My original account of a warm snow event :

On 9 October 1994 (I subsequently discovered this occurred 10 October) at about 1330 EST, I saw a snowshower at a temperature of 54 °F in Dearborn Heights, MI (near Detroit, and about 10 mi NNE of DTW). The temperature is based on my reading of a Taylor mercury thermometer (shadowed by our garage) immediately after the snowshower. I calibrate the thermometer to 32 °F in an ice bath periodically. Below are a description of the event and a brief discussion regarding favorable conditions for this phenomenon.

On the evening of the 8th (9th actually), a cold air mass moved into our region, and the next morning our sky was mostly clear. Solar energy heated up the atmospheric boundary layer (ABL), and cumulus clouds began to form in late morning. They continued to form thruout early afternoon, and their sky coverage averaged about .4 from 1210-1530 EST. At about 1330 EST, my mother and father noticed snow, and called to me so that I could see it also. It lasted about 1-2 min, and I saw its final 20 sec or so. The air was relatively calm during the snow, and we saw no significant downdraft. The snowflakes had a silvery appearance (probably covered with water), although they were definitely "plates". Some also looked more like sawdust, and appeared to be evaporating (and subliming) very quickly. The cumulus cloud from which they fell covered about half of the sky, and seemed to have a typical base of about 5000 feet. Immediately after the snowshower I read our thermometer, which indicated 54 °F. If the ground became wet, I did not notice it. A difference between this event and other times I have experienced snowshowers with temperatures up to 48 °F is that in every other case, there seemed to be a certain "chill" in the air - in this case, the air felt warm.

2 considerations regarding the observed temperature are : 1) it is likely to have decreased slightly during the snowshower, and 2) the mercury thermometer would have responded very little to any temperature change that may have occurred. The high temperature reached 59 °F earlier in the day, and was 49 °F by about 1400 EST. The proximity of the Great Lakes and perhaps large scale subsidence also contributed to warming the ABL. I do not measure humidity, but the air did not seem excessively dry. One can check observations at DTW, but I seem to recall their dew point being slightly above freezing that afternoon.

Much discussion regarding this phenomenon has focused on dry air. Evaporation and sublimation cause snowflakes to diminish in size, but also result in cooling, allowing them to survive in a 'hostile' environment well above freezing. It was suggested to me that air can never be so dry that size reduction overcomes cooling. Snowflake size, shape, and composition also significantly affect this phenomenon, perhaps even more so; mainly regarding breakup, melting and cooling, and fall speed. I state the following as favorable conditions for snow at high temperatures :

  • High temperature lapse rate, due to cold air aloft and various factors warming air near the ground

  • Cloud processes which create snowflakes that can survive the plunge into a 'hostile', warm environment; and perhaps a cloud base at or near freezing

  • Dry air below cloud base, to cool melting snowflakes

  • Partly cloudy cumulus skies, such that solar energy significantly heats the ABL

  • High elevation, since for a given amount of evaporation/sublimation, cooling of air is inversely proportional to its density (the highest possible temperature with a wet-bulb temperature of 32 °F is about 49.2 °F at a surface pressure of 980 mb, and about 53.2 °F at 800 mb)

    • Perhaps someone into weather modification can look for such conditions and go to it ! Maybe someone already did


    Text is copyright of Joseph Bartlo, though may be used with proper crediting.

    Home Page