Thursday, January 12, 2012

An update on the snow forecast for the upper midwest tonight

As I talked about in yesterday's blog, the first significant snowfall event of the season is underway across much of the upper midwest.  The areas around Lake Michigan look to be the hardest hit with this storm in terms of snowfall amounts today.  Here's the latest regional radar image:

Lots of precipitation falling throughout the area, and it's pretty much all snow once you get west of central Indiana.  Notice the banding of the precipitation coming off of the eastern side of Lake Michigan and into western Michigan--that's evidence of strong lake effect enhancement going on there.  However, as we'll see in a moment, it's a bit of a tricky situation going on there...

Here are some of the snowfall totals as of around 12:15 this afternoon (Central Standard Time) (from the WGN weather center).

Lincolnshire 1.0 inches
Crystal Lake 1.2 inches
Woodstock 1.0 inches
Romeoville 1.8 inches
Downers Grove 2.1 inches
Machesney Park 2.4 inches
Montgomery 1.5 inches
Yorkville 1.2 inches
Rockford 2.0 inches
O'Hare 0.9 inches
Oak Brook 1.4 inches
Midway Airport 1.4 inches
Batavia 1.5 inches
Poplar Grove 2.0 inches
Aurora 0.5 inches
Arlington Heights 1.0 inches

I supplement this with the reports from the Milwaukee forecast office for southern Wisconsin:
Milwaukee  0.9 inches
Beaver Dam 2.1 inches
Madison  2.1 inches
Oregon 1.5 inches
DeForest  1.8 inches

The big winners so far seem to be points west of the city, including my home territory of the Rockford area and its suburbs which are approaching 2-2.5 inches so far.  I have seen a few scattered 4+ inch reports coming in from eastern Iowa, but I haven't found a compiled list from the Davenport forecast office yet.
** Just got a list from the Davenport office of a few observations there***
Dubuque  3.3 inches
Cedar Rapids  4.8 inches
Stockton  3.5 inches
 I'm also waiting to see what the Grand Rapids office reports about lake effect snow amounts on the Michigan side.

So how much more snow are we going to get in these areas?  Often we turn to high-resolution, short-range models to try and figure that out. One of the newest of these models is something called the High Resolution Rapid Refresh, or HRRR. This model was initialized at 16Z this morning (10AM CST) and forecasts out for the next several hours.  Here's the HRR forecast for total snowfall accumulation from 10AM through 3PM this afternoon:
You can see that the HRRR is highlighting a band from the south side of Chicago back through north central and into northwestern Illinois as the area with the most snow (disregarding the snow related to the cyclone further east over eastern Ontario).  A few isolated spots in central Michigan and central Wisconsin are also highlighted.  One thing I initially found peculiar about this image is that it's forecasting little to no snow accumulation over the Grand Rapids area in southwestern Michigan.  Surely one would assume with that lake-effect-style banding coming off the lake that they'd be getting buried with snow.  But, if you check the latest surface map...
You'll see that in that area of southwestern Michigan the temperatures are above freezing-- 37 to 40 degrees --  with rain being reported.  An onshore component to the wind, combined with the fact that that area is somewhere very close to the surface low pressure center is helping keep the low-level temperatures there warm enough to support rain.  So, what started out as my mild disbelief of the HRRR model turns into a mild respect for its ability to pick out that particular phenomenon.

Just to keep following up on this, it turns out that the Grand Rapids radar is indeed upgraded to dual-polarization.  We can check the correlation coefficient product from that radar to get an idea of whether or not what's hitting the ground is rain, snow, or a rain/snow mix.  Here's the latest correlation coefficient image.
The radar is right in the center of the image.  Anywhere you see white on the image it indicates that the radar returns were highly correlated--this means that what the radar is seeing in those areas is all rain or all snow.  Since the radar beam gets higher above the ground the further it moves from the radar, the large areas of white at a distance from the radar are also pretty high off the ground.  This probably means we're looking at all snow up there (since it gets colder aloft).  But, right near the radar we see a transition to the pinks and oranges.  Those indicate lower correlations, which points to a mix of rain and snow.  This means that as we're getting closer to the ground (closer to the radar) the snow falling from above is melting--becoming a mixture.  And right near the radar there's a small area where we're all white again--so, that could mean right at the ground we've suddenly switched to all rain  (if you believe the signal that close to the radar, but those surface observations seem to confirm this).  It's very close to being snow, but it's just warm enough near the surface to melt that snow into rain.  This illustrates one of the great benefits of having a dual-pol radar--we can use it to monitor the transitions between rain and snow.  The more that pink and orange melting layer shrinks in closer to the radar, the lower the snow level is becoming.

So now that we've verified that the HRRR is actually being somewhat accurate, let's press a little further with our forecast.  Here's the forecast snowfall accumulations between 10AM and 6 PM CST tonight (including the evening commute...):

That area from the south side of Chicago westward still looks to be getting the brunt of things.  In the previous map they had accumulations of about 2-3 inches, and hear they're starting to get into the 3-6 range. So that's saying that another 1-4 inches of snow are forecast to fall between 3 and 6 PM local time.  Also notice that the model does start to show some accumulations in that area of southwestern Michigan--the model is forecasting it to cool down enough for that snow to finally make it all the way to the surface in the Grand Rapids area.

Another complicating factor will be the winds.  Here's the HRRR forecast for the winds at 4 PM CST tonight:

This is showing 20 knot winds or so around the Chicago area and points west.  The color coding indicates the wind speed.  Notice the winds are stronger out over the lakes where this is less friction.  They also weaken significantly over the bulk of Michigan as that area is pretty much right underneath the low pressure center.

With lots of snow falling, strong winds and darkening skies, visibility is going to be an issue for the commute home tonight across northern Illinois.  Take all necessary winter driving precautions and take it slow.

If you'd like to look at graphics from the HRRR yourself, they can be found at:
(Note: if there don't appear to be a lot of images available, you can change the initialization time to the previous hour from the drop down box at the top of the page.  You can also use the region drop down box up there to select a different region to zoom the map in.  I used the Great Lakes region for the maps I showed here.)

1 comment:

  1. Where did you find the Dual-Pol image like the one you link too? thanks!