1. Boundary layer depth, entrainment, decoupling, and clouds over the eastern Pacific Ocean
Robert Wood, Atmospheric Sciences,
University of Washington

2. Motivation
Processes in the subtropical MBL have enormous impact upon the structure and coverage of clouds
Subtropical MBL clouds strongly impact the earth’s radiation budget
These MBLs and the clouds therein are poorly represented in large scale numerical models

4. MBL depth, entrainment and decoupling
Integrative approach to estimate MBL and cloud properties in regions of low cloud
Combines observations from MODIS and TMI with reanalysis from NCEP and climatology from COADS
Results in estimates of MBL depth and decoupling (and climatology of entrainment)

5. MBL depth, entrainment, and decoupling

6. Methodology Independent observables: LWP, Ttop, SST
Unknowns: zi, q (= )
Use COADS climatological surface RH and air-sea temperature difference
Use NCEP reanalysis free-tropospheric temperature and moisture
Iterative solution employed to resulting non-linear equation for zi

7. Mean MBL depth (Sep/Oct 2000)
NE Pacific SE Pacific

8. Mean decoupling parameter q
Decoupling scales well with MBL depth

9. q vs zi-zLCL

10. Deriving mean entrainment rates
Use equation: we=uzi+ws
Estimate ws from NCEP reanalysis
Estimate uzi from NCEP winds and two month mean zi

11. Mean entrainment rates
Entrainment rate [mm/s] ◄ NE Pacific SE Pacific ► Subsidence rate [mm/s]

12. Summary of MBL depth work
Scene-by-scene estimation of MBL depth and decoupling
Climatology of entrainment rates over the subtropical cloud regions derived using MBL depth and subsidence from reanalysis
Decoupling strong function of MBL depth
Next step: deriving links between turbulence, inversion strength and entrainment by coupling to simple model forced with realistic boundary conditions

13. IPCC estimates of radiative forcing changes (present-preindustrial)
Tropospheric aerosol “indirect”
effect upon cloud reflectivity

14. ….cloud feedbacks remain the largest uncertainty in the prediction of future climate change
from Cess et al. 1989, 1996

16. Second aerosol indirect effect
AEROSOL INDEX
LIQUID WATER PATH [ g m-2]
GCM simulations show systematic increases of cloud liquid water content
with increasing aerosol
Simulations with prescribed cloud microphysical properties (for use in the precipitation scheme) do not
Aerosol Index =
3.915t865ln(t670/t865)
from Lohmann and Lesins, 2002,
Science, 298,

17. Aerosol-cloud microphysical observations
….first measurements (1967)
Aerosol-cloud microphysical observations
+ wind from land
 wind from sea
….recent measurements (1994)
Observed cloud droplet concentration [cm-3]
Cloud droplet concentration [cm-3]
Predicted droplet concentration
from aerosol spectrum [cm-3]
From Twomey and Warner,
J. Atmos. Sci., 24, , 1967
Aerosol concentration [cm-3]
From Martin et al.,
J. Atmos. Sci., 51, , 1994

18. Aerosol loading and cloud droplet radius…. ….35 years on
Aerosol index
from Breon et al. 2002, Science, 295,
Cloud droplet radius [micron]

19. Droplet concentration and drizzle
“maritime”
cloud base drizzle rate
[mm day-1]
“continental”
cloud droplet concentration [cm-3]
From Wood, 2005
J. Atmos. Sci., in press

20. EPIC 2001 – drizzle, cloud LWP, cloud droplet concentration Nd

21. Fundamental difficulties
Reflected SW for cloudy regions is a function of cloud optical depth t, itself a function of both cloud droplet size (re) and liquid water path (LWP):
Effect upon reflected SW depends upon both cloud optical depth and cloud cover C:
R=C RCLD + (1-C) RCLEAR
LWP re
t 
RCLD = f(t)

22. } SHIPTRACKS natural laboratories for the study of aerosol-cloud
interactions
SHIP TRACK
From Durkee et al.,
J. Atmos. Sci., 57,
, 2000
ALBEDO
DROPLET RADIUS [mm]
DROPLET CONC. [cm-3]
AEROSOL CONC. [cm-3]

23. t/t = -re/re + LWP/LWP
implying: lnt/lnre = 1 (for constant LWP) -lnt/lnre < 1 (for decreasing LWP) -lnt/lnre > 1 (for increasing LWP)
From Coakley and Walsh,
J. Atmos. Sci., 59, , 2002
-lnt/lnre

24. Confounding results? In shiptracks….
increased aerosols  smaller droplets
 decreased drizzle (e.g. Ferek et al. 2000)
 reduced LWP?
Important to remember that LWP is controlled not only by precipitation but also by radiation, entrainment, and surface fluxes….feedbacks that are poorly represented in many climate models

25. Bistable CCN concentrations?
Baker and Charlson (1990) suggest two regimes in the MBL:
R1: low CCN conc., drizzle is major CCN sink
R2: high CCN conc., aerosol coagulation is major CCN sink

26. stable equilibria unstable equilibrium
reproduced from Baker and Charlson,
Nature, 345, , 1990

29. particle diameter [m]
400
300
200
100
Na (D>0.1 mm) [cm-3]
0.7
0.3
0.1
0.05
0.03
0.01
dN/dlogR
[cm-3]
accumulation mode
particle diameter [m]
Aitken mode
day in November 2003

32. Coalescence
Probability that two drops, radii R and r, will coalesce in unit time is:
KR,r  (R+r)2 × [vt(R)-vt(r)] × HR,r × CR,r
KR,r  R5 × f(r/R)
(R2-r2)
R × f(r/R)
Area swept out
Relative fall speed of droplets
Collision efficiency
Coalescence efficiency
Coalescence probability increases
rapidly with droplet size

33. Remote sensing of cloud droplet concentration
[85W, 20S]

34. Effect of assuming uniform droplet size on TOA reflected SW flux
W m-2

35. Sea surface temperature
Southeast Pacific very poorly sampled
Net cloud forcing

36. Scanning C-band radar (EPIC 2001) drizzle is intermittent with cellular structure max reflectivity of dBZ 0
90
180
270
15 km
30 km
dBZ

37. Can drizzle affect MBL dynamics?

38. Cloud structural properties
Pockets Of Open Cells (POCs) are frequently observed in otherwise unbroken Sc. These may be formed as a cloud response to drizzle-induced decoupling
POCs

39. Microphysical-macrophysical links?
MODIS brightness temperature difference, GOES thermal IR, scanning C-band radar

40. POCs and aerosols POC CCN [cm-3] solid cloud -50 0 50 0 400 800 1200
Figures from Sharon et al.
(J. Atmos. Sci., in revision)
CCN [cm-3]
Ultrafine particles
[3-12 nm diameter]
solid
cloud
DISTANCE from POC edge [km]
Altitude [m]
Figure from
Petters et al.
(ICCP conf.
2004)
Concentration [cm-3]

41. DIURNAL CYCLE

42. IMET Buoy (85W, 20S) drizzle index (data courtesy of Robert Weller, WHOI)

43. Fraction of MODIS 256x256 km regions containing open cells
Open cellular convection frequency of occurrence
Fraction of MODIS 256x256 km regions containing open cells
Northeast Pacific
Southeast Pacific
…more likely to find open cells closer to land in SEP

44. Can precipitation affect MBL dynamics?
Mixed layer budget analysis using observationally-derived forcings (EPIC 2001) and entrainment, shows subcloud negative buoyancy fluxes ( decoupling) only when drizzle is included