Ewan O

1
Radar/lidar observations of boundary layer clouds

• Ewan OConnor, Robin Hogan, Anthony Illingworth,
  Nicolas Gaussiat

2
Overview

• Radar and lidar can measure boundary layer clouds
  at high resolution
• Cloud boundaries - radar and lidar
• LWP microwave radiometer
• LWC cloud boundaries and LWP
• Cloudnet compare forecast models and
  observations
• 3 remote-sensing sites (currently), 6 models
  (currently)
• Cloud fraction, liquid water content statistics
• Microphysical profiles
• Water vapour mixing ratio - Raman lidar
• LWC - dual-wavelength radar
• Drizzle properties - Doppler radar and lidar
• Drop concentration and size radar and lidar

3
Vertically pointing radar and lidar
Radar ZD6 Sensitive to larger particles
(drizzle, rain) Lidar bD2 Sensitive to small
particles (droplets, aerosol)
4
Statistics - liquid water clouds

• 2 year database
• Use lidar to detect liquid cloud base
• Low liquid water clouds present 23 of the time
  (above 400 m)
• Summer 25
• Winter 20
• Use radar to determine presence of drizzle
• 46 of clouds detected by lidar contain
  occasional large droplets
• Summer 42
• Winter 52

5
Dual wavelength microwave radiometer

• Brightness temperatures -gt Liquid water path
• Improved technique Nicolas Gaussiat
• Use lidar to determine whether clear sky or not
• Adjust coefficients to account for instrument
  drift
• Removes offset for low LWP

LWP - initial
LWP - lidar corrected
6
LWC - Scaled adiabatic method

• Use lidar/radar to determine cloud boundaries
• Use model to estimate adiabatic gradient of lwc
• Scale adiabatic lwc profile to match lwp from
  radiometers

http//www.met.rdg.ac.uk/radar/cloudnet/quicklooks
/
7

• Compare measured lwp to adiabatic lwp
• obtain dilution coefficient
• Dilution coefficient versus
• depth of cloud

8
Stratocumulus liquid water content

• Problem of using radar to infer liquid water
  content
• Very different moments of a bimodal size
  distribution
• LWC dominated by 10 ?m cloud droplets
• Radar reflectivity often dominated by drizzle
  drops 200 mm
• An alternative is to use dual-frequency radar
• Radar attenuation proportional to LWC, increases
  with frequency
• Therefore rate of change with height of the
  difference in 35-GHz and 94-GHz yields LWC with
  no size assumptions necessary
• Each 1 dB difference corresponds to an LWP of
  120 g m-2
• Can be difficult to implement in practice
• Need very precise Z measurements
• Typically several minutes of averaging is
  required
• Need linear response throughout dynamic range of
  both radars

9
(No Transcript)
10
Drizzle below cloud

• Doppler radar and lidar - 4 observables
  (OConnor et al. 2005)
• Radar/lidar ratio provides information on
  particle size

11
Drizzle below cloud

• Retrieve three components of drizzle DSD (N, D,
  µ).
• Can then calculate LWC, LWF and vertical air
  velocity, w.

12
Drizzle below cloud

• Typical cell size is about 2-3 km
• Updrafts correlate well with liquid water flux

13
Profiles of lwc no drizzle

• Examine radar/lidar profiles - retrieve LWC, N, D

14
Profiles of lwc no drizzle
260 cm-3
90 cm-3
80 cm-3

• Consistency shown between LWP estimates.

15
Profiles of lwc no drizzle

• Cloud droplet sizes lt12µm
• no drizzle present
• Cloud droplet sizes 18 µm
• drizzle present
• Agrees with Tripoli Cotton (1980) critical size
  threshold

16
Conclusion

• Relevant Sc properties can be measured using
  remote sensing
• Ideally utilise radar, lidar and microwave
  radiometer measurements together.
• Cloudnet project provides yearly/monthly
  statistics for cloud fraction and liquid water
  content including comparisons between
  observations and models.
• Soon - number concentration and size, drizzle
  properties.
• Humidity structure, turbulence.
• Satellite measurements
• A-Train (Cloudsat Calipso Aqua)
• EarthCARE
• IceSat

17
Importance of Stratocumulus

• Most common cloud type globally
• Global coverage 26
• Ocean 34
• Land 18
• Average net radiative effect is about 65 W m-2
• Cooling effect on climate

Mean annual low cloud amount ISCCP
18
Cloud Parameters

• Use radar and lidar to provide vertical profiles
  of
• Cloud droplet size distribution (N, mean D,
  broad/narrow)
• Drizzle droplet size distribution (N, mean D,
  broad/narrow)
• Relate drizzle to cloud N
• Is stratocumulus adiabatic? Entrainment rates

19
Data
20
Drizzle-free stratocumulus

• Z ND6 LWC ? ND3
• ? Z ? LWC2/N
• Assume adiabatic ascent and constant N
• LWC increases linearly with height (z)
• If we know T and p
• ? dLWC /dz

Assume dLWC /dz is a constant, a ? LWC(z) az
Z(z) ? (az)2 / N
Adiabatic profile Z should vary as z2
21
Aircraft data - ACE 2 Brenguier et al. (2000)
22
Refined technique

• Allow dilution from adiabatic profile of LWC

LWC(z) k LWCad(z) N k Nad
D(z) Dad(z)
Z(z) ? k (az)2 / Nad
23
Plots of N
High N, small D ? low Z
Nad 264 cm-3
24
Plots of N
Nad 91 cm-3
25
Plots of N
Nad 82 cm-3
26
Presence of drizzle can lead to an overestimate
of N ? an overestimate of LWC (and LWP)
27
(No Transcript)
28
Conclusion

• Consistency shown between LWP estimates from this
  technique, and from microwave radiometers.
• Additional techniques to investigate Sc are also
  available
• Doppler radar/lidar Drizzle properties
  (OConnor et al. 2004)
• Dual wavelength radar LWC profile (Gaussiat et
  al.)
• Doppler spectra
• Raman humidity measurements WV structure, mixed
  layer depths
• Aircraft verification?
• CloudNet 3 years, 3 sites, provide climatology
  of Sc properties

29
Dual wavelength microwave radiometer

• Brightness temperatures -gt Liquid water path
• Improved technique Nicolas Gaussiat
• Use lidar to determine whether clear sky or not
• Adjust coefficients to account for instrument
  drift
• Removes offset for low LWP

LWP - initial
LWP - lidar corrected
30
LWC - Scaled adiabatic method

• Use lidar/radar to determine cloud boundaries
• Use model to estimate adiabatic gradient of lwc
• Scale adiabatic lwc profile to match lwp from
  radiometers

http//www.met.rdg.ac.uk/radar/cloudnet/quicklooks
/
31

• Compare measured lwp to adiabatic lwp
• obtain dilution coefficient
• Dilution coefficient versus
• depth of cloud

32
Stratocumulus liquid water content

• Problem of using radar to infer liquid water
  content
• Very different moments of a bimodal size
  distribution
• LWC dominated by 10 ?m cloud droplets
• Radar reflectivity often dominated by drizzle
  drops 200 mm
• An alternative is to use dual-frequency radar
• Radar attenuation proportional to LWC, increases
  with frequency
• Therefore rate of change with height of the
  difference in 35-GHz and 94-GHz yields LWC with
  no size assumptions necessary
• Each 1 dB difference corresponds to an LWP of
  120 g m-2
• Can be difficult to implement in practice
• Need very precise Z measurements
• Typically several minutes of averaging is
  required
• Need linear response throughout dynamic range of
  both radars

33
(No Transcript)
34
Drizzle below cloud

• Doppler radar and lidar - 4 observables
  (OConnor et al. 2005)
• Radar/lidar ratio provides information on
  particle size

35
Drizzle below cloud

• Retrieve three components of drizzle DSD (N, D,
  µ).
• Can then calculate LWC, LWF and vertical air
  velocity, w.

36
Drizzle below cloud

• Typical cell size is about 2-3 km
• Updrafts correlate well with liquid water flux

37
Profiles of lwc no drizzle

• Examine radar/lidar profiles - retrieve LWC, N, D

38
Profiles of lwc no drizzle
260 cm-3
90 cm-3
80 cm-3

• Consistency shown between LWP estimates.

39
Profiles of lwc no drizzle

• Cloud droplet sizes lt12µm
• no drizzle present
• Cloud droplet sizes 18 µm
• drizzle present
• Agrees with Tripoli Cotton (1980) critical size
  threshold