Pet Peeves and Rear-End Collisions
by Lee Grenci

Of all the unscientific drivel that I sometimes hear on television weathercasts, none gets my blood boiling more than the maddening references to 32 degrees Fahrenheit as the freezing point of water (weathercasters typically refer to 32 degrees as 'freezing'). For the record, water can exist in the liquid phase more or less indefinitely at temperatures below 32 degrees Fahrenheit. For proof, check out the results (see photograph below) of an experiment I conducted in our refrigerator ... thanks to my wife for the temporary use of our freezer.

For the record, I placed nine drops of water on the bottom side of an empty can (after I rubbed a little oil on the can's bottom). Then I placed the experiment into the freezer for about ten minutes (at approximately 15 degrees Fahrenheit). When I removed the can, only five of the drops had frozen.

In the atmosphere, tiny cloud drops resist freezing down to temperatures as low as -40 degrees Fahrenheit. During episodes of freezing rain, falling raindrops must be 'supercooled', meaning that their temperature is lower than 32 degrees Fahrenheit (during times of freezing rain, the atmosphere makes a 'warm-air sandwich', with a layer of relatively warm air several thousand feet aloft sandwiched between thicker layers of cold air; in such an environment, snowflakes melt into raindrops, and the drops become supercooled as they fall through the layer of cold below the melting layer). Of all people, meteorologists should know that they are on firm scientific ground calling 32 degrees Fahrenheit the melting point of ice and not the freezing point of water. Assuming a typical air pressure, all ice melts at 32 degrees Fahrenheit (actually, an immeasurable very tiny bit above 32 degrees), but not all water freezes at 32 degrees Fahrenheit.

You see, for water to freeze, freezing nuclei must be present. Without getting into too much detail, freezing nuclei are tiny particles that initiate the growth of ice crystals (akin to condensation nuclei, which must be present for cloud drops to form).

I suspect that students are taught in physics (and, perhaps, chemistry too) that water invariably freezes at 32 degrees Fahrenheit. That's because physicists, in general, probably don't know much about the importance of freezing nuclei. In fairness, the reason for this apparent lack of interest is that freezing nuclei are ubiquitous. You have to work to get rid of them. And so there is often no need to worry about freezing nuclei. Nature provides them for free. Only in meteorology does the rubber really meet the road. Indeed, freezing nuclei are the only way to explain the irrefutable fact that clouds of water drops can exist at temperatures well below 32 degrees Fahrenheit.

Which leads me to another pet peeve. Television weathercasters sometimes lend the impression that surface temperatures must be 32 degrees Fahrenheit or lower in order for it to snow. That's hogwash ... I've seen it snow at temperatures as high as 44 degrees Fahrenheit. Perhaps this mistaken notion explains why some folks believe that it never snows in Florida.

On January 19, 1977, 0.2 inches of snow fell on Tampa (see the above historic photograph). Here's the pertinent weather data from Weather Source's Weather-Warehouse:

Tampa International Airport, Tampa, FL 33631

According to Andy Devanas, the Science Officer at the National Weather Service in Key West, Florida: 

"I was in Tampa Florida on January 19, 1977, and remember that day well. We got about a quarter inch of snow in the morning and pretty much gone by noon. There were hundreds of accidents across the city from people rear-ending the car in front of them while looking up at the falling snow."

Needless to say, it was a rare event. There were unconfirmed reports of snowflakes falling in Key Largo in the Florida Keys, which is about as far south as you can go in the United States.

Below is the historic weather map for 7 A.M. EST on January 19, 1977. At the time, a low-pressure system had departed the Southeast Seaboard, while a high-pressure system was centered along the Texas-Louisiana border. The low and high worked in tandem to usher low-level cold air southward over the Florida peninsula, essentially bypassing the relatively warm waters of the Gulf of Mexico (note the north-south orientation of the isobars and the northerly and north-northwesterly surface winds). Such surface patterns are a recipe for cold air to penetrate far south over the Florida peninsula. In winter, the winds are typically more westerly in the wake of low-pressure systems and their associated cold fronts that form over the Deep South, thereby allowing Gulf waters to modify cold air before reaching the Florida peninsula.

Keep in mind, however, that temperatures aloft largely govern the type of precipitation observed near the ground. At the time, 850-mb temperatures had plummeted to approximately -10 degrees Celsius (in this case, the 850 mb altitude over Tampa was approximately 1425 meters). The vertical profile of temperatures and dew points from the radiosonde released at 12Z from Tampa (below) shows a precipitating layer roughly between 850 mb and 700 mb (where the relative humidity is 100%).  

The sounding looks convective to me (superadiabatic lapse rate near the ground and nearly dry adiabatic up to about 900 mb), so I envision precipitating cumulus over Tampa at the time. Moreover, this sounding indicates that the precipitating clouds could have been sufficiently tall to be cumulus congestus, but there's no way for me to determine, just by looking at sounding data, whether the radiosonde rose through a cumulus congestus cloud or precipitating stratocumulus. Indeed, we would need to check satellite imagery and/or the remarks on the surface METARS from all of the observation sites in the Tampa Bay area (anybody out there game?).

The bottom line here is that the Tampa temperature sounding lies completely to the left of the skewed zero-degree isotherm. It's a snow sounding, no doubt about it.

If you look again at the isobars over Florida on the 12Z surface analysis, you'll note that they are anticyclonically curved, meaning that the isobars have a curvature consistent with the approaching high (the surface low and its cyclonically curved isobars had exited Florida by this time). At 700 mb (an altitude roughly near 3000 meters or, approximately, 10000 feet), a trough in the 700-mb heights was nearly overhead, accounting for the cold air aloft that helped to destabilize the lower troposphere and set the stage for lingering precipitation over the southern half of Florida.

When I zoomed into the area of precipitation on the 12Z surface analysis, I noticed that the observation at Ft. Myers was 37 degrees with freezing rain. Hmmmm .... And the observer at Key West reported 57 degrees, with a snow shower having ended during the previous hour.

I have serious doubts about the snow shower reported at Key West. To my knowledge, it has never snowed at Key West since official weather records began there (confirmed by Andy Devanas). Moreover, it's physically impossible for it to snow at 57 degrees (or any temperature close to this reading). Assuming that the lapse rate above the ground is dry adiabatic (for all practical purposes, the fastest rate of cooling with increasing altitude), the melting level would have been at an altitude near 10,000 feet. That's way too high for a falling snowflake to survive to the ground without melting. Not even close. Maybe it was just wishful thinking on the part of the observer?

I also have serious doubts about the freezing rain at Fort Myers. Perhaps the unsaturated layer above the ground on the Tampa sounding was also present at Fort Myers. Such an unsaturated layer might have promoted enough evaporational cooling to allow some frozen precipitation to reach the ground.  If true, such a sight would have been very, very rare at Fort Myers, so I'm willing to cut the observer a little slack.