some cases) as the minimum of actual snowfall along I-96 (central part of SE MI, if that makes sense) ended up being 8-9 inches with a max of around a foot (13 inches was reported around Jackson I believe).

Overall I think this storm serves a good example of how fgen snow events can really overachieve. This is almost always due to mesoscale banding within the snow shield itself, which is exactly what happened here.

Any thoughts on the storm or this summary?

Dan Robinson

Staff member
Jan 14, 2011
St. Louis
I wondering if the thermal profile up there was as colder than forecast as ours was here (STL area). The Arctic air advancement was about 2 hours ahead of schedule here, with the freezing line arrival and ptype changeover complete by 10AM instead of noon-1pm that all models had been showing even in the hours before. I didn't look at upper air soundings for this event at all in preparation for it, but the assumption was that we'd be below freezing through the column with the low levels the last to change as the cold advection continued. The impact for that here was that we had 3 hours more of below-freezing snowfall than expected, with the last few hours of moderate precip falling at 26-28F instead of 30-31F. That resulted in significant road icing with the DOT crews caught off guard. We reached a little over 2" in spots, which was in line with short term guidance, but the road impacts were much more severe than I had been anticipating thanks to the faster arrival of cold air in the low levels.
Aug 18, 2018
Novi, MI
I am not 100% sure but I’m fairly certain we were colder than expected. I think the low pressure that formed along the arctic front strengthened more quickly than anticipated and increased CAA earlier consequently. We were forecasted to get to a high of around 30 in the late morning but we never made it out of the mid 20s (at least where I was). This definitely increased SLRs a bit.

Also the QPF from the models ended up on the low side as most areas near me ended with around 0.8-0.9’ liquid equivalent, but the models had only been advertising up around 0.7’ at most.