Robin Hogan

1
Cloud and Climate Studies using the Chilbolton
Observatory

• Robin Hogan
• Department of Meteorology
• University of Reading

2
Introduction

• Cloud feedbacks remain the largest source of
  uncertainty in predicting the global warming
  arising from increased CO2 (IPCC 2007)
• Better observations of clouds are needed to
  tackle this problem
• More than a decade of observations at Chilbolton
  have been used to
• Directly evaluate cloud representation in weather
  climate models
• Improve understanding of physical processes in
  clouds
• Develop algorithms for spaceborne radar (CloudSat
  and EarthCARE)
• This has involved the combination of
• Near-continuous vertically pointing radar and
  lidar observations (e.g. ESA C2 project, EU
  Cloudnet project)
• Focussed field campaigns together with
  meteorological aircraft (e.g. CLARE98, CWVC,
  CSIP)

3
Cloud observations at Chilbolton

• Cloud radars
• 35-GHz since 1994 (Rabelais then Copernicus)
• 94-GHz since 1996 (Galileo)
• Can also use 3-GHz CAMRa for clouds
• Cloud lidars
• 905-nm since 1996 (CT75K)
• 1.5-?m Doppler lidar since 2006 (HALO)
• 355-nm RAMAN and polarization lidars
• plus many other passive instruments!
• Chilbolton has led the way in methods to combine
  instruments at different wavelengths to retrieve
  cloud properties

4
Basics of radar and lidar
Radar ZD6 Sensitive to large particles (ice,
drizzle) Lidar bD2 Sensitive to small
particles (droplets, aerosol)
Radar/lidar ratio provides information on
particle size
5
Target classification

• First task use different radar and lidar
  sensitivities to identify different types of
  clouds and other atmospheric targets
• From this we can estimate cloud fraction and
  other model variables

Cloud radar Cloud lidar
6
Cloud fraction comparison for a month
Observations
7
Evaluation of 7 forecast models

• Cloud fraction and ice water content for 2004

Good news ECMWF and Met Office ice water
contents are within observational errors at all
heights
Bad news all models except DWD underestimate
mid-level cloud fraction, and there is a wide
range of low-cloud amounts
Bulletin of the American Meteorology Society, in
press
8
Liquid water content

• LWC derived using the scaled adiabatic method
• Lidar and radar provide cloud boundaries,
  adiabatic LWC profile then scaled to match liquid
  water path from microwave radiometers

0-3 km
9
Cloud overlap
Most models assume maximum-random overlap

• Cloud fraction and water content alone is not
  enough climate models need to know how clouds
  overlap

10
Cloud overlap global impact

• Chilbolton overlap retrievals were tested in the
  ECMWF model effect on radiation budget is
  significant, particularly in the tropics

Difference in outgoing infrared radiation between
maximum-random overlap and new approach
5 Wm-2 globally
ECMWF model run by Jean-Jacques Morcrette
11
Mixed-phase clouds

• Clouds containing a mixture of super-cooled
  liquid droplets and ice particles are a major
  headache in climate prediction
• In a warmer atmosphere these clouds are more
  likely to be liquid, making them more reflective
  and longer lasting, a negative feedback
• Chilbolton can identify them using lidar and
  radar
• Liquid droplets are much smaller and much more
  numerous than ice, so are much more reflective to
  lidar than to radar

35-GHz radar
Large falling ice particles
905-nm lidar
Small supercooled liquid droplets
12
Supercooled water occurrence

• Chilbolton lidar was used to estimate occurrence
  of supercooled water over a 1-year period
• 15 of mid-level ice clouds contain significant
  liquid water, decreasing with temperature
• Similar results were obtained from a lidar in
  space
• Radiative transfer calculations reveal that the
  liquid water interacts much more strongly with
  solar and infrared radiation than ice, so it is
  crucial to get the phase right
• These results are informing the development of
  models, which poorly represent this behaviour

13
Mixed-phase clouds

• Physics very uncertain
• Represented very crudely in models
• Layers detected during CLARE98 experiment
• Highly reflective to lidar ? optically thick
• Low depolarisation ? spherical particles
• Invisible to radar ? very small particles
• In situ confirmation of liquid water droplets

Lidar backscatter (from aircraft above)
C-130 liquid water (-7ºC)
Lidar depolarisation (from aircraft above)
Radar reflectivity
14
Radiative effects of ice and liquid

• We use radar and lidar to derive profiles of IWC
  and effective radius, used in radiation
  calculations
• Supercooled water most significant in short-wave
• Can reduce net absorbed radiation by more than
  100 Wm-2
• In daylight, usually more important than any ice
  present

Liquid water layer
?
Hogan et al. (QJ 2003a)
15
The future

• Information for high-resolution models
• Both forecast and climate models are becoming
  more sophisticated in their representation of
  clouds but not necessarily more accurate!
• Use Chilbolton to evaluate model representation
  of turbulence intensity, cloud particle fall
  speeds, cloud variability etc.
• Cloud processes need to be understood in more
  detail, e.g. the interaction of aerosols with
  clouds (NERC APPRAISE project)
• Assimilation of cloud radar data into forecast
  models?
• Exciting new technology for cloud observations
• E.g. development of the first cheap,
  continuously operating Doppler lidar for cloud
  and boundary-layer studies, now at Chilbolton
• Spaceborne cloud radar and lidar
• Algorithms developed at Chilbolton will be used
  by the CloudSat and Calipso satellites (launched
  a year ago)
• Chilbolton observations have been used to build
  the science case for the ESA EarthCARE
  satellite (to be launched in the next 5 years)