cooling, there is a tendency for the initial condensation
to contain a higher proportion of water droplets, which
leads to a mixed cloud that will convert to ice or snow
in time.
Presumably, dense cirrus, fine cirrus,
cirrocumulus, and anvil cirrus clouds are of this type. It
is assumed that fine cirrus clouds (proper) are formed in
shallow layers that are undergoing rapid convection due
to advection of colder air at the top of the shear layer.
On the other hand fine cirrus and cirrostratus clouds
are so often associated, and cirrostratus clouds are so
often reported by pilots as developing from the merging
of fine cirrus clouds, that there is a question whether the
process of formation in cirrus and cirrostratus clouds are
essentially different.
Nevertheless, the prevailing
crystal types in cirrus and cirrostratus clouds seem to
differ, though this may not be universal, or may merely
represent different stages in cirrus cloud evolution.
Horizontal risibilities within cirrostratus clouds are
generally between 500 feet and 2 nm. Thin cirrus haze,
invisible from the ground, often reduces the visibility to
3 nm.
A rule of thumb for forecasting or estimating the
visibility within thin cirrus or other high cloud
(temperatures below 30°C) follows:
Visibility = 1/2 nm times dewpoint depression in
degrees C. For example; temperature is 35°C,
dewpoint is 38°C, and visibility = 1/2 x 3 = 1 1/2 nm.
This rule has been used successfully only in the Arctic
where poor visibility in apparently cloudfree air is often
encountered.
THE CIRRUS CLOUD FORECASTING
PROBLEM
Many forecasters have attempted to forecast cirrus
clouds by using frontal or cyclone models. This
procedure is not always satisfactory. There are a
number of parameters, both surface and aloft, that have
been correlated with cirrus cloud formation. A few of
the more prominent parameters are mentioned in the
following text.
Surface Frontal Systems
Frontal and cyclone models have been developed
that embody an idealized cloud distribution. In these
models, the cirrus clouds are lowering and thickening to
form altostratus clouds, which indicates an advancing
warm front.
Fronts Aloft
Above 500-hPa the concepts of air masses and
fronts have little application. Most of the fine cirrus
clouds observed ahead of and above warm fronts or lows
initially form independent of the frontal middle cloud
shield, though later it may trail downward to join the
altocumulus and altostratus cloud shields. With
precipitation occurring in advance of a warm front, a
60-percent probability exists that cirrus clouds are
occurring above. Cirrus clouds observed with the cold
front cloud shield either originate from cumulonimbus
along and behind the front or from convergence in the
vicinity of the upper trough. In many cases there is no
post cold front cirrus clouds, probably due to marked
subsidence aloft.
Contour Patterns Aloft
One forecasting rule used widely states that the
ridge line at 20,000 feet, about 500 hpa, preceding a
warm front marks the forward edge of the cirrus cloud
shield.
For a typical 500-hPa wave pattern, the following
information applies:
. No extensive cirrostratus clouds will occur
before the surface ridge line arrives.
l Extensive cirrostratus clouds follow the passage
of the surface ridge line.
. No middle clouds appear before the arrival of the
500-hPa ridge line.
. Middle clouds tend to obscure the cirrus clouds
after passage of the 500-hPa ridge line.
. When the 500-hPa wave has a small amplitude,
the cirrus cloud arrival is delayed and the clouds are
thinner.
. The greater the 500-hPa convergence from
trough to ridge, the more cirrus clouds between the
surface and 500-hPa ridge lines.
Cirrus Clouds in Relation to the Tropopause
Experiences of pilots have confirmed that the tops
of most cirrus clouds are at or below the tropopause. In
the midlatitudes, the tops of most cirrus cloud layers are
at or within several thousand feet of the polar
tropopause. Patchy cirrus clouds are found between the
polar tropopause and the tropical tropopause. A small
percentage of cirrus clouds, and sometimes extensive
4-20