UK informal Definition * to leave a place in order to avoid a difficult

Report
EarthCARE:
The next step forward in
global measurements of
clouds, aerosols,
precipitation &
radiation
Robin Hogan
ECMWF & University of Reading
With input from many of the
EarthCARE science team
Earth Clouds, Aerosols and
Radiation Explorer (EarthCARE)
Joint ESA/JAXA mission, launch 2017, budget 700 MEuro
 EarthCARE will quantify will measure clouds, aerosols,
radiation and precipitation with unprecedented accuracy
 EarthCARE will be used to retrieve profiles of radiative
fluxes and heating rates at a target accuracy of 10 W m-2
 It will provide essential data to evaluate weather and
climate models, and to improve their representation of
cloud-aerosol-precipitation-radiation interactions
 Level-1 data available in real time for data assimilation
EarthCARE combines four instruments: radar, lidar,
imager and broad-band radiometer
The A-Train versus EarthCARE
The A-Train (fully launched 2006)
–
–
–
–
–
–
–
NASA
Multiple platforms
700-km orbit
CloudSat 94-GHz radar
Calipso 532/1064-nm lidar
CERES broad-band radiometer
MODIS multi-wavelength radiometer
EarthCARE (launch 2017)
–
–
–
–
–
–
–
ESA and JAXA
Single platform
393-km: higher sensitivity
94-GHz Doppler radar
355-nm High spectral res. lidar
3-view broad-band radiometer
Multi-spectral imager
EarthCARE
Viewing geometry
•
•
•
•
Satellite mass: 2000 kg
Solar panel area: 21 m2
Altitude: 393 km to maximize sensitivity
Radar and lidar power consumption: 2.5 kW
Principle of high spectral
resolution lidar (HSRL)
• EarthCARE will derive extinction profile (crucial for radiative
transfer) unambiguously from attenuated air backscatter
Extinction-to-backscatter ratio, sr
Aerosol classification from lidar
• Exciting potential for aerosol classification via two independent
pieces of information:
– Lidar depolarization (a measure of non-spericity)
– Extinction-to-backscatter ratio (affected by size and absorption)
Example A-train observations
Classification:
Ceccaldi et al. (JGR 2013)
• CAPTIVATE retrieval: Cloud, Aerosol and Precipitation from
Multiple Instruments using a Variational Technique
– Error estimates include contrib. from measurement and model error
– Looks good but is it right?
Illingworth et al. (BAMS 2014)
CAPTIVATE
retrieval
Applied to an ATrain profile
containing
ice/snow and rain
• Synergy of radar, lidar and visible/infrared narrow-band
radiances used to retrieve best estimate of cloud, precipitation
and aerosol properties
• Works also with EarthCARE instrument specification
What would EarthCARE see?
•
Use radar and lidar forward models for the EarthCARE instrument specifications:
•
Note EarthCARE radar’s higher sensitivity: most ice clouds will be detected by both
radar and lidar, enabling accurate particle size in a much higher fraction of clouds
3D scene
reconstruction
• CAPTIVATE retrievals
under the radar-lidar
curtain will be used to
reconstruct the 3D scene
based on visible/infrared
radiances
• 3D radiation calculations
then used to continually
test reconstructed scene
by comparing to 3-view
broadband shortwave and
longwave radiances
Barker et al. (QJRMS 2011)
1D-Var assimilation of cloud radar reflectivity
Case 20070123
over Pacific
Observations
First guess
Analysis
Improved
match to
observations
after
assimilation
Reading, UK
Janiskova et al. (2012)
ECMWF are
ready to
assimilate
EarthCARE
radar and
lidar data in
realtime
Tested
offline on
CloudSat
and
CALIPSO
© ECMWF 2013
Impact of 1D+4D-Var assimilation on subsequent forecast
a) 10.7µm100°W
TB GOES12 2008042421
Mean=278.1380°W
K
90°W
Case
20080424
K
320
312
304
10m
simulated TB
from 9-hour
forecast:
|obs-exp||obs-ref|
296
288
40°N
40°N
280
GOES-12
observations
272
264
256
248
240
232
30°N
30°N
224
216
208
200
100°W
c)
90°W
80°W
Precipitation
20080424 12:00 for t9
- t6
100°W- NEXRAD OBS:90°W
80°W
33
333
33
33
33
333
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
333
33
33
33
33
33
333
33
333
33
333
33
33
333
33
33
33
33
333
33
33
333
333
333
333
333
33
33
333
33
333
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
333
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
333
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
33
3
40°N
3-hour
accumulated
NEXRAD
precipitation
30°N
12
10
8
6
40°N
4
• Positive impact on forecasts
felt up to 48 hours ahead
• Next step: full 4D-Var
2
1
30°N
0.5
0.25
100°W
90°W
Reading, UK
80°W
Janiskova et al. (2012)
© ECMWF 2013
EarthCARE: UK Involvement
• Lead European Scientist
– Professor Anthony Illingworth (University of Reading)
• Development of synergy algorithms
– Professor Robin Hogan (ECMWF and University of Reading)
• Development of Doppler radar simulator and radar retrievals
– Dr Alessandro Battaglia (University of Leicester)
• Prime contractor
– Astrium UK
• Multi-Spectral Imager
– Surrey Satellite Technology Ltd. (Sevenoaks)
• Broad-Band Radiometer
– SEA, Bristol
• Thermodynamic data for for EarthCARE retrievals;
Real-time assimilation of EarthCARE data
– European Centre for Medium Range Weather Forecasts (Reading)
Key pre-launch scientific questions
• How can we exploit the new radar information for better rain
and snow retrievals?
– High latitude and mountain snowfall rates are very important but
uncertain. Radar reflectivity from space ambiguous: riming leads to
uncertain snow density. Rimed snow falls faster so constrained by
Doppler observations but pre-launch validation urgently needed.
– Extra constraint on rain rates available from Doppler, path-integrated
attenuation from the ocean reflectivity and the 94-GHz brightness
temperature (using radar as a radiometer). How do we best combine
this information as well as accounting for multiple scattering?
• How can we exploit new radiative closure information?
– Balance of longwave and shortwave radiative effect of ice clouds very
important for climate but sensitive to uncertain ice scattering
properties. How can we exploit EarthCARE (particularly lidar and
imager) to constrain assumptions on ice scattering in models?
– Long-standing error in marine cumulus in models is that they are too
few and too bright. How can we use EarthCARE’s 3D cloud
reconstructions to solve this problem?
• Currently no pre-launch funding to address these questions!
Thank you!
Doppler in convective clouds
Ice fall speed:
important uncertainty
in global models
Convective updrafts:
first time they will be
sampled globally
− NASA ER2:
high flying
aircraft
Atmospheric Lidar (ATLID)
EarthCARE Payload
– 355 nm (UV) with depolarization channel
– High spectral resolution capability providing
direct cloud/aerosol extinction measurements
Broadband
Radiometer (BBR)
– Long-wave & totalwave
– 3 views to get
fluxes
Cloud Profiling Radar (CPR)
–
–
–
–
94 GHz
2.5-m dish
Doppler capability
Min detectable signal -35 dBZ
Multispectral Imager (MSI)
– 4 solar and 3 thermal infrared channels
– 150-km swath
24/01/2013
ECARE overview KNMI
Single
Raw
instrument
measurements
products
EarthCARE products
Lidar
Radar
Imager
BB Radiometer
ATLID Level 1
Attenuated backscatter in
• Rayleigh channel
• Co-polar Mie channel
• Cross-polar Mie channel
CPR Level 1
Radar reflectivity profile,
Doppler profile
MSI Level 1
TOA radiances for 4 solar
channels, TOA brightness
temperatures for 3 thermal
channels
BBR Level 1
TOA long-wave and total-wave
radiances
ATLID Level 2
Feature mask and target
classification, extinction,
backscatter and depolarization
profiles, aerosol properties, ice
cloud properties
CPR Level 2
Radar echo product, feature
mask, cloud type, liquid and ice
cloud properties, vertical
motion, rain and snow
estimates
MSI Level 2
Cloud mask, cloud microphysical parameters, cloud top
height, aerosol parameters
BBR Level 2
Unfiltered TOA short-wave and
long-wave radiances,
TOA short-wave and longwave fluxes
Synergy
products
Synergistic Level 2
Cloud and aerosol products
derived from synergistic
retrievals using combinations
of ATLID, CPR, MSI
Radiation and
closure
products
Radiative Transfer Products
Fluxes, heating rates, TOA
radiances calculated from
constructed 3D cloud-aerosol
scenes (1D & 3D rad. transfer)
Assessment
Comparison of Radiative
Transfer Products (fluxes,
radiances) to BBR radiances
and fluxes
Evaluation of model ice clouds using A-train retrievals
Gridbox-mean ice water content
In-cloud mean ice water content
•
•
•
Both models lack high cirrus; Met Office has too narrow a distribution of in-cloud IWC
Using this work, ECMWF have developed a new scheme that performs better
Ice water content and particle size will be considerably more accurate from EarthCARE
Delanoe et al. (QJRMS 2011)

similar documents