Cirrus C’loud Properties and Detection..
K.N. Lieu, S.C. Ou, Y. Takano, N; Rae,P. Yang, and B. Barkey
University of Utah
● Light Scattering by Ice CrystaIs: Remote Sensing Implications
● Recent Development of Light Scattering by Ice Crystals
● Remote Sounding of the Optical and Blicrophysical Properties
of Cirrus Clouds Using AVHRR Data
\
‘-’&3
“<
co.-Gc
103
102
10’
s11-
10-’
———————Cirrostratus
——— —— Ice spheres
Laborato~ Data for ColumnsuVolkovitsky et al. (1980)
k = 0.65 pm
\
\/’
\ I
\ /\ .//’
II
1 OzI 1800
Sca;;ering Angle (d~gree)
Sa+ell;te RemOti SensinJ ,.
13
6
Fig. 2.
I
N=15 00 “
.9
I I6 7 8 9 10 11 12 13
LIDAR CLOUD CENTER (km)
Comparison of surface I.idar and satellitederived cloud-center heighk over Parsons, KS. ( ~f’t’er M/”nn}~et ●!~t~fz )
n
x-
i
●
6
5
4
3
2
1
0
/
/
/
/
/
“*” ●“k
/
f
{/
/ /●.4... .
. . /. .
,.:{ /
. .. /.*
“-z. .. . . f /
/
Spheroid~exagon //
(V=3 & V-5) /
:irrus /
Sphere
/
/
/
( I 1 I 1 I.
1 2 3 4 3
111.16 - T12 (K)
;
,.
Zenith Angie, @ (deg)
o 90I i I I I I I I I I 1 I I 1
+-+.=180° +-+O=U”
II \
—
/\1
#
“1-1I
,/i
. . . . . . . #measurement (Ci~*fi&7*I I I I I I I I I I I I I I I I I *
.1
0 60 120 18~phase Angle (deg)
c.
(
CLOUD TEMPERATURE (“c)
- 0’0”0959:55 0959:57
2.00
-9 -; 4 -9 -134 -9 -13”4
‘1
~~
0959:53
0 ALTITUDE (km )
~
1000:018“2”
M24 3’0
Fro. 4. Lidar field data (from ~lall et al 1~ showing Ihe considerable dacreasa in planar crystal backmttaring (in terms of Ihebackacatler coafficienl ~’) ana increasein’linear ~polarizalion (A used hare) as the lidar is scanned gfl the vertical direction (note thetirtad lima~ and lidar angles from the vertical direction), The cloud temperatures are within the expectad range for piate crystals, andan aanal sighting Of a subsurr conlirmad Wr uniform horizontal orientations,.— - —. .—-—. -. .. . .
Solid Column Solid Plate7’
Hollow Column- Dendrite
++
Fig. 2.
Bullet Rosette
Aggregate
The second and fourth columns &pict the
cbuds. The first and third
geometric model Our current
of “~e crystals includes all the
discussion).
Double Piate
o8*m,,*s,Quasi-spherical
ice crystal shape observed in cirrus
wbmns represent the shape simulated from a
capability of Aing I“ght scattering and absorption
shapes, except the aggregate (see text for further
10”
10
~o:: 10’L
u’$
f 10”
CL
1(J
10;
+“ 02
(a) A.= (~<r i,m
1./2a = ?rtn/ROum
14c,llnw(rJ.50pm) ~
—- Sollrl E
J I 1
0 60 I 20 180
(b)
——— —
(c) k = 0.55 pm
——.#
U2a= 120148pm
— ~ L/2a.200/80pm
L
\1~\
\ .-‘-\
0 60 120 180
Scattering Angle (degree)
(d)
——— —
I I b
o 60 120 180
ScatteringAngle (degree)
e) k = 0.55 pm
—.
@L/2a= 32/80pmbb= 121tm, bt=40p
– @ L/2a.32/8opnl
0 60 1?0 180
(f)
\1 I
(? 60 120 180
S,:atten~,] Anqle (clegree~ ‘
Fig. I Phase function and degree of linear polarization ab functions of the scattering angle for hollow columns (a and b), bullet rosettes
(c and d), and dendrites (e and f). The wavelength and the sizes of these ice crystals are depicted in the figure, where
hollow depth and bt and bb are the height and width (along the side of the plate surface) of the branches of a dendrite.d is the
.
t “-
*.
P - 9 $**”. P’w ..
i. ●●
T+”+:.””~~”,.. .. . ..
,’ . . .“:-+ :.’ ‘ -:”
$..e . ..*. “. -“””:. “,-:;,:: . . ..
w, . .. . . ~“.>,:”..:..”..< ,.. .. :-~._ -. . .“.. + . .
.4-..
&“””““.>...?
.g<d
. .-. —- .—
1iI
“i—
i
I
I
I
I
I
.
(a)
(b)
Light Scattering by Small Ice Crystals
100
n=z 10. m=l .31 + iO.Oo1=u ka=20z3uw
1-,m$
— FDTD
aa ...... GOM2~ o.1-i<
z0 :.:
z 0.01.
0.001 ,,,
Fig. 3
,,, ,.0 20 40 m Bo 100 120 140 160 Ieo
SCATTERING ANGLE ( 0 )
n=z 10 m.1.31 +iOO
oFg ka=40
3uw 1:rna
— FDTD
Ino .... GOM2
2
$
g001.
.=.
Oml I I I 1 I 4 ( 10 20406000100120140160 1
SCATTERING ANGLE ( 0 )
0
(a), The conceptual diagram for the scattering of waves by a small hexagon using thegeometric ray-tracing for the computation of the near field. The far field results are deteminedexactly by means of the electromagnetic wave theory. (b). Comparison of the phase functionsfor 2-D hexaaonal ice columns com~uted from the finite difference time domain (FDTD) and thenewgeomet;c optics model (GOM”2) for the size parameters of 20 and 40. ResuHs “from thetime consuming FDTD are considered to be exact (Yang and Lieu. 1994).
4rn=l.31+i0.0 FDTD
. GOM2
-- GOM1
“~— I-lJlu
m=l .31 +iO.O
3.5- ● GOM2
O;. !l’’’’ l’’’’’’’’” I
10 15 20 2530 35 40
24 I
TM POLARIZATION -- GOM1g
L
z
8f
TE POLARIZATION - -GOM1
I
16 ● 0(n
0.5-o~o10 15 20 25
2,
IJ-1.5:
-1.5-1
I -2 I I I I-2 - I I I I
io 15 20 25 3035 40
10 15 20 25 30 35 40
SIZE PARAMETERSIZE PARAMETER
— —
REMOTE SENSING OF CIRRUS
(CLOUDS: SCIENTIFIC OBJECTIVES
● Development of a retrieval scheme for the inference
of cirrus cloud parameters, including mean effective
ice crystal size and visible optical depth based on the
principle of radiative transfer and
parameterizations using the scattering and
absorption properties of hexagonal ice crystals and
AVHRR data, in particular, the 3.7 and 10.9 pm
channel radiances. (Ou et al., Applied Optics, April,
1993)
“ Removal of solar radiance in the 3.7 pm channel
data for application
on the correlation
channel radiances.
November, 1994)
to daytime cloud retrieval based
betw’een the 0.63 and 3.7 pm
(Rao et al., J. Appl. Meteor.,
● Verification of the retrieved cirrus cloud parameters
using available balloon-borne replicator and
sounding observations (FIRE-11-IFO).
“ Application of retrieved cirrus cloud parameters to
the computation of surface radiative fluxes using a
detailed radiative transfer model (Fu and Lieu,
1992, 1993), and comparison with ground-based
radiometer measurements. \
Algorithm Development(AVHRR Chs. 3 and 4)
● Upwellingradiances
observed,?
B(q) (l-E)Ia,.
////////////L
13 = 1=3(1- E3)+ E3B3(TC)
tIa
14 z IU4(1- E4) + E4B4(TC) > //’////’/////////’/// /
%4 = cirrus cloud IR emissivities9B3,4 = Planck intensities
Ia3.4 = upwelling cloud-base radiances
(1) Correlation between B3 and B4
(2) Parameterization of IR emissivities
&34 = 1 –exp(–k34T),9T = visible opti;al depth
Ed = 1 -(1 - &3)k”k’,k4/k3 = effective extinction ratio
Tc
(3) Evaluation of 1,3,4 from three-dimensionaldisplay of the frequency of occurrence of theAVHRR radiance pair(~3,]4)
Solution equation for cloud temperature
..
Ice crystal mean effective size is defined by
D = ice crystal width,~ = maximum dimension.
Based on radiative transfer calculations, &/k3 isparameterized in terms of I/D; as follows:
k4/k3= ;b,, (1 / De)”*co
From cloud microphysics measurements and theretrieved optical depth:
De = {T/[Az(~+~/De)IWC]]ED,Retrieve ~c and E from Chs. 3
and determine Deand T fromequations
and 4 radiances
parameterization
\
ALGORITHM DEVELOPMENT(Continued)
Ice crystal mean effective size is defined by,.
D = ice c~stal width, ~ = maximum dimension
Based on radiative transfer calculations, &/k3 isparameterized in terms of I/D, as follows:
k4 /k3 = ~b,l(l/De)n
3.5
2.5
2
1.5
1
\
Heymsfield & Platt (1 984)
Takano & Lieu (1989)
FIRE-I-IFO Measurement 1986)
■
■
o 50 100 150
De (Pm)
Kb.410J’AL OF SOLAR COSIPOX-ENT
IN THE 3.7 pm RADIANCE
Solar Reflectance for Chs. 1 (0.63 pm) and 3 (3.7pm)
,,
Construction of a Look-up Table ReIating r] and r3
Radiative transfer program (Takano and Lieu, 1989)
Sun-satellite geometry
Mean effective ice crystal size
0.126.=71°, 6=40’,4)= 146-
Cold Cirrus (23.9 pm). Ral =0,12, Ra3=O,M6
0.08
$/ t
:l~m Cirrostratus (41.5 pm)m
m I
D D m m8
‘0”04[~. .Nov. 1 (75.1 pm)
I
A 4w w NOV. 2 (930 pm)
FCi Uncinus (123.6 pm)
nu t I I I I
o 0.2I
0.4 0.6 0.8Ch.1 Reflectance
Verification (FIRE I IFO)
(a). Mean values of the retrieved Tc, E3, E4, r, k4/k3, De, and ZC for
cirrus pixels within a l“xl” ‘“scene westUTC, 28 October 1986, along with observed
of Fort McCoy at 0930
values.
Parameter Observation Retrieval
Tc (K) ..- 244 (233-255)
&3--- 0.36 (0.17-0.70)
&4 --- 0.42 (0.20-0.76)--- 1.08 (0.41-2.77)
k4;k3 --- 1.25 (1.07-1.44)De (pm) 107 (King Air) 104 (68-156)
ZC(km) 6-8 (Lidar) 7.5* (6-9)
* based on the Fort McCoy sounding at 0930 U’T’C.
(b). Cirrus cloud temperature, cloud height, and optical depth
determined from the present retrieval program.
TC(K) zC(km) TLocation present lidar* present lidar* present GOES*
Wausau 226,5 230.7 9.5 9.0 1.54 1.43Fort
McCoy 229.7 226.5 9.1 9.5 1.5 1.41
Madison 225.6 228.1 9.6 9.3 0.6 0.56
*from Minnis et al. (1990) using lidar and GOES data at 2100UTC, October 28, 1986
.
— -.,10 19 20 21 22 2:1 24 —-.–.–-L.—- r,!........./t,f-l.,, 1 ,..’
37.25
37.15
36.85
Satellite Data and Retrieval Results at2108 UTC, 5 December, 1991
1280
270
260
250
> 240.{.:
~30
220
36.7595.8 95.6 954 95.2 950 948
Longitude (’W)
13.0
2.5
.;.. 2.0
.’,,
:, 1,5
1.0
0.5
0.0
95.8 956 95.4 952 950 948
Longitude (’W)
37.25
37.15
36.75
95.8 95.6 954 952 950 948
Longitude (W)
(b),, (a)
Fig.1.
(1) Cl:small, quasi-spherical z = 12.96 kmT = -65.4 “C
(3) A: columns+ bullets+roseltes z = 11.32 kmO: quasi-spherical, irregular T = -54.5 “C
,6‘Y
-\1 I t I 1
(4) A: columns+ bullela+plalea z = 9.77 km
O: quasi-spherical, irregular T = -42.7 “C
(5) O: sublimated, irregular z = 9.64 kmT = -42.5 “C
0 100 200Maximum ~~msion (p%~
500 600
1
0.1
E~ 0.01
c
0.001
O.ml
1
0.1
z
‘~ 0.01
c
0.001
0.0001
b Dec. 5
●: smoothed.>: raw data
0s 50%AG 30%
C, P, BR 200/0
o 100 200 ‘+ 400Maximum3D%meter (pm)
500 600
INov. 26
c: smoothedo: raw data
Qs 59%
AG 37%
C, P, BR 470
0 100 200 300 400 500 mMaximum Dimension (pm)
(a). Five representative ice crystal size distrfbutims derived from the replicator sounding on 5 December 1991 during FIR E-11-IFO. (b). Averaged icecrysatal size distributions and smoothed curves for 5 December and 26 November data sets. The percentages of shapes for each distribution are also
shown: QS denotes quasi-spherical particles, including frozen droplets and distorted plates and columns; AG, C, P, and BR denote aggregate, column,plate and bullet rosette, respectively.
Schematic diagram of replicatorsounding system
95.8 95.6 95.4 95.2
Longitude (OW)
95 94.8
(a) RELATIVE HUMIDITY (%)
o 40 80 12016 I I
[ :k14 ““”....#
......
I
Cloud Top.---- A&. ----- ----- _____ _
..12 ....O
_loEx
4
“T......,...L
.....”----- ---=- .- ”___ L-----__________.. ~
~“”””-”l--xudBa:ir’.. R.H. j
NbAti-GIA. . . . . . .
~ Retrieved Tc (average...4. . of 81 pixels):#,.. * 10.9pm BT..“-.. . . .
. . . . . . . . .. <1\ ..”,. . . . . . . . . .
...”
2 -......,...................................................
0 I T“ 1 ,,,1-
-100 -80 -60 -40 -20 0 20TEMPERATURE (“C)
16
14
12
_loEx;B
$iii16
4
2
0
(b) MEAN EFFECTIVE SIZE (De, pm)
o 40 80 120 160 200 240
.- —_____ _____0.
1[1-----___
‘..
‘*, I Columns, bullet
i rosettes, and
‘k.: quasi-spherical
crystals
i “wSublimation zone----- _____ ___ M-- ----------
Cloud BaseI
i
● : De from replicator data
I : De from replicator data weighted
by number densityT : De from retrieval
A : T from retrieval
o : T from replicator data
7OA
1 1 I I i
0.6 0.8 1 1.2 1.4 1.6 1.8VISIBLE OPTICAL DEPTH
Ftg. 5. (a). Temperature and humidity profiles obtained fmm the NCAR-CLASS sounding system on 5 December 1991. Overlapped with the
temperature profile are the mean retrieved cloud temperature and mean Ch. 4 brightness temperature over a 0.05° x 0.2° domain
around tiffeyville. (b). Display of the replicator-derived mean effective sizes at the selected height levels, their verlical average, and
the retrieved value. Also shown on the bottom scale are the optical depths derived from the replicator data and from the retrieval.
Radiation hlodel
● Radiative Transfer
Delta-four-sUeam (Lieu, et al., 1988)
.Application to nonhomogeneous atmosphere (Lieu, 1975)
● Nongray Gaseous Absorption in Scattering AtmosphereCorrelated k-distribution approach (Fu and Lieu, 1992)
Spectral Intervals: 6 solar and 12 IR bands (Fu and Lieu, 1993)
● Single-Scattering Properties of Hydrometer
Hexagonal ice crystal (Takano and Lieu, 1989)
Water droplet (Fu, 1991)
Rain, Snow, and Graupel
● Vectorized Version on Supercomputer
FIRE-11-IFO (December 5, 1991)
400 ‘a)‘E~ 380‘!
320
300I
A
A
•1
●
o
NOAA radiometer
Model simulations:
satelltie cloud parameters
replicator cloud parameters
cloud climatology
black cloud at 255 K
clear
4
260 I I I I
20 20.5 21 21.5 22
nKa 2503
g 200”
a
(b)
o
w~ 150 :Lu~ 100 I I
20 20.5 21 21.5 22
TIME (UTC)
FIHE-11-IFO (November 26, 1991)
(a) RELATIVE HUMIDITY (~0)
120
“o~
. 12
10
4
2
:. f
..... NCAR-CLASS
“.....
A.,.. -
....+.
......
-----------
.....
..... .
.... . . . .
. . . . . ..-.. . . . . . . . . . . . . . . . . . . . . .. .
b-____+ -----t.;.. . . . . . . . . . . . . . . . . . . . . .
...... . . .... . . . . . . . .
.. . . . . . . . . . .. ..#...
. ..-.. . ...
. . . . . . . ... ... ..
J.. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . ..
...”.. . . ... ... . . .... .
....”.>:......I I I I I
:100 -80 -60 -40 -20 0 20TEMPERATURE (°C)
(b) MEAN EFFECTIVE SIZE (De, pm)
12
10
8
6
I ● : De from replicator data
4 I :De from replicator data weighted !
t:
by number densityT : De from retrieval 1
:~
o 1 2 3 4 5
VISIBLE OPTICAL DEPTH
FIRE-11-IFO (November 26, 1991)
460(a)
A
A
❑
+
o
Obse watio~:
NOAA radiometer
Mod el simulations:,.
satellke cloud parameters
repl’~ator cloud parameters
cloud climatology
black cloud at 248 K
clear
oI I I
20 20.5 21 21.5 22
(b)
w:150
Lu~ 100
20 2120.5 21.5 22
TIME (UTC)
Future Research
..
● Light Scattering by Small Xce Crystals (10 < a < 30)
c Validation: Remote Sounding of the Cirrus OpticalDepth and Ice Crystal Size in Cirrus/Low CloudConditions
● Development of h’ew Techniques for the Determinationof Ice Crystal Sizes Using 1.6 pm and VisibleWavelengths
——--- .—