Transcript OU Seminar

RFC/HPC Hydromet Course 02-1:
Precipitation Type
Rain/Snow Lines
John Cortinas, Jr., Ph.D.
[email protected]
University of Oklahoma-NOAA Cooperative Institute for
Mesoscale Meteorological Studies
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Forecasting Precipitation Type
QPF: A forecast of the location and
the liquid equivalent amount of
precipitation during a given period.
Since the phase of the precipitation
affects the amount of liquid water
that reaches the ground, these
forecasts must consider
precipitation type during the cold
season.
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Forecasting Rain/Snow Lines and
Other Precipitation Types
* Have snowflakes formed aloft?
* Will the snowflakes melt
before reaching the ground?
* Will the snowflakes sublimate
before reaching the ground?
* Are the snowflakes going to
melt totally or partially?
* Will frozen precipitation fall
through layers of supercooled
water?
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Outline
* Brief review of precipitation microphysics associated with
producing:
 Snow (ice)
 Freezing rain (supercooled droplets)
 Ice pellets
* Review forecasting techniques for forecasting precipitation
type.
* Explain representation of precipitation in numerical models.
* Introduce SPC products that may be helpful in forecasting
precipitation type.
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Factors that Determine Precipitation
Type at the Ground
* Ground temperature (melting at the ground?)
* Precipitation microphysics (snowflake/ice crystal
characteristics)
* Thermodynamic stratification (temperature and moisture
vertical profiles)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Why is an understanding of
precipitation microphysics important
for QPFs?
* Precipitation microphysics determine precipitation type!
* This knowledge reduces forecaster reliance on “rules of thumb” that
may be inaccurate.
* Anticipate short-term changes in precipitation type and intensity.
* Understand shortcomings of precipitation-type numerical guidance.
* Knowledge of precipitation microphysics helps forecasters interpret
numerical model guidance intelligently (apply adjustments to model
forecast).
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Ice Physics - Basic Concepts
* ice nuclei - microscopic particles that serve as nuclei for ice
crystal formation
* ice (snow) crystal - a small ice particle that results from
deposition of water vapor onto an ice nucleus
* snow crystal habit - the shape of an ice crystal
* snowflake - an aggregate of ice crystals
* snowflake ice density - ice density of snowflake (bulk ice = 0.9
g/cm3)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Ice Nucleation
* Homogeneous nucleation - the spontaneous initiation of an ice
crystal caused by random collisions of water vapor
 theoretically occurs when temperature is less than -40˚C and RH is
several hundred percent!
* Heterogeneous nucleation - the spontaneous formation of ice
upon ice nuclei
 requires slightly supersaturated (with respect to ice) conditions and
temperatures less than 0˚C, usually less than -10˚C
 Three modes of nucleation:
» Contact
» Deposition
» Freezing
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Ice Nuclei
Most ice nuclei are clay particles
that originate in desert and arid
regions.
The concentration of active ice
nuclei is temperature dependent.
A very small concentration of
active freezing nuclei can exist
in air with a temperature as
high as -5˚C (compared to ~1 X 108 L-1
for cloud condensation nuclei.)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Heterogeneous Nucleation
How does an ice crystal form?
* Contact mode: ice nuclei initiates the ice phase at the moment
of contact with a supercooled droplet (riming)
* Deposition mode: water vapor diffuses directly to ice nuclei
* Freezing mode: ice phase is initiated from within a
supercooled water droplet because of the presence of ice
nuclei within the droplet
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Snow Crystal Habits
bullet
simple plate
solid column
combination
of needles
dendrite
hollow column
crystal with broad
branches
combination
of bullets
sheath
(Pruppacher and Klett, p. 33)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Snow Crystal Habits
Habit is a function
of temperature and
humidity.
(Puppacher and Klett, p 32)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Ice Particle Multiplication Processes
* In the presence of air saturated with respect to ice, ice
multiplication processes increase the number of ice crystals
and ultimately create more or larger snowflakes.
* Two processes create additional ice nuclei from a single ice
crystal:
 mechanical fracturing
 ice splintering
 fragmentation of individual cloud drops during freezing (controversial)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Mechanical Fracturing
Occurs when fragile ice
crystals, such as
dendrites and other
plates, collide
with other crystals,
graupel particles, or
cloud drops.
The process requires
turbulence or
differential crystal fall
speeds.
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Ice Splintering
* Occurs during riming conditions
* Splintering is caused by collisions with relatively large cloud
droplets
* Depends on drop size distribution, supercooled water content,
velocity of the drops impacting the riming particle
* Pronounced maximum occurs at a temperature of -5 ˚C and
drop impact velocity of 2.5 m/s
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Snowflakes
Snowflakes are an
aggregation of ice
(snow) crystals.
Avg. diameter of component
crystals: (1) < 1.5 mm, (2) 1.5
to 2.5 mm, (3) 2.5 to 3.5 mm,
(4) > 3.5 mm
(Rodgers 1974)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Crystal Growth by Vapor Diffusion
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Growth Rate
Growth rate depends on temperature (T), pressure (p), and size (r) of ice crystal
P=constant
(Ryan et al. 1976)
Plot shows growth rate dependence on size and temperature
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Growth Rate
Growth rate is
temperature
and pressure
dependent.
(Rogers, p. 126)
Plot shows growth rate dependence on pressure and temperature
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Crystal Growth Among Supercooled
Water Droplets
Ice crystals grow at the
expense of supercooled
water droplets because the
saturation vapor pressure is
lower over ice than it is
over liquid water.
© Wadsworth Publishing Company 1998
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Saturation Vapor Pressure
Excess Saturation Vapor Pressure
0.3
0.25
0.2
Water - Ice (mb)
Difference between
saturation vapor pressure
over water and over ice
is greatest near -12˚C.
0.15
0.1
0.05
0
-40
-35
-30
-25
-20
-15
-10
-5
T (ÞC)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
0
Aggregation
Aggregation increases the
mass of a snowflake and is a
function of temperature,
maximizing near 0˚C because
of sticky dendrites.
Aggregates are composed
mostly of dendrites and
some thick plates.
Secondary maximum
occurs near -12˚C, the
temperature at which
the growth rate for
dendrites is largest.
(Rogers 1974)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Ice Crystal Decay
dM
 4rD(   r )
dt
sublimation - air cools
deposition - air warms
evaporation - air cools
condensation - air warms
melting - air cools
freezing - air warms
conduction - air cools or warms
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Sublimation
* Occurs when ice crystals or snowflakes fall through icesubsaturated environments
theoretical
observed
(1) 0.3 g/cm3
(2) 0.5 g/cm3
(3) 0.75 g/cm3
(4) 0.9 g/cm3
(A) column,
800X164 mm
(B) sphere, r=160 mm
SLE Radiosonde
Air craft
Comparison between survival distance of cirrus ice particles
(Hall and Pruppacher 1977)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Snowflake Melting
* The melting rate is a function of:




air temperature near the hydrometeor surface
relative humidity
size of hydrometeor
amount of liquid water present
* Temperature of air near snowflake surface is determined primarily
by latent heating and conduction.
* Begins when temperature at surface of snowflake is > 0˚C
 stage 1: small drops of tens of microns in diameter appear at the tips of crystal
branches
 stage 2: capillary forces and surface tension draws liquid water to center
 stage 3: small branches of interior melt
 stage 4: main ice frame collapses, pulls itself together into drop
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Melting Model
RH = 90%
Theoretical and observational studies show snowflakes can
descend 800 m below the melting level before
complete melting in a subsaturated layer.
air temperature
(I) Cooling from sublimation exceeds warming from
conduction, therefore snowflake temperature remains
at freezing and no melting occurs.
(II) Warming from conduction exceeds cooling from
sublimation and snowflakes begin to melt. Melting rate
is slow because of additional cooling from evaporation.
(III) Air is supersaturated with respect to ice and
deposition occurs. Heating from conduction and
deposition exceed the cooling from evaporation, and
the snowflake melts completely.
Vapor density
Dewpoint
temperature
(Matsuo and Sasyo 1981)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Melting Experiments (Numerical)
density = 0.005 g/cm3
density = 0.02 g/cm3
RH = 100%, no sublimation, some evap.
density = 0.04 g/cm3
RH = 90%
RH = 80%
(Matsuo and Sasyo 1981)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Melting Experiment (Laboratory)
(Mitra et. al 1990)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Dependence of Melting on Air
Temperature and Relative Humidity
(Matsuo and Sasyo 1981)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Cooling by Melting
* the latent heat used to
melt precipitation cools
the atmosphere
* significant and continuous
melting can cool an entire
vertical column below
freezing, causing the
melting level to descend to
the surface
(Stewart and McFarquhar 1987)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Forecasting Rain/Snow Lines
* Examine forecast soundings when possible:







wet-bulb profile
melting level
amount of precipitation expected
general type and size of snowflakes
presence of adiabatic cooling caused by upward motion
ground temperature
surface air temperature and relative humidity
* Given the existence of ice crystals, snow usually occurs when:
 the wet-bulb temperature throughout the entire troposphere is expected to be
equal to or less than 0 ˚C and
 the melting layer is within 0-800 m of the ground
» high melting levels require low humidity (< 70%) surface layers or large
snowflakes for snow to reach the surface
» melting level can descend to surface with significant precipitation (in the
absence of other thermal processes) : accum. Rain (in.) = [dT (˚C)*dP
(mb)]/500 (Kain and Goss 2000), e.g. (2 * 200)/500 = 0.8”
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Partial Thickness
* Thickness
dz=-(R/g p) [Tv dp]
500 mb
* Advantages:




only requires 4 data values
more accurate than 1000-500 thickness
easy to compute
some skill with rain/snow decisions
* Disadvantages:
 does not identify shallow warm layers
 does not work for high terrain
 most accurate when “tuned” for specific
regions (no national uses)
700 mb
850 mb
1000 mb
}
}
}
dz=-(R/g p) [Tv dp]
Heppner (1992)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Partial Thickness (Heppner 1992)
* Determine the thermal profile of the troposphere and its
static stability.
* Examine the 850-700-mb thickness. (thickness > 1550 m,
snow unlikely).
* Use care when evaluating 1000-500-mb thickness (snow can
occur with thickness < 5400 m).
* Always consider effect of diabatic processes (evaporation or
melting) on changing thermal profile.
* At 850 mb, 0 deg C isotherm doesn’t always work well to
delineate rain and snow, especially with unstable lapse rates.
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Freezing Rain
* Freezing rain occurs when snowflakes
melt completely in an elevated warm
layer and continue to fall into
subfreezing surface layer with a
temperature greater than -5 ˚C
* The lack of ice nuclei at -5 ˚C inhibits
the liquid water droplets from
freezing in the cold surface layer
(become supercooled)
S
FZRA
3 mm Maximum Diameter
* Contact of supercooled droplets with
subfreezing surface causes nearly
instant freezing upon the cold surface
(Stewart and King 1987)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Sounding Analysis (Freezing Rain)
Depth of Warm Layer
Height of Maximum
Temperature
Depth of Cold Layer
Tw
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Sounding Analysis (Freezing Rain)
Reports of FZRA at 00 and/or 12 UTC (1976-1990)
* Albany, New York - 18
* Bufffalo, New York - 16
* Greensboro, North Carolina - 18
* Peoria, Illinois - 10
* Spokane, Washington - 6
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Warm Layer Depth (Tw)
3000
2500
Depth (m)
2000
1500
1000
500
0
ALB
BUF
GEG
GSO
PIA
COMET RFC/HPC Hydrometeorology Course 02-1
All
12/3/01
Rawinsonde Data
* Data distributions at most stations are similar
* Median values:






Depth of warm layer = 1500 m
Depth of cold layer = 500 m
Height of maximum temperature = 1500 m
Maximum inversion temperature = 4˚C
“Warm” area of sounding = 2000 deg - m
“Cold” area of sounding = 125 deg - m
* Some variability exists between stations and events
* Important: Examine all sounding data to determine freezing
rain potential
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Ice Pellets
* ice pellets form when:
 snowflakes melt partially in an
elevated warm layer and fall into
a subfreezing surface layer
 fully-melted particles fall into a
deep (.5 - 1 km), cold (T < -10 ˚C)
surface layer.
* surface layer must be
sufficiently cold (T < -10˚C)
and deep enough to activate
ice nuclei and cause liquid
drop to freeze as it descends to
the ground
3 mm Maximum Diameter
(Stewart 1987)
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
November 8, 2000 12 UTC
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Microphysics Summary
* The amount of snowflake melting, and corresponding cooling,
is dependent upon air temperature, relative humidity,
snowflake size and density.
* Numerical and laboratory experiments show that complete
melting occurs between 0 and 800 m below the melting level
in saturated or unsaturated conditions.
* Rain/snow lines can exist on the warm side of the 0˚C
isotherm when the RH is less than 100% and the melting level
is within several hundred meters of the ground.
* Freezing rain conditions include a elevated deep warm layer
and a subfreezing surface layer that is usually no colder than
~ -5 ˚C.
* Ice pellets conditions include a elevated, shallow warm layer
and a cold (T < -5 ˚C) deep surface layer.
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Precipitation Type Forecasts Using
Numerical Model Data
* MOS
* Bulk microphysics scheme in numerical models
 Precipitation type through AWIPS browser (maybe) and NTRANS
* Experimental precipitation type algorithms
<www.spc.noaa.gov/exper/ptax>
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
MOS
* Precipitation-type
determined by
relationship between
model variables and
observed precip. type
* Advantages:
 Automated system
 3 hour output available
 Provides initial guess
* Disadvantages:
 Equations must be
derived for each location
 Only available every
three hours
 Accuracy uncertain
IAD E NGM MOS GUIDANCE 1/01/98 0000 UTC
DAY /JAN 1
/JAN 2
HOUR 06 09 12 15 18 21 00 03 06 09 12 15
MX/MN
37
22
TEMP 19 17 16 26 33 37 32 30 27 26 25 36
DEWPT 6 6 7 10 11 12 14 14 14 14 16 22
CLDS SC SCSCSCSC SCCLCLCLCLSCCL
WDIR 30 29 24 24 21 21 24 22 23 22 23 22
WSPD 08 06 04 08 10 11 12 13 13 12 09 11
POP06
0
0
1
0
0
POP12
1
0
QPF
0/ 0/ 0/0 0/ 0/0 0/ 0/0 0/ 0/0
PTYPE S S S S S S S S S R Z Z
POZP 0 0 5 4 3 13 21 19 17 12 38 25
POSN 89 97 95 96 97 87 64 55 42 21 0 0
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
ETA: Cloud/Precipitation Prediction
Water Vapor
Conde nsation
of Clouds
Evaporation
of Clouds
Clouds
Liquid Water
Ice Particles
Aggregation
Accretion Autoconversion
Rain
Autoconversion
Snow
Melting
Evaporation/Sublimation of Precipitation
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Experimental Algorithms
* Precipitation-type algorithms evaluate the entire
thermodynamic profile to determine the most probable type
of precipitation based upon precipitation microphysics.
* Algorithm currently run with ETA data does not use cloud
and precipitation data.
* Algorithm used with RUC data uses cloud and precipitation
data, as well as thermodynamic data.
* Current algorithms are superior to thickness rules since they
use all thermodynamic data and they incorporate physical
processes.
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Precipitation Type Algorithms
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
SPC Mesoscale Discussion
WINTER WEATHER MESOSCALE DISCUSSIONS
GOAL IS TO PROVIDE SHORT-TERM (0-6 HOUR) GUIDANCE
ON HAZARDOUS WINTER WEATHER FOR LOCAL NWS
OFFICES AND OTHER USERS BOTH BEFORE AND DURING
THE EVENT
- ISSUED FOR BLIZZARDS, HEAVY SNOW AND FREEZING RAIN
- FIRST PARAGRAPH PROVIDES THEWHAT, WHEN AND
- SECOND PARAGRAPH PROVIDES THE MESOSCALE
REASONING (THE WHY) OF THE FORECAST
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
Previous Mesoscale Discussion
SAMPLE WINTER WEATHER MESOSCALE
DISCUSSION
ZCZC MKCSWOMCD ALL;345,0867 385,0773 365,0773 325,0867;
ACUS3 KMKC 240455
MKC MCD 240455
NCZ000-SCZ000-TNZ000- 240800SPC MESOSCALE DISCUSSION #1124 FOR WRN NC...ERN TN...AND NWRN SC
CONCERNING...FREEZING
RAIN...
AREAS OF LIGHT FREEZING RAIN/DRIZZLE ARE EXPECTED
ACROSS
PARTS OF W RN NC AND NW RN SC. LIGHT TO MODERATE
TO CONTINUE
RAIN IS LIKELY OVER PARTS OF ERN TN. ICE
FREEZING
EXPECTED
TO VARY
BETW EEN 0.10 AND 0.50 AN INCH
ACCUMULATIONS
ARE
THROUGH 24/08Z.
24/00Z RUC2 AND ETA MODEL FORECAST SOLUTIONS SUGGEST AN AREA OF
ENHANCED VERTICAL MOTION ACROSS NRN AL/SRN MIDDLE TN WILL MOVE
ENEWD ACROSS ERN TN INTO WRN NC. LATEST IR SATELLITE IMAGERY
INDICATES COOLING CLOUD TOPS NEAR TCL TO HSV. ALTHOUGH BOTH
MODELS SUGGEST A WELL FOCUSED AREA OF UVV OVER ERN TN...
SATELLITE/RADAR AND SOUNDING DATA SEEM TO INDICATE A GENERAL AREA
OF UVV FROM NRN AL INTO SRN/SERN VA.
EXPECT AREAS OF LIGHT TO BRIEFLY MODERATE FREEZING RAIN WILL
GRADUALLY MOVE ENEWD AS UPPER LEVEL DIVERGENCE MOVES ACROSS
ERN TN AND WRN NC. FSL/AIRCRAFT SOUNDING DATA CONTINUES TO INDICATE
THERMODYNAMIC PROFILES REMAIN FAVORABLE FOR
PRECIPITATION.
FREEZING/MIXED
..VAN SPEYBROECK..
12/24/98
...PLEASE SEE WWW.SPC.NOAA.GOV FOR GRAPHIC
NNNN
PRODUCT...
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
References
* Baldwin, M., R. Treadon, and S. Contorno, 1994: Precipitation type
prediction using a decision tree approach with NMCs mesoscale eta
model. Preprints, 10th Conf. on Numerical Weather Prediction,
Portland, OR, AMS, 30-31.
* Czys, R., et al., 1996: A physically based, nondimensional parameter
for discriminating between locations of freezing rain and ice pellets.
Wea. Forecasting, 11, 591-598.
* Erickson, M., 1995: Evaluation NWS precipitation type forecasts.
Preprints, Sixth Conf. On Aviation Weather Systems, AMS, Dallas,
TX, 219-224.
* Hall, W.D., H.R. Pruppacher, 1976: The survival of ice particles
falling from cirrus clouds in subsaturated air. J. Atmos. Sci., 33,
1995–2006.
* Heppner, P., 1992: Snow versus rain: Looking beyond the “magic”
numbers. Wea. Forecasting, 7, 683-691.
* Keeter, K. and J. Cline, 1991: The objective use of observed and
forecast thickness values to predict precipitation type in North
Carolina, Wea Forecasting, 6, 456-469.
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01
References
* Matsuo, T., and Y. Sasyo, 1981: Melting of snowflakes below the
freezing level in the atmosphere. J. Met. Soc. Japan, 59, 10-25.
* Nakaya, U., 1954: Snow crystals. Harvard Univ. Press, 521 pp.
* Pruppacher H., and J. Klett, 1980: Microphysics of clouds and
precipitation. D. Reidel, 714 pp.
* Roger, R., 1979: A short course in cloud physics. Pergamon Press,
New York, 235 pp.
* Stewart, R., and P. King, 1987: Freezing precipitation in winter
storms. Mon. Wea. Rev., 115, 1270-1279.
* Stewart, R., and G. McFarquhar, 1987: On the width and motion of
the rain/snow boundary. Water Res. Res., 23, 343-350.
* Zerr, R., 1997: Freezing rain: An observational and theoretical
study. Wea. Forecasting, 36, 1647-1661.
* Zhao, Q., T. Black, and M. Baldwin, 1997: Implementation of the
cloud prediction scheme in the Eta model at NCEP. Wea.
Forecasting, 12, 697-712.
COMET RFC/HPC Hydrometeorology Course 02-1
12/3/01