Characterization and Radiative Impact of a

1
Characterization and Radiative Impact of a
Springtime Arctic Mixed-Phase Cloudy Boundary
Layer observed during SHEBA
Paquita Zuidema
University of Colorado/ NOAA Environmental
Technology Laboratory, Boulder, CO
2
Surface Heat Budget of the Arctic
SHEBA
3
Early May 76N, 165 W
4
WHY ?

• GCMs indicate Arctic highly responsive to
  increasing greenhouse gases (e.g. IPCC)
• Clouds strongly influence the arctic surface and
• atmosphere, primarily through radiative
  interactions
• Factors controlling arctic cloudiness not well
  known
• Observational evidence may support predictions
  (Serreze et al. 2000)

5
Arctic Sea Ice Extent in 2002 strongly diminished
relative to 1987-2001 mean
6
Annual warming dominated by winter and
spring spring warming 0.5 C/decade in
SHEBA region
Spring 1966-1995 Temperature Trends (Serreze et
al., 2000 Jones 1994)
7
spring
Increased Spring And Summer Cloudiness
summer
1982-1999 AVHRR data (WangKey, 2003)
annual
Persistent springtime cloud cover may advance
snowmelt onset date (e.g., modeling study of
Zhang 1996)
8
Project Goal

• characterize a multi-day arctic cloud sequence
  as best possible
• elucidate the underlying cloud physical
  processes
• assess the clouds radiative impact.

The Case May 1- May 10, 1998. Surface-based,
mixed-layer, mixed-phase cloud Overlaps with the
first two FIRE.ACE flights
Arctic Clouds Experiment
The challenge both ice and liquid phases are
present
9
Surface-based Instrumentation May 1-8 time series
8
dBZ
-5
-45
-20
6
35 GHz cloud radar ice cloud properties
km
4
2
depolarization lidar-determined liquid cloud base
Microwave radiometer-derived liquid water paths
100
g/m2
2
3
4
5
6
7
8
1
day
day
4X daily soundings. Near-surface T -20 C,
inversion T -10 C
4
1
8
lidar cloud base
z
-30C
-10C
10
(No Transcript)
11
LIQUID FIRST

• Liquid/ice discrimination
• based on
• depolarization ratio value
• backscattered intensity
• gradient

Depolarization ratio
ice
water
Aug
Nov
Monthly-averaged percentages of Vertical columns
containing liquid (grey bars)
May 6. Intrieri et al., 2002
12
adiabatic calculation constrained by
Microwave-radiometer-derived Liquid water paths

• microwave radiometer responds
• to integrated water vapor and
• liquid water
• physical retrieval also utilizes
• - cloud temperature
• - soundings
• decreased uncertainty
• (good for Arctic conditions)
• Yong Han, unpublished data

13
May 4 7 NCAR C130
Research Flights
range (micron)
parameter
instrument

• FSSP-100 2-47
  liquid, ice size distribution
• 1D OAP-260X (May 4) 40-640 ice size
  distribution
• 2D OAP (May 7) 25-800 ice
  shape, size
• Cloud Particle Imager 5-2000 particle
  phase, shape, size
• King hot-wire probe liquid water
  content

establishes liquid droplet concentration and
distribution width
14
Liquid Characterization
May 4
Cloud radar reflectivity
dBZ
0
-50
-50
2
Height (km)
1
Temperature inversion
Aircraft path
Lidar cloud base
2400
UTC
2200
2300
time
15
Liquid Water Content Adiabatic Ascent
Calculation
May 4

• lidar-determined liquid cloud base parcel
• interpolated sounding temperature structure
• constrained w/ microwave radiometer-derived
  liquid water path

adiabatic LWC
1.0
King LWC
CB
0.6
Z (km)
excellent correspondence between adiabatic calc.
and King probe LWC
0
0.5
Liquid water content g/m3
16
Derivation of liquid volume extinction
coefficient b and effective particle radius re
May 4

• Lognormal droplet size distribution
• ltrkgt ltrokgtexp(k2s2/2) (Frisch et al.,
  95,98,02)
• mean aircraft cloud droplet conc.
• N222 (14)
• mean aircraft lognormal of geometric standard
  deviation of droplet size distribution
• s 0.242 (0.04)

re
b
adiabatic
aircraft
17
May 7 thin cloud, low LWP
Lidar depolarization ratio
High aerosols ! Max of 1645/L (Rogers et al.,
2001)
Backscattered intensity
18
May 1 10 liquid b, re, t time series
b
0
30
re
0
re
2
3
4
5
8
6
7
1
day
day
30
t
0
Mean liquid cloud optical depth 10, mean r_e
4.5
19
May 1-3 Mean Sea Level Pressure
Weak low N/NW of ship followed by weak/broad
high moving from SW to NE
Data courtesy of NOAA Climate Diagnostics Center
May 4-9 Mean Sea Level Pressure
Boundary-layer depth synchronizes w/
large-scale subsidence
20
ICE microphysics retrieval

• radar only (Matrosov et al. 2002 2003)
• particle size retrieved from Doppler velocity
• particle mass retrieved from reflectivity
• particle size

MAY 5
21
ISSUES

• Radar retrieval developed for ice clouds, not
  iceliquid clouds
• Radar not sensitive to the smaller particles
• Another degree of freedom Particle shape
• for bulk aircraft measurements, complete size
  distributions difficult to form

Comparison to aircraft data uncertain IWC
comparison most reliable (not D or b)
22
May 4 Cloud Particle Imager data
pristine ice particles from upper cloud
...super-cooled drizzle
23
May 4 complete size distribution FSSP (), CPI
(line), 260X (triangles)
24
Robust conclusions

• Radar reflectivity insensitive to liquid
• when ice is present
• Radar retrievals agree with aircraft-derived
  values given large uncertainties (4x ?)
• Ice cloud optical depth almost insignificant

dBZ
b
radar
liquid
Ice aircraft
Deff
IWC
25
What is the radiative impact of the ice ?

• Direct impact negligible mean ice cloud optical
  depth 0.2
• BUT
• 1) upper ice cloud sedimentation associated with
  near-complete or complete LWP dissipation (May 4
  6)
• 2) local IWC variability associated with smaller
  LWP changes, time scale few hours

At T-20C, air saturated wrt water is 20
supersaturated wrt ice
26
Ice water content/LWP time series
27
Mechanism for local ice production

• Liquid droplets of diameter gt 20 micron freeze
  preferentially, grow, fall out
• New ice particles not produced again until
  collision-coalescence builds up population of
  larger drops
• Only small population of large drops required
• Hobbs and Rangno, 1985 Rangno and Hobbs, 2001
  Korolev et al. 2003 Morrison et al. 2004
• Little previous documentation within cloud radar
  data

28
Local ice production more evident when boundary
layer is deeper and LWPs are higher
May 3 counter-example variable aerosol
entrainment ?
Quick replenishment of liquid longer-time-scale
variability in cloud optical depth related to
boundary layer depth changes
29
Project Goal

• characterize a multi-day arctic cloud sequence
  as best possible
• elucidate the underlying cloud physical
  processes
• assess the clouds radiative impact.

The Case May 1- May 10, 1998. Surface-based,
mixed-layer, mixed-phase cloud Overlaps with the
first two FIRE.ACE flights
Arctic Clouds Experiment
30
Radiative flux closure and cloud forcing
Implement derived cloud properties within
radiative transfer model
Streamer (Key Schweiger Key 2001).
Medium-band code, utilizes DISORT (Stamnes et al.
2000)
Strength comprehensive, adapted for Arctic
climate problems

• Both phases represented within a single volume
• Shortwave ice cloud optical properties
  parameterized for 7
• particle habits
• Arctic aerosol profile available
• surface albedo spectral variation adequately
  represented

Weakness 4 gases only, outdated gaseous line
information
31
Clear-sky comparison (May 7 April 25)

• SHEBA spectral surface albedo data (Perovich et
  al.)
• time-mean broadband albedo 0.86
• (matches surface-flux albedo)

• Arctic haze aerosol profiles constrained with
• sunphotometer measurements (R. Stone, unpub.
  data)
• Aerosol optical depth 0.135 _at_ 0.6 micron

• Ozone column amount 393 DU (TOMS J. Pinto
  pers. comm.)

Shortwave and infrared calculated and measured
Downwelling surface fluxes agree to within 1
W/m2
32
Most common ice particle habit aggregate
aggregates, smallbig
spheres
(below liquid cloud base)
33
Comparison of modeled to observed surface
downwelling radiative fluxes, May 1 -8

• modeled LW gt observed LW by 1 W m-2 RMS dev.
  13 W m-2 or 13 of observed fluxes
• modeled SW gt observed SW by 3 W m-2 RMS dev.
  17 W m-2 or 12 of observed fluxes
• Bias slightly larger for low LWP cases
• Small bias encourages confidence in data (better
  agreement cannot be achieved w/out exceeding
  estimated uncertaintities)

longwave
shortwave
observed
modeled
W m-2
34
How do clouds impact the surface ?
Surface albedo0.86 most SW reflected back
Clouds warm the surface, relative to clear skies
with same T T RH, by time-mean 41 W m-2
(little impact at TOA)

• Can warm 1m of ice by 1.8 K/day, or melt 1 cm of
  0C ice per day,
• barring any other mechanisms !

35
For cloud optical depthlt3, net cloud forcing
dominated by longwave gt Sensitive to optical
depth changes
Longwave
For cloud optical depth gt 6, net cloud forcing
dominated by shortwave gt Sensitive to solar
zenith angle, surface reflectance changes
Shortwave
30 of cloud optical depths lt 3 60 gt 6
Net
Cloud optical depth
36
How sensitive is the surface to cloudiness
changes ?

• Satellite-based study concludes surface cloud
  forcing most
• sensitive to changes in cloud amount, surface
  reflectance, cloud
• optical depth, cloud top pressure (Pavolonis and
  Key, 2003)

D LWP (g m-2) 5 20 -5 -20
D CF (W m-2) 2 3 -3.5 -10
D surface 0.05 -0.05
D CF (W m-2) 4.5 -3.8

• Little radiative impact from additional water
• Surface reflectance changes may be more
  radiatively significant

37
Why is this cloud so long-lived ????

• Measured ice nuclei concentrations are high
  (mean 18/L, with
• Maxima of 73/L on May 4 and 1654/L (!) on May 7
  (Rogers et al. 2001)
• This contradicts modeling studies that find
  quick depletion w/ IN
• conc of 4/L (e.g. Harrington et al. 1999)

One part of the explanation

• Cloud-top radiative cooling rates can exceed 65
  K/day
• Strong enough cooling to maintain cloud for any
  IN value (Pinto 1998)
• Promotes turbulent mixing down to surface,
  facilitating surface fluxes

How did this cloud finally dissipate ????
Strong variability in subsidence rates part of
answer
38
Most interesting results

• Radiative flux impact of this mixed-phase cloud
  is close to
• that of a pure liquid cloud
• Two mechanisms by which ice regulates the
  overall cloud
• optical depth
• Sedimentation from upper ice clouds
• A local ice production mechanism, though to
  reflect
• the preferred freezing of large liquid droplets
• ..but liquid is quickly replenished
• Longer-time scale changes in cloud optical depth
  appear
• synoptically-driven
• CONCLUSIONS DERIVE THEIR AUTHORITY FROM A
  COMPREHENSIVE
• CHARACTERIZATION OF BOTH LIQUID AND ICE PHASE

39
What might a future climate change scenario look
like at this location ?
Recent observations indicate increasing
springtime Arctic Cloudiness and possibly in
cloud optical depth (Stone et al., 2002, Wang
Key, 2003, Dutton et al., 2003)
At this location (76N, 165W) an increase in
springtime cloud optical depth may not
significantly alter the surface radiation
budget, because most cloudy columns are already
optically opaque.
A change in the surface reflectance may be more
influential
40
Acknowledgements Brad Baker Paul Lawson Yong
Han Jeff Key Robert Stone Janet
Intrieri Sergey Matrosov Matt Shupe Taneil
Uttal
Submitted journal article available
through http//www.etl.noaa.gov/pzuidema
41
Summary Conclusions

• Arctic mixed-phase clouds are common, radiatively
  and climatically important
• Can characterize the liquid with an adiabatic
  ascent calculation using a saturated air parcel
  from the lidar-determined liquid cloud base,
  constrained with the microwave radiometer-derived
  liquid water path
• The ice component can be characterized with cloud
  radar retrievals, even when LWC is high
• This was applied to a May 1-10 time series with
  some success, judging from comparison to aircraft
  data and comparison of calculated radiative
  fluxes to those observed.
• For May 1-10 radiative flux behavior is
  practically that of a pure liquid cloud
• The low ice water contents are consistent with
  what is required for the maintenance of a
  long-lived super-cooled ( -20 C) liquid water
  cloud (e.g., Pinto, 1998, Harrington, 1999)
• Usefulness of the technique can be improved even
  further by improving the microwave radiometer
  retrievals of liquid water path

42
Characterization and Radiative Impact of a
Springtime Arctic Mixed-Phase Cloudy Boundary
Layer observed during SHEBA
Paquita Zuidema
Height (km)
University of Colorado/ NOAA Environmental
Technology Laboratory, Boulder, CO
Temperature inversion
1.0 km
Aircraft path
Cloud radar reflectivity
Lidar cloud base
time
43
Liquid phase top agrees well with the location of
the temperature inversion
2 km
Cloud radar top
1 km
Temperature inversion
day
1
1
10
9
8
7
6
5
4
3
2
44
Aircraft-adiabatic calc. optical depth comparison
with microwave, agreement to 10 w/out
microwave, agreement to a factor of 2
Uses microwave LWP
tadiabatic
taircraft
45
ICE (radar)

• Remote retrieval depends only on cloud radar
• Radar-based retrieval developed for all-ice
  clouds (Matrosov et al. 2002, 2003) extended to
  mixed-phase conditions, relies only on Z, V.
• IWCZ/(GD3) where G assumes exponential size
  distribution, Brown and Francis bulk density-size
  distribution
• EXTZ/(XD4) X also assumes a mass-area-size
  relationship for individual particles
• Correction accounts for dry air density variation
  with height
• DEFINE D_effective1.5IWC/(rA) (Mitchell 2002
  Boudala et al. 2002)