mesoscale variability and drizzle in stratocumulus kim comstock general exam 13 june 2003
DESCRIPTION
Data set Meteorological measurements on ship and buoy (T, q, U, LW, SST) Ceilometer MMCR and C-band radar GOES satellite imagery EPIC 2001 Sc dataTRANSCRIPT
Mesoscale variability and drizzle in stratocumulus
Kim Comstock
General Exam13 June 2003
EPIC 2001 Sc cruise
image courtesyof Rob Wood
x
EPIC 2001 Sc cruise
Data set• Meteorological measurements on
ship and buoy (T, q, U, LW, SST)• Ceilometer • MMCR and C-band radar• GOES satellite imagery
EPIC 2001 Sc data
Why are Sc important?• Areal extent and persistence• Effect on radiation budget
Key parameter: Sc albedo
• mean droplet size– CCN aerosols
• cloud thickness– turbulence, entrainment, drizzle
•diurnal and mesoscale variations• horizontal variability
– mesoscale circulations– drizzle?
Central QuestionsTo understand the physical processes
that govern variability in Sc albedo, we must answer the following questions:
• What is the structure and life cycle of Sc?
• What is the role of drizzle in mesoscale variability?
• What role does the diurnal cycle play?
Goals: using EPIC data to address central
questions• Determine drizzle cell properties
from C-band radar.• Obtain and physically interpret
signatures of mesoscale variability from ship and buoy time series.
• Estimate amount of drizzle and relate to mesoscale variability.
• Analyze diurnal cycle and determine how it modulates all of the above.
MMCR time-height section
-60 -40 -20 0 15 dBZ
hourly cloud top
hourly LCL
hourly cloud base
Quantifying drizzle• We have reflectivity (Z) over a wide area
around the ship from the C-band radar, but we want to know rain rate (R) information.
• No suitable Z-R relationships exist for drizzle.• We developed Z-R relationships, Z=aRb , from in-situ DSD data at cloud base and at the surface:– aircraft (N Atlantic) and surface (SE Pacific) data– linear least squares regression (log10Z, log10R)
• Ideally, we want to know R at the surface.
Quantifying drizzle - method
• Evaporation-sedimentation model– assumes truncated exponential drop-size
distribution (DSD) with mean size r – run with various r’s and drop concentrations
• Obtain model reflectivity profiles (Z(z)/ZCB) and compare with MMCR profiles.– infer DSD for each MMCR profile– use model to extrapolate cloud base DSD
characteristics to the surface (get surface R)• Develop “bi-level” Z-R relationship using
cloud base ZCB to predict surface Rs.
Quantifying drizzle - results
• Apply bi-level Z-R to C-band cloud reflectivity data to obtain area-averaged rain rate at the surface.
• Average drizzle rates for EPIC Sc– 0.93 mm/day at cloud base (range 0.3-3)– 0.13 mm/day at the surface (range 0.02-
0.6)• Uncertainties due to
– C-band calibration (2.5 dBZ)– Z-R fitting procedure
Diurnal cycle• At night the BL tends to be well
mixed (coupled). • During the day, the BL is less well
mixed (decoupled). • It tends to drizzle most during the
early morning.
Coupled BL
U
UT
cloud thickness 410 ± 60 mcloud base 930 ± 30 m
~ 30 km
Decoupled BL
cloud thickness 310 ± 110 mcloud base 930 ± 60 m
Drizzling BL
cloud thickness 415 ± 150 mcloud base 890 ± 110 m
Mesoscale variabilityGoes 8 Visible19 October
0545 Local Time
Summary of previous work
• Though the diurnal signal is dominant, mesoscale structure is an integral part of the dynamics of the Sc BL.
• BL time series classified as coupled, decoupled or drizzling.
• There is a significant amount of drizzle in the SE Pacific BL, and it is associated with increased mesoscale variability
Future work• Compare Sc mesoscale structure with
previous studies of mesoscale cellular convection (MCC)
• Further examine radar data for 2-D and 3-D information– circulations (also use DYCOMS II and
possibly TEPPS Sc) – compositing/tracking
• Analyze buoy time series for mesoscale variability in relation to “drizzle”.
MCC comparisons Compare our coupled cell with
closed cell from Rothermel and Agee (1980)
q
Radial velocities• EPIC C-band volume-scan radial velocities
are probably unusable due to pointing errors associated with these scans.
• Vertical RHI scans appear less susceptible to error, so the radial velocity data (in the RHIs) may be useful for qualitatively looking at 3-D circulations in the BL.
• TEPPS volume scans and DYCOMS II vertically-pointing radar data are other possibilities.
Example
0
90
180
270
15 km
30 km
EPIC Sc RHIs 17 October 2001 1058 UTC0
90
180
270
2 km 19 km
dBZ
EPIC Sc RHIs 17 October 2001 1058 UTC0
90
180
270
2 km 19 km
m/s
Comparison with DYCOMS II
• Anticipate receiving DYCOMS II aircraft data (vertically-pointing MMCR data and time series)– look for circulations associated with
closed cells and drizzling conditions– look at variability associated with
drizzle (flight RF02)
C-band compositeCell 1Cell 2
Compositing/tracking: preliminary results
• Examples from tracked drizzle cells
Time avg PDF of dBZ Average reflectivity
Time (hr UTC)dBZ
Drizzle’s signature• Air-sea temperature difference
appears to be a good indication of drizzle occurring in the area.
Drizzle’s signature
Drizzle climatology• Will apply air-sea T analysis to
year-long buoy time series to determine – frequency and persistence of drizzle – diurnal cycle information– cloud fraction associated with drizzle
•Longwave radiation can be used as a proxy for cloud fraction in the buoy data series.
– relationship to satellite images
Buoy data• Example of SST-Ta for 15
September 2001 satellite overpasses
Buoy dataGOES 8 IR 1145 UTC
WHOI BUOY
Buoy dataGOES 8 Vis 1445 UTC
WHOI BUOY
Buoy dataGOES 8 Vis 1745 UTC
WHOI BUOY
Buoy dataGOES 8 Vis 2045 UTC
WHOI BUOY
ScheduleDate GoalSummer 03 Submit Z-R paperSummer 03 Compositing & sizing of drizzle
cellsSummer-Fall 03 Contribute to broken
cell/drizzle paperFall 03 Submit mesoscale variability
paperWinter-Spring 04 C-band radial velocity analysisSpring-Summer 04 DYCOMS II data analysisSummer-Fall 04 Satellite – time series analysisWinter 05 Finish
LW as a proxy for cloud fractionLW
-T a4 (
W/m
2 )
Drizzle and open cellsGOES image (color) and C-band reflectivity (gray
scale)GOES image only
(Less) drizzle and closed cells
GOES image (color) and C-band reflectivity (gray
scale)GOES image only
Evaporation-sedimentatio
n model
r (m)
N (#/L)
C-band Sc Volume Scan
MCC – closed cell
Moyer&
Young 1994
Tracking algorithm
Williams and Houze 1987