WAARSCHUWINGEN%20GEVAARLIJK%20WEER - PowerPoint PPT Presentation

Title: WAARSCHUWINGEN%20GEVAARLIJK%20WEER


1

LidarRadar ice cloud remote sensing within
CLOUDNET.
D.Donovan, G-J Zadelhof (KNMI) and the CLOUDNET
team With outside contributions from Z. Wang
(NASA/GSFC) D. Whiteman (NASA/GSFC)
2
Introduction

• Background/Rational
• Approaches used in CloudNET
• How we have used data within Cloudnet
• Testing using Raman lidar data (the future)
• Summary

3
Active (lidar/radar) cloud remote sensing
Radar
Lidar
l 3-100mm
l 350-1100nm
Difference in returns is a function of particle
size !!
4
Basic Considerations

• The lidar extinction must first be extracted from
  the lidar signal (or, equivalently, the observed
  lidar backscatter must be corrected for
  attenuation).

Observed signal
Backscatter
Calibration Constant
Extinction

• Ze used to link backscatter and extinction and
  facilitate extinction correction/determination
  process.
• The retrieved extinction (corrected backscatter)
  can then be used to estimate an effective
  particle size and IWC.

5
No Rayleigh No Raman
Must use Klett (Fernald Rayleigh)

• Must estimate extinction at zm(cloud top)
• Very difficult to do directly if one only has IR
  lidar info
• If have Radar then use smoothness constraint on
  derived lidar/radar particle size, or extinction,
  or No.
• But solutions converge if optical depth is above
  1 or so !!

6
Forward inversion is more direct but it is
unstable !!
Forward
Backward
10 error
Radar
Lidar
10 error
In CLOUDNET most lidar data is IR? No
usable Rayleigh signal ?Radar quite helps
7
Effective Particle Size for Ice Crystals
Exact treatment of scattering difficult
(impossible?)However

• Ice particles are large compared to llid (Optical
  scattering regime)
• Ice particles are small compared to lrad
  (Rayleigh scattering regime)

Confirmed using DDA and RT calculations
8
Two main approaches within CNET (for IR lidars)
MS correction
KNMI Approach Chooses BV Such that Variation of
Reff Is minimized
Reff IWC
9
CETP approach uses concept ofNormalized number
density No associatedwith scaled size
distribution
CETP Approach Chooses BV Such that Variation of
No Is minimized
Based on In-situ Aircraft Observations
10
How we have used results within CNET ?
Current approaches Limited by fact that Must have
BOTH Good Radar and Lidar Data even to get
Extinction
11
So coverage is incomplete

• But we still have105s of data points !!
• Good for parameterization development !
• Red?ARM SGP
• BLUE?CNET

12
Can also use limited LidarRadar data as
benchmark to assess accuracy of Radar only IWC
estimates
13
The Future ?

• The super IR celiometers used at CABAUW and
  Chilbolton really are not optimal.
• Ideal is a high power 24/7 Raman system !!
• But that is expensive (but will be coming to
  CABAUW!)
• Settle for 24/7 visible or UV system of good
  sensitivity for ice clouds.

14
Elastic vs Inelastic scattering
15
(No Transcript)
16
A Test Case Using GSFC Raman lidar data and ARM
MMCR.
Test various approaches w.r.t Raman results
Eo ? S-S E1? Force R1 where no cloud. E2?
Minimize derivative of ext E3? Minimize
derivative of Reff E4?Minimize derivative of No
17
MS effects Consistency between approaches ? Can
be accounted for
Signature of MS
18
Comparison of Techniques
19
Comparison of Techniques (In terms of OT)
20
OPTIMAL APPROACH (FOR ELASTIC RAYLEIGH
LIDARS) Combine Methods 13(4) ! Should work
well in thicker Clouds also.
21
Conclusions

• Lidar Radar ice cloud remote sensing is becoming
  mature
• Limitations and strengths of technique becoming
  more understood.
• Increasing body of comparisons with In-Situ
  measurements
• Most useful in CNET for statistical
  parameterization of ice cloud effective radius
  parameterizations and to help estimate accuracy
  of Radar only IWC estimations.
• Ideal is to use Raman Lidar. If this is not an
  option then a good vis/uv lidar with a Radar is a
  good option.

22
Ext vs- Ze
23
If we have Useful Rayleigh above the cloud.
Then (effectively) can find S and Clid so
that The scattering ratio R is 1.0 below and
above cloud
24
If We have good Raman data then
Direct but noisy
Less noisy but indirect
25
Implementation
Cost Eo W1E1 W2 E2 W3E3 W4E4
Eo ? S-S E1? Force R1 where no cloud. E2?
Minimize derivative of ext E3? Minimize
derivative of Reff E4?Minimize derivative of No