Current Icing Potential (CIP) and the CIP sounding
technique
Meteorologists at the National Center for Atmospheric Research developed a
multiple data source, hybrid approach to the diagnosis of icing, the “Current
Icing Potential” (CIP; Bernstein et al 2004). In March 2002, CIP became an
official Federal Aviation Administration and National Weather Service product.
The latest version of CIP combines satellite, radar, surface and lightning
observations with numerical model output and pilot reports of icing to create
an hourly, three-dimensional diagnosis of the potential for icing, and
SLD. CIP uses observations to diagnose the locations of clouds and
precipitation, and then combines them with numerical model output in a
situationally-based, fuzzy logic system. The physical structure of the
atmosphere is assessed for each vertical column and an icing scenario is
identified using a decision tree. Data are treated differently for each
scenario, and fuzzy logic “membership functions” are applied to fields such as
temperature and relative humidity in an appropriate manner. The algorithm
determines icing and SLD potentials for each level in the sounding. The
potential is essentially a confidence or likelihood that such conditions were
present. Their presence becomes increasingly more (less) likely at the
higher (lower) thresholds, and the algorithm is more (less) efficient at
capturing pilot reports while warning for a relatively small (large) volume of
airspace in this range. A more complete description of CIP, its components and
application of fuzzy logic is found in Bernstein et al (2004).
A special version of CIP was developed that diagnoses icing and SLD potentials
in a column using the vertical profile of temperature and moisture from a
balloon-borne sounding in combination with coincident, nearby surface
observations. The CIP sounding technique essentially mirrors that of the
real time CIP grids. The technique is quite complex, so only the aspects
that are unique to the CIP sounding system are described here. In short,
the cloud top height and temperature are determined using the moisture profile
in the sounding, rather than satellite data, while the occurrence of
precipitation and precipitation type are determined solely from surface observations,
rather than both surface observations and radar reflectivity. Pilot
reports of icing and forecasts of vertical velocity and explicit supercooled
liquid water that are used in the real time CIP grids are not used in the
sounding CIP. Stricter versions of the relative humidity and cloud top
temperature membership functions were applied because the improved quality of
moisture data in soundings (as compared to numerical weather model forecasts)
provided more accurate estimates of cloud locations and cloud top temperatures.
Determining Cloud Presence and Candidate
Altitudes
The first step in finding icing and SLD is to see if clouds and precipitation are
present. If the surface observations indicate that the sounding ascended
through a “cloud free” environment, then the icing and SLD potentials are set
to zero throughout the column. If clouds are present, then the altitudes
that are likely to contain clouds and precipitation are identified and all
other altitudes are considered to be free of icing and SLD. The lowest
altitude considered for icing and SLD is set to ground level when any form of
liquid or freezing precipitation was present, since it is possible for icing
and SLD to exist from cloud base down to the ground (see example soundings in
Fig. 1).
Cloud top height is estimated by searching downward through the profile until
relative humidity (RH) with respect to water or ice either A) meets or exceeds
87% or B) meets or exceeds 84% and the RH of the level above is at least 3% lower
(Wang and Rossow, 1995). The presence of multiple
cloud layers is inferred if the dry layer beneath a cloud deck appears to be
adequate for complete sublimation or evaporation of any particles falling from
it. If adequate dry air is present and an additional layer meeting the RH
thresholds is found beneath the dry layer, then a new cloud deck is identified
(see example soundings in Fig. 1). The cloud top temperature (CTT) of
each layer is set to the temperature where the RH threshold is first met.
All altitudes above the highest cloud top are indicated as free of icing and
SLD. SLD potential is only determined for the lowest cloud layer and altitudes
between it and the surface, since only this layer is likely to have formed the
precipitation reported at the surface.
Once cloudy and/or precipitating altitudes are found, the sounding structure
and precipitation are examined to determine the meteorological situation that
is present. This situational approach is critical, since the meaning of an
individual piece of data can be very different for different situations.
Fig. 1a - Example skew-T plot for sounding
taken at
Fig. 1b - Example skew-T plot for sounding
taken at
Icing and SLD Scenarios
While icing forms via several mechanisms, two primary mechanisms are
responsible for the formation of SLD: classical and non-classical. By
determining the mechanism and examining the temperature and moisture profiles
as well as the observed precipitation type, the “sounding CIP” can be used to
diagnose the SLD potential. The classical and non-classical mechanisms are
described below because of their criticality for SLD. Alternative icing
mechanisms are described in Bernstein et al (2004). They include single-
and multi-layer non-precipitating clouds, and deep convection.
Classical
“Classical” icing and SLD occurs when a layer of T>0oC
("warm nose") is located between two layers with T<0oC,
surface freezing or liquid precipitation is observed and the CTT (cloud top
temperature) is less than -12oC (Fig. 2a). The relatively cold CTTs
indicated that an ice process is likely to be active above the warm nose. Snow falling
into the warm nose melts to form liquid precipitation, which subsequently falls
into the lower subfreezing layer to form classical SLD, usually in the form of
FZRA. The precipitation typically reaches the surface in the form of
FZRA, PL or RA, depending on the strength of the warm nose and the T and RH
within and beneath the lower subfreezing layer (Hanesiak
and Stewart 1995). It occasionally reaches the surface in the form of FZDZ or
DZ when very light snow starts the process above the melting zone or the
process is entirely non-classical, but this is uncommon.
Non-classical
“Non-classical” (collision-coalescence) icing and SLD occurs in one of
three situations (Fig. 2b). A) Freezing or liquid precipitation is observed
when a classical warm nose is present, but CTT is greater than -12oC,
indicating a good chance that an all-liquid process is responsible for the
precipitation formation. B) No warm nose is present, only RA and/or DZ are
observed at the surface, and CTT is greater than -12oC. C) No warm
nose is present and freezing precipitation (FZDZ,
Fig.
2. Examples
of a) classical and b) non-classical SLD. The “warm nose” is marked in (A). SLDPOT
profiles (gray lines) are given in the right-hand
panels, with the SLD layer tops and bases indicated. Grey shaded areas
indicate the altitudes with clouds.
References
Bernstein, B.C., F.M. McDonough, M.K. Politovich, B.G. Brown, T.P. Ratvasky,
D.R. Miller and C.A. Wolff, 2004: The Current Icing Potential (CIP). Conditionally accepted to J. Appl.
Meteor.
Hanesiak, J.M. and R.E. Stewart, 1995: The mesoscale
and microscale structure of a severe ice pellet storm. Mon. Wea.
Rev., 123, 3144-3162.
Wang, J.W. and W.B. Rossow, 1995: Determination of
cloud vertical structure from upper-air observations. J. Appl.
Meteor., 34, 2243-2258.