Post on 28-Mar-2015
Blind tests of radar/lidar retrievals: Assessment of
errors in terms of radiative flux profiles
Malcolm BrooksRobin Hogan and Anthony Illingworth
David Donovan and Claire Tinel
Introduction
• First blind test showed that – Both Donovan and Tinel algorithms could retrieve extinction
coefficient very accurately– Effective radius and IWC depend on assumption of habit (i.e.
“density” or the mass-size relationship)
• Second blind test included multiple scattering, molecular scattering and instrument noise:– Reasonable extinction profiles were generally obtained if
multiple scattering was included in the retrieval, otherwise extinction was underestimated
– Retrieval only possible where lidar still has good signal
• What are the radiative implications?• How do these retrievals compare to radar-only?
Blind test 1
(From aggregation study)
• No instrument noise• No multiple scattering• No molecular scattering• High lidar sensitivity• Two versions of each
profile provided, with variable or constant extinction/backscatter ratio “k”, which was not known by the algorithms
Blind test 1:Results 1
• Constant k:– Both Donovan and Tinel
(after modification) algorithms produce highly accurate extinction
• Variable k:– Error in extinction varies
with k, but not unstable
Blind test 1:Results 2
• Effective radius:– Good, but difficult if re >
80 microns because of radar Mie scattering
– Sensitive to particle habit
• Ice water content:– Extinction ~ IWC/re
– Hence if extinction is correct then the % error in effective radius is equal to the % error in IWC
Best case: radar/lidar retrieval
• Excellent extinction, good re if same mass-size relationship is used (otherwise 40% too low)
Mitchell relationship
Francis et al. relationship
Radar only retrieval
Extinction coefficient Effective radius
Best case: radiation calculations• Used Edwards-Slingo radiation code
• Excellent longwave, good shortwave but slight effect of habit and k; better than radar alone
Longwave up
Shortwave up
Clear sky profile
Cloudy profile
Error 20-40 W m-
2 depending on habit and k
Worst case: radar/lidar retrieval
• Radar/lidar extinction excellent, re underestimated
• Extinction poor from radar only
Extinction coefficient Effective radius
Effective radius underestimate
Poor radar-only
retrieval, particularly at cloud top
Worst case: radiation calculations
• Excellent longwave, still good shortwave!• Effective radius not very important?
Longwave up Shortwave up
Error 20-30 W m-2
Blind test 1: Heating rates• Radar/lidar: very accurate• Radar alone: OK but some biases
Best case Worst case
Error due to higher Z here
Blind test 2
(From EUCREX)
• Instrument noise• Multiple scattering• Molecular scattering• True lidar sensitivity• Constant extinction to
backscatter ratio• Note: radar-only
relationships derived using this dataset so not independent!
Donovan retrieval: with multiple scattering
Tinel retrieval: no multiple scattering
Good case: radar/lidar retrieval
• Extinction and effective radius reasonable when use same habit and include multiple scattering
Extinction coefficient Effective radius
Full profile retrieved
Difference between Mitchelland Francis et al.
Good case: radiation calculations
• OLR and albedo good for both radar/lidar and radar-only (but radar-only not independent)
Longwave up
Shortwave up
Mass-size relationship has modest
effect:Error<10
Wm-2
Underestimate radiative effect if
multiple scattering neglected
Poor case: radar/lidar retrievals
• No retrieval in lower part of cloud
Extinction coefficient Effective radius
Wild retrieval where lidar runs
out of signal
Good retrieval at cloud top
Poor case: radiation calculations
• At top-of-atmosphere, lower part of cloud important for shortwave but not for longwave
Longwave up Shortwave up
OLR excellent despite lower
part not retrieved Albedo
underestimated (90 W m-2):
lower part of cloud is
important
Blind test 2: Heating rates
• Heating profile reasonable if full profile retrieved
Best case Worst case
Erroneous 80 K/day heating
No cloud observed so no heating by cloud
here
Sensitivity of radiation to retrievals
• Longwave: easy!– Sensitive to extinction coefficient– Insensitive to effective radius, habit or extinction/backscatter– OLR insensitive to lower half of cloud undetected by lidar
• Shortwave: difficult to get to better than 20 W m-2
– Most sensitive to extinction coefficient– Need full cloud profile to get correct albedo– Some sensitivity to habit and therefore effective radius– Slight sensitivity to extinction/backscatter ratio– Note: not included habit dependence of asymmetry
parameter or single-scattering albedo
Conclusions• Extinction much the most important parameter:
– Good news: this can be retrieved accurately independent of assumption of crystal type
– But need to include multiple scattering in retrieval
• Need to retrieve something when no more lidar:– Switch to radar-only retrieval?– Assign error to both radar/lidar and radar-only retrievals and
produce a consensus value, weighted accordingly?– Must avoid erroneous spikes where lidar loses signal!– Use imager (VIS & IR) synergy to give top-of-atmosphere
radiances and provide a constraint for the retrieval: how would this be incorporated into the algorithms?
– Do SW radiances provide multiple-scattering information?
Scaling the radar-only retrieval
• Where radar/lidar retrieval fails, can we scale the radar-only retrieval to get a seamless join?– Dubious: the profiles are not real but simulated!
Good fit Partial fit