I would like to know how to determine whether or not frost
will appear on the ground. I assume you need dew to get frost, and dew is
dependent on the dew point, which is easy to find information about (e.g.
http://paroscientific.com/dewpoint.htm). Basically then, my question is
how can you estimate ground temperature given air temperature or possibly
air temperature history? Does heat transfer between the ground and air
play a significant role?
Let me answer just a part of your question. Your assumption is incorrect.
Frost is not made by freezing dew. It is formed by sublimation - water vapor
moving directly to a solid... ice. Frost is formed when the dew point is near
or below freezing and the temperature of the air falls to within a few degrees
of the DP.
Your thought about ground temperature is interesting. Generally the air is
warmed or cooled by the surface. The air is largely transparent to radiant
energy so heat released by the surface at night is radiated off into space.
But as the surface cools, the heat from the lowest layer of air is
the ground. Cooling the air from below tends to make the air more stable and
keep the cooler air low.
Hope that helps.
I published an article several years ago in the Journal of Meteorology, U.K.
on how to estimate frost from previous night temperatures. It is called
"Predicting Frost At Your House". You don't usually
get dew before frost, although you can on nights when the temperature starts well
above freezing and moves to below freezing during the night. In the colder
months frost develops as the temperature decreases to below the frost point,
which is essentially equivalent to dew point but at temperatures below freezing.
On nights with heavy frost, it is not uncommon to have an ice fog, in which
supercooled water vapor forms ice crystals in the air and hoarfrost on exposed
Heat transfer between the ground and air is important, as the ground
supplies energy to heat the air near the ground, unless of course it has been
very cold for a long time and the ground is frozen (or there is snow on it).
Much of the water vapor that goes into dew and frost actually comes from
from the soil and transpiration from plants and grass (or snow sublimation);
evaporation continues at night, although to a much lesser degree than
during the day.
A very dry soil and surface vegetation will produce much less or no dew or
frost on the vegetation.
If you are interested in the article, I can send you a copy. Please contact
NEWTON again to pass along your address.
The citation is:
D. R. Cook, "Predicting Frost At Your House",
Journal of Meteorology, Vol. 15, no. 153, November 1990.
This is a popular journal published in Great Britain.
Argonne National Laboratory
Frost formation is a complex process, and conditions have to be "right"
for it to occur. Frost forms on surfaces directly from the vapor state,
without condensing as dew. If dew forms, frost formation is unlikely, even
if the temperature drops below freezing.
Frost is more likely to form on surfaces above the ground first, such as
house roofs, or automobiles, because the air immediately above the ground is
usually a few degrees warmer than air a few feet higher. There is some heat
transfer from the ground to the air a few centimeters above it. If there is
much wind, frost will not form either. (Neither will dew, as both these
occurrances require little or no wind, so the atmosphere will not stay
mixed.) If the skies are cloudy, usually dew or frost will not form either,
as the clouds reflect the radiated heat from the ground, which helps in
keeping the lower layers mixed.
So the ideal conditions for frost formation is a night with clear skies,
light winds, and a temperature forecast to be near or a little below
freezing. Standard temperature measurements are taken from about 2 meters
above ground. On a calm night the ground temperature can be as much as 5-7
degrees cooler than the standard temperature reading. If there is some wind,
the air stays mixed, and the temperature difference disappears.
Wendell Bechtold, Meteorologist
Forecaster, National Weather Service
Weather Forecast Office, St. Louis, MO
Click here to return to the Weather Archives
Update: June 2012