Transcript Lidar for Atmospheric Remote sensing

Lidar for Atmospheric
Remote sensing
Philippe Keckhut et Andrea Pazmiño
LATMOS, Institut Pierre Simon Laplace, CNRS-UPMC-UVSQ, Paris, France
Outline
 Historical Overview
 Lidar Basics
The Lidar Equation
 Lidar systems
 Summary
Historical Overview (1)
 1930 Synge proposed a method to
determine the atmospheric density
with an antiaircraft searchlight and
a telescope (bistatic configuration)
 1936 First reported results of
density profiles: Duclaux (3.4 km),
Hulbert (28 km)
 1938 First reported use of a
monostatic configuration for cloud
base height, using a pulsed light
source (Bureau)
 1953 First retrieval of
temperature profiles from density
profiles (Elterman)
Emitted beam
Detector
field of view
~km
Monostatic
co-axial
Monostatic
bi-axial
Transmitter & receiver collocated
(pulsed light source  range of
scattering)
Bistatic
Receiver’s FOV scanned along
the transmitted beam
(geometry  range of
scattering)
Historical Overview (2)
 1956 Friedland et al. reported
the first pulsed monostatic
system for atmospheric density
measurements
 Early 1960s Invention of laser
 powerful new light source for
lidar systems
 1962 First use of laser in a lidar
system (Smullins & Fiocco)
 1977 First ozone measurements
by lidar (Mégie et al.)
Gérard Mégie
Historical Overview (3)
Present
 Networks of groundbased lidar systems as
NDSC, EARLINET, etc
 Lidars on aircraft
 Space-based lidar
(ALISSA, LITE, …
CALIPSO program)
Lidar Basics (1)
LIDAR: LIght Detection And Ranging
Active remote sensing technique for measuring atmospheric parameters
 (T, , wind and different constituents: H2O, O3, …, clouds, aerosols)
Same principle as radar but 0.1 <  < 10 m
 Principle: emission of a light beam that interacts with the medium &

detection of radiation backscattered towards the instrument
Interactions with the Atmosphere:
elastic (Rayleigh, Mie, Resonance scattering)

inelastic (Raman scattering, Fluorescence)
Absorption
Lidar Basics (2)
Some Optical Interactions of Relevance to Laser Environmental Sensing
Elastic Interactions
Inelastic Interactions
Virtual Level
•  ~ d  Mie = C/a
Rayleigh Scattering
Mie Scattering
•  >> d  Rayleigh = C/4
•  ~ d  Mie = C/a
Vibrationally
Excited
Level
Raman Scattering
• Interaction with the quantized vibrational
& rotational energy levels of the molecule
+ = Differential Absorption & Scattering
Absorption
Lidar Basics (3)
Block diagram of a generic lidar system
Emitted light
Laser
Beam expander
(optional)
Transmitter
Backscattered light
Light
collecting
telescope
Optical
filtering
Receiver
Synchronization
control
Electrical
recording
system
Optical to electrical
transducer
Detector & Recording
Lidar Basics (4)
Ranging of pulsed monostatic lidar
Laser beam
Altitude
Z2
Aerosol
layer 2
Z1
Aerosol
layer 1
Each light pulse fired

tup
Scattered
light
Complete altitude scattering profile
tdown
Time
Emission
impulsion
tup + tdown = T1=2.Z1/C
Signal
Noise
level
Rayleigh
scattering
T2=2.Z2/C
Mie
scattering
z = ct/2  1 s 150 m
Time
The lidar equation (1)
 Number of photons detected by a lidar system
o instrumental parameters
o geophysical variables
 If Ne is the total number of photons emitted by the laser at L
Total number of photons
transmitted into the
atmosphere
Nt  Net t  L 
transmission coefficient of optics (0-1)
 The number of photons available to be scattered at the distance r
Nt Ta (  L , r )
optical transmission of the atmosphere at L
along the laser path to the range r
 The number of photons backscattered, per unit solid angle due to scattering of
type i, from the range interval R1 to R2
R2
Nt  Ta (  L .r )i (  L , r )dr
R1
backsatter coefficient for scattering
of the type i and L
The lidar equation (2)
 Number of photons incident in the collecting optic of the lidar due to
scattering of the type i
R2 1
Nt A  2 ( r )Ta ( s , r )Ta (  L , r ) i (  L , r )dr
R1 r
area of the collecting optic
overlap factor
wavelength of the scattered light
decreasing illuminance of the
telescope by the scattered light
 The number of photons N(s,r) after the detection
R2
1
R1
r
N( , r )  Nt At r ( s )Q( s ) 
transmission coefficient of the
reception optics at s
i

(
r
)
T
(

,
r
)
T
(

,
r
)

(  L , r )dr
a s
a L
2

quantum efficiency of the
detector at s
The lidar equation (3)
 In many cases, approximations allow simplification of lidar equation …
o
L = s 
Ta(L) = Ta(s)
o
integral range  cte, during acquisition (t
o
(r)  1
= 2 (R2-R1)/c)
 Then, the lidar equation …
N( , r )  Nt ATr (  L )Q(  L )
where
r
r2
instrumental dependency
p
  m



atmospheric dependency
Ta (  L , r )  exp[ (  L , r )]

contribution of
molecules & particles
i (  L , r )Ta2 (  L , r )
r
optical depth
  m (  L , r )   p (  L , r )   k (  L , r )n k ( r )dr
0
extinction coefficient of
molecules & particles
k
cross section of
concentration of
constituent k at L constituent k
Rayleigh-Mie Aerosol Lidar (1)
2
Nr  K e
•
•
•
Receiver
2 
contribution of
aerosols
Application of inversion method to the lidar
equation (Klett)  p(,z), p(,z) (hypothesis
on p(,z)/p(,z) )
Polarization technique: measure the
polarization ratio  indication of aerosols
shape (liquid or solid)
Multi-wavelength lidar  Spectral
dependence of aerosol optical thickness
(AOT)
X Measurements of aerosols & clouds in the
troposphere & lower stratosphere
Transmitter
Detector
(polarization technique)
Aerosol and molecular scattering (2)
• Both molecular and aerosol
contribution are present
• Aerosols are identified
through their vertical shape
• Aerosol analysis consists in
estimating
– Molecular contribution
– Aerosol attenuation
Rayleigh-Mie Aerosol Lidar (3)
Measurements in the troposphere
ln(Nr2)
10000
10000
Altitude [m]
15000
Altitude [m]
15000
5000
0
8
9
10
11
12
Time [UTC]
13
14
15
(Pietras et al., 2004)
flag
5000
0
8
9
10
11
12
Time [UTC]
13
14
15
Temporal evolution of lidar signal at 532 nm (linear polarization component) corrected in distance [ln(Nr 2)] for April 1st 2003, (left panel). Classification
of the atmospheric layers: noise (flag 0), zone with molecules (flag 1), ABL (flag 2), zone with particles (flag 3 & 4), and indefiended zones (flag > 4)
•
•
•
•
•
Transmitter: Nd:YAG at 532 nm (second harmonic) & linear polarization + expander
Receiver: 2 telescopes (0.1-7 km & 2-15 km)
Detection: 532 nm linear & cross polarization components par PMT, 1064 nm par avalanche photodiodes
Vertical resolution: 15 m & temporal resolution: 30’
Classification of the atmospheric structure from backscattering lidars signals corrected from noise & total overlapping:
(Identification of atmospheric boundary layer (ABL), the zones with particles (aerosols & clouds) & finally the zones with molecules)
Multiwavelenght lidar (4)
Médiane du nuage filtré
Minimum de la fonction de coût
No = 7.71 cm-3
rm = 0.29 µm
σ = 1.45
Size distribution ->
Aerosol surface and volume
Temperature measurements (5)
• Required pure molecular
scattering
• Density and pressure are
relative measurements
• Temperature is absolute
 (z)  f (N(z)
dP(z)  g (z)dz
90
85
80
75
MP(z)
T(z) 
R (z)
z
T(z) 
g ( )z

M
0
R
 (z)
Altitude (km)
70
z

N( )z

Mg
0
R
N(z)
65
60
55
50
OHP 5 Dec 1991
17:38-03:21
45
40
35
30
180
200
220
240
Temperature (K)
260
280
Rayleigh Lidar (6)
Temperature measurements
(Alpers et al., 2004)
(a) (background corrected) raw lidar backscatter profiles with the Rayleigh/Mie/Raman (RMR) and Potassium
(K) lidar at Kühlungsborn, Germany on 23 February 2003. (b) Temperatures profiles retrieved from (a)
•
•
•
•
•
•
Transmitter: Nd:YAG at 532 nm (second harmonic) & 355 nm (third harmonic) for T measurements
532 nm high Rayleigh signal : 4 telescopes of 50 cm diameter  40-90 km (blocking chopper at 40 km)
532 nm low Rayleigh signal : 1 telescope of 50 cm diameter  20-50 km (blocking chopper at 20 km)
1 h integration time
Vertical resolution of 1 km & a heigh-variable smooth filter (0.6-3 km width)
Statistical T error < 10 %
To discriminate species:
Raman scattering (7)
• Raman consists in a spectral
shift of the returned wavelength
• Raman shift is characterized by
the molecules considered
• Only attenuation of the bean is
required
• Technique useful for pollution
Raman Lidar (8)






Spectral shift of the returned wavelength (Raman= L   )
Raman shift is characterized by the considered molecules (unique spectral signature)
The vibrational Raman lines are generally selected for detection  concentrations
High-quality of narrow-band interference filters
High-blocking filter for elastic backscatter of molecules & aerosols
Small cross-section of Raman scattering  molecules with a relatively high abundance
(H2O, N2, O2)
X Measurements of temperature using Raman scattering from N2
X Cloud & aerosols can also be studied by this technique
H2O Raman Lidar: q(z) H2O mixing ratio is specified as:
qz  

n H 2O ( z )  M H 2O


 rN2  

n N2 ( z )  M N2

Mass H2O /
Dry air mass
k

T 
Calibration
constant

, z K
N H2O Ta  RamN2 , z K N2  N2
N N2
Lidar Raman
signal for
nitrogen &
water vapor
a
Ram H 2O
Atmospheric
transmission at Ram
H 2O
 H 2O
k
Differential Raman
backscattering cross
sections for water
vapor & nitrogen
H2O Raman: Calibration (9)
H2O at 660 nm
N2 at
607 nm
Filters for
Rayleigh rejection
At 532 nm
Beam spliter
2 channels: H2O and N2
Sky background calibration
during daytime (SZA=60°)
Raman Lidar (10)
Water vapor measurements in the lower troposphere
a)
b)
(Tarniewicz et al., 2003)
29/10/2002
(a) Comparisons with collocated radiosonde. Data are summed over 20 minutes. (b) Height time series for the water vapor mixing ratio for night
29 October 2002. Profiles are summed over 5 minutes. (right column).
•
•
•
•
Vertical resolution is variable from 50 to 500 m in order to maintain a good signal to noise ratio
Good agreement between lidar and RS water vapor mixing ratio below 5 km
Same water vapor structures are seen by the two instruments.
Slight overestimation of lidar profile after 4 km due to an undetermined instrumental bias. Relative precision of
lidar < 5% (up to 2 km) & 10% (up to 4 km)  requirements for boundary layer applications.
Lidar Retrieval for O3 Measurements (11)
DIfferential Absorption Lidar technique for
stratospheric ozone measurements

Measurements of the stratospheric ozone vertical distribution

Two laser wavelengths (on, off) characterized by a different ozone
absorption cross section (UV spectral range, great ozone absorption)

Self calibrating technique, no instrumental constants
nO3
 N( off , z )  N bg ( off , z ) 
1
d
  nO ( z )

 Ln 
3
 N( on , z )  N bg ( on , z ) 
2  O3 ( z ) dz


O3 number
differential O3
density absorption cross-section
O3 (on , z )  O3 (off , z )
number of detected
photons at i
background
radiation at i
correction term
Rayleigh & Mie
differential scattering
Rayleigh & Mie
differential extinction
Absorption by others
constituents (SO2, NO2)
OHP stratospheric ozone DIAL system (12)
Mechanical
chopper
Multiple-fiber collector concept
optical fibers
Beam expanders
Moveable fiber mounts for the
alignment of the XeCl laser radiation
spectrometer
4 Collecting mirrors:  0.53 m,
F 1.5 m, Ap. F/3
Ref. line: 3rd harmonic Continuum
Nd:Yag (355 nm)
Abs. radiation : XeCl Lambda Physics
EMG 200 Excimer laser (308 nm)
Example of ozone profile (13)
Courtesy S. Godin-Beekmann
•
Ozone measurements performed during the night
•
Temporal resolution 3 – 4 hours
•
Require clear skies
Wind measurements (14)
• Wind is based on the
Doppler shift of the
return signal
Limitations (1)
• Dynamic of the signal
: 5-6 orders of
magnitudes
• Emission-reception
geometry
– Parallax
– defocalisation
• Noise and signalinduced-noise
Lidar subsystems (2)
Transmitter sub-system
Strategy of measurement
 Altitude range and concentration to
be detected
 Specific Wavelengths (absorption)
 Energy & repetition rate
 Concentration and spectral
characteristics of other gases
 Reliability, ease of operation in
monitoring applications
 Operating costs

laser
beam expander
choice of the laser source
Common examples:
Gas laser (ex: excimer laser)
optical medium: gas of molecules only stables in
an excited state. Electrical discharge
Solid-state laser (ex: Nd:YAG)
Impurity ions (Nd3+) in a glassy material (YAG).
Optically pumped by a flash lamp  stimulated
emission (1.06 m)
Lidar subsystems (3)
Receiver sub-system
 Collection of scattered laser light back from the
atmosphere and focuses it to a smaller spot
  ~ 10 cm, lenses or mirrors (close range)
  ~ few meters, mirrors (middle & upper atmosphere)
 Processing of scattered laser light
chopper
 spectral filtering schemes: centered in a specific
wavelength (dichroic, gratings, mirrors, narrowband
interference filters  < 1 nm)
 separation based on polarization (aerosols)
 protection of detector (fast mechanical shutter,
electrical gating)
4 fibers
grating
308 nm
high & low energy
387 nm
355 nm
high & low energy
347 nm
332 nm
Lidar subsystems (4)
Signal Detection & Acquisition sub-system
 Conversion of light into an electrical signal & recording in electronic device
 Photomultipliers Tubes (PMTs) are generally used in incoherent lidar systems
(direct detection)
 Output of PMT:
current pulses produced by photons
+
thermal emission of electrons (dark current)
 2 Techniques:
- Photon counting Mode (individual pulses)
- Analog Mode (multitude of pulses)
 Selecting the PMT:
- PMT structure  optical measurement conditions
To the electronic device
Hamamatsu
- Photocathode Quantum Efficiency  high QE in the wavelength range
- Gain  > 106
- Dark count  lower detection limit
- Response time  maximum count rate, time resolution
Lidar subsystems (5)
Signal Detection & Acquisition sub-system
Photo Counting
• Preamplifier  amplification + pulse shape (ringing)
• Main amplifier (if it is necessary)
• High speed comparator (discriminator)  remove
a substantial number of the dark current
• Pulse shape & counter
x Generally for low signals detection
Hamamatsu
Analog Detection
Typical Photon Counting System
• Pulse pair resolution of the detector 10 to 100 MHz
• Fast analog-to-digital converter
x Generally for tropospheric measurements
Coherent Detection
• Mixing of backscattered laser light with light from local oscillator on a photomixer  radio frequency (RF) signal
• Frequency of RF signal  Doppler shift of the scattered laser light  wind velocity
x Frequency stability & short laser pulse length is required
Photon counting (6)
• Measurement = Histogram
• D  1
HV
Cooling system
NbPhotons
• Improvements = increase the
number of collected photons
Discriminator lev el
Photon
PHD

Counter
t
t
–
–
–
–
Size of the telescope
Laser power
Vertical resolution
Temporal resolution