(C) 2007 DEP. OF SIGNAL THEORY AND COMMUNICATIONS HD2 HD2 - - 4: 4: Introduction Introduction to to LIDAR ( LIDAR ( laser laser radar) Remote radar) Remote Sensing Sensing Systems Systems Francesc Rocadenbosch Remote Sensing Lab. (RSLAB) Universitat Politècnica de Catalunya http://www.tsc.upc.edu Campus Nord, D4-016, E08034, Barcelona (SPAIN) [email protected]
61
Embed
Introduction to LIDAR (laser radar) Remote Sensing Systems · PDF fileIntroduction to LIDAR (laser radar) Remote Sensing Systems ... Based on their APPLICATION ELASTIC-BACKSCATTER
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
(C) 2007
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS HD2HD2--4:4:
IntroductionIntroduction toto LIDAR (LIDAR (laserlaserradar) Remote radar) Remote SensingSensing SystemsSystems
Francesc Rocadenbosch
Remote Sensing Lab. (RSLAB)Universitat Politècnica de Catalunya
http://www.tsc.upc.eduCampus Nord, D4-016, E08034, Barcelona (SPAIN)
N is the molecule number densityn is the refraction indexλ is the radiation wavelengthδn is the depolarization ratio
( )g
n
ng N
nβ
π=⎟⎟
⎠
⎞⎜⎜⎝
⎛δ−δ+
⎥⎥
⎦
⎤
⎢⎢
⎣
⎡
λ−ππ
≈α3
876361
38
4
222
SEE ALSO: B.A. Bodhaine, N.B. Wood, E.G. Dutton, J.R. Slusser, “On Rayleigh Optical Depth Calculations,” J. Atmospheric and Oceanic Technology 16(11), 1854-1861 (1999).
Reminder: INTERACTION MECHANISMS
1) Rayleigh scattering (molecules, r << λ)
2) Mie scattering (aerosols, r ≈ λ)3) Others: Absorption
4
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
7
LID
AR
REM
OTE
SEN
SIN
GOPTICAL AND TECHNOLOGICAL CONSIDERATIONSOPTICAL AND TECHNOLOGICAL CONSIDERATIONS
RAYLEIGH SCATTERING (i.e., molecular/gas scattering, r << λ)
Reminder: INTERACTION MECHANISMS
1) Rayleigh scattering (molecules, r << λ)
2) Mie scattering (aerosols, r ≈ λ)3) Others: Absorption
( ) ( ) ( )( )zTzPzg
42234 10·6.61109154.2 −−−− λλ+×=α
SOURCE: P. Menéndez-Valdés, “Atmospheric Transmission and Climatic Effects in the Assessment of Atmospheric Losses on an Optical Link Budget.” UPC, UPM, IAC, ONERA, Final Report (A. Comerón, Ed.). ESA contract no. 8131/88/NL/DG, Barcelona, Oct. 1989.
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
8
LID
AR
REM
OTE
SEN
SIN
G
Lidar ratio defined as:
OPTICAL AND TECHNOLOGICAL CONSIDERATIONSOPTICAL AND TECHNOLOGICAL CONSIDERATIONS
MIE SCATTERING (i.e., aerosol/particle scattering, r ≈ λ)
SOURCE: The Infrared and Electro-Optical Systems Handbook,SPIE Press, (1993).
Mie scattering diagram• x= (2π/λ)r= 8, m=1.25+j0.0• i1 and i2 are the ⊥ and || components
( ) ( )( )RRRS aer
aer
Mλ
λ
βα
=λ,
5
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
9
LID
AR
REM
OTE
SEN
SIN
GOPTICAL AND TECHNOLOGICAL CONSIDERATIONSOPTICAL AND TECHNOLOGICAL CONSIDERATIONS
MIE SCATTERINGTypical extinctions and Deirmendjian’s distribution function,
( ) ( )( )1,0exp=γ∞<<−= γα
rbrarrn
Shaping parameters
N: Aerosol numberdensity
W: Aerosol weightdensity
RN, RM are themodal radii fornumber densityand mass, respectively
SOURCE: The Infrared and Electro-Optical Systems Handbook,SPIE Press, (1993).
( ) ( )drrnrQr extaer λπ=α ∫
∞λ ,0
2
Where:
And:
2rQ ext
extπσ
=
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
10
LID
AR
REM
OTE
SEN
SIN
G
OPTICAL AND TECHNOLOGICAL CONSIDERATIONSOPTICAL AND TECHNOLOGICAL CONSIDERATIONS
ATMOSPHERIC OPTICAL COEFFICIENTSConcerning total components note that:
( ) ( ) ( ) ( )RRRR absmolaerλλλλ α+α+α=α
≈ 0
( ) ( ) ( )RRR molaerλλλ β+β=β
Fig. SOURCE: R.T.H. Collis and P.B. Russell, “LidarMeasurement of Particles and Gases by Elastic Backscattering and Differential Absorption,” Chap.4 in Laser Monitoring of the Atmosphere, E.D. Hinkley, Ed., (Springer-Verlag, New York, 1976), pp.71-102.
6
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
11
LID
AR
REM
OTE
SEN
SIN
GOPTICAL AND TECHNOLOGICAL CONSIDERATIONSOPTICAL AND TECHNOLOGICAL CONSIDERATIONS
1 – 5 kmTemperature
Chemical species concentration in the atmosphere (SO2, NO, CO, H2S, C2H4, CH4, H2CO, H2O, N2,
O2 ...)
Ruby (λ =694.3, 347.2 nm)N2 (λ =337 nm)
Nd:YAG (λ =1064, 532, 355 nm)
Raman
1 – 90 km
Chemical species concentration, especially in upper atmosphere (OH, Na, K, Li, Ca Ca+), oil on water
surface, chlorophyll
DyeN2 (λ =337 nm)
NeFluorescence
1 – 5 km
Temperature, pressure
Chemical species concentration (SO2, O3, C2F4, NH3, CO, CO2, HCl, ...
SOURCE: R.M. Measures, Laser Remote Sensing: Fundamentals and Applications,(Krieger, Malabar, Fla., 1992), Chap.4, pp. 146-204.
7
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
13
LID
AR
REM
OTE
SEN
SIN
G
LASER SOURCES VS. WAVELENGTH
OPTICAL AND TECHNOLOGICAL CONSIDERATIONSOPTICAL AND TECHNOLOGICAL CONSIDERATIONS
SOURCE: J. Hecht,Understanding Lasers,IEEE Press.
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
14
LID
AR
REM
OTE
SEN
SIN
G
OPTICAL AND TECHNOLOGICAL CONSIDERATIONSOPTICAL AND TECHNOLOGICAL CONSIDERATIONS
DETECTORS vs. SPECTRAL BANDS
Fig. DETECTORS OF INTEREST IN LIDAR SCIENCE.Spectral dependence of Detectivity, D*(λ) [cm·Hz1/2W-1], for photoconductors (PC) and photodiodes (PD) of interest in LIDAR (0.3 ≤ λ ≤ 10 μm typ.) for different materials.
PMTs PIN, APD Thermal
SOURCE: R.M. Measures, Laser Remote Sensing: Fundamentals and Applications,(Krieger, Malabar, Fla., 1992), Chap. 6, pp. 205-236.
Airborne (helicopter, plane, satellite), mobile (van, truck), or ground-based.
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
16
LID
AR
REM
OTE
SEN
SIN
G
UPC BACKSCATTER LIDARUPC BACKSCATTER LIDAR
ΔR = 7.5 m, Δt = 1 min
LASER RECEIVER SYSTEM SPECSGain medium Nd:YAG Focal length 2 m Configuration Vertical biaxialEnergy 0.5 J/532 nm Aperture ∅ 20 cm System NEP 70 fW·Hz-1/2
Divergence 0.1mrad Detector APD (EGG C30954) Min. Det. Power < 5 nWPulse length 10 ns Net Responsivity 6×101-3×106 V/W Acquisition 20 Msps/12bitPRF 10 Hz Bandwidth 10 MHz Spatial resolution 7.5 m
Fig. SOURCE: R.T.H. Collis and P.B. Russell, “LidarMeasurement of Particles and Gases by Elastic Backscattering and Differential Absorption,” Chap.4 in Laser Monitoring of the Atmosphere, E.D. Hinkley, Ed., (Springer-Verlag, New York, 1976), pp.71-102.
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
12
LID
AR
REM
OTE
SEN
SIN
G
THE LIDAR EQUATION (V)THE LIDAR EQUATION (V)
FURTHER COMMENTS:
1) Assuming a homogeneous atmosphere and ideality systemconditions, the lidar equation takes its simplest form:
( )RRK= P(R) α−β 2exp2
transmittancebackscatter
2) Note the LIDAR optical thickness (COT) and related transmissivity!
( )[ ] ( ) r)dr(RCOTRCOTR)T(R
0,;2exp, λα=−=λ ∫
7
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
13
LID
AR
REM
OTE
SEN
SIN
GTHE LIDAR EQUATION (VI)THE LIDAR EQUATION (VI)
OPTICAL OVERLAP FACTOR (OVF)
The telescope cannot “read” the full atmospheric cross-section illuminated by the laser beam (i.e., does not lie within its FOV)
It is a function of many geometrical and optical parameters of both the laser and telescope.
Fig. SOURCE: Measures (1992).
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
14
LID
AR
REM
OTE
SEN
SIN
G
THE LIDAR EQUATION (VII)THE LIDAR EQUATION (VII)
[ ]R d R + W Rf = y o
222oi δθ −+±−
[ ] ( )[ ]443442143421
positionysizei Rf f = y
−
−Ω−±− δθ
[ ] ( )[ ]4434421
43421 positionysize
i R f W Rf = y
−
−Ω−±− δ0
Atmospheric laser foot-print imaged is (telesc. far-field):
• For optically “clear” atmospheres, the “range-corrected” (R2P), “backscatter” and “extinction” representations look very much alike.
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
18
LID
AR
REM
OTE
SEN
SIN
G
SIGNAL CONDITIONING (I): RECEIV. CHAINSIGNAL CONDITIONING (I): RECEIV. CHAIN
R’V: Net Voltage Responsivity (V/W)Vos: Total system offset (user+drift+background)ntot: Total noise (photodetection + electronic)εq: Quantization noisexa,s: A/Synchronous interferences
R’V Ideal ADC
ntot Vos εq xa+xs
V(R)P(R)
∑+++′= Δtdt
dVVVPRV driftuserdriftBackvOS
unwanted terms
Sampling at fs, detection time τd=1/fs, so that
s
d
fccR
22=
τ≈Δ
)()()()( RxRxnVRLPRRV saqtotosv ++ε+++=
10
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
19
LID
AR
REM
OTE
SEN
SIN
GEXAMPLE OF GOOD SIGNAL CONDITIONINGEXAMPLE OF GOOD SIGNAL CONDITIONING
OTHER PROJECTS• ATLID (from ESA): Similar to LITE (NASA)• ALADIN (from ESA): Wind lidar space-borne sensor• CALIPSO (from NASA-CNES): Aerosol and clouds• ADM-AEOLUS: Wind sensing
LASER RECEIVER SYSTEM SPECSGain medium Nd:YAG Focal length 2 m Configuration Vertical biaxialEnergy 0.5 J/532 nm Aperture ∅ 20 cm System NEP 70 fW·Hz-1/2
Divergence 0.1mrad Detector APD (EGG C30954) Min. Det. Power < 5 nWPulse length 10 ns Net Responsivity 6×101-3×106 V/W Acquisition 20 Msps/12bitPRF 10 Hz Bandwidth 10 MHz Spatial resolution 7.5 m
5) Reencounters the system (atmospheric) impulse response →
• System identification• Demodulation is
substituted by correlation
( ) ( ) ( ) ( )RRTRR
Actg r ξλξβ= 22
12
)(
( ) ( ) ( )( ) ( ) ( )⎩
⎨⎧
∗=
′==tgtxty
taTPtEatx b~~
~0
( ) ( ) ( ) ( )tgtatytg ≈′∗= ~~̂
( )⎩⎨⎧
≠−=
=′′ 001~
1 jj
jRN
aa
( )2
,11 b
baa
NTNTN
NR ≤τ−⎟⎟⎠
⎞⎜⎜⎝
⎛ τΛ
+=τ′′
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
32
LID
AR
REM
OTE
SEN
SIN
G
PSEUDOPSEUDO--RANDOM SYSTEMSRANDOM SYSTEMS
Fig. SOURCE: Bundschuh et al., “Feasibility study of a compact low cost correlation LIDAR using a pseudo-noise modulated diode laser and anAPD in the current mode”, IEEE (1996).
THE ATMOSPHERIC ID. PROBLEM• The impulse excitation is substituted by
OPERATIONAL PRINCIPLE1) In contrast to elastic systems, the return wavelength, λR, is shifted from the incident one, λ0.2) Wavelength shift, κ, depends oneach molecular species.
3) Very faint returns.• requires photon counting• very often, night-time operation
0
0
1 κλ−λ
=λR
Fig. ADAPTED FROM: Inaba, H. Detection of Atoms and Molecules by Raman Scattering and Resonance Fluorescence. In Laser Monitoring of the Atmosphere, Hinkley, E. D., Ed.; Springer-Verlag: New York, 1976; Chap. 5, 153-236.
2
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
3
LID
AR
REM
OTE
SEN
SIN
GRAMAN LIDAR SUMMARYRAMAN LIDAR SUMMARY
APLICATIONS1) Self-calibrated lidar (N2 shift)
• Absolute concentration of anyatmospheric species can be determined by comparison to theN2-atmospheric return
2) Temperature profiler (±2K)
3) Spectroscopic sensing (COMPARISON WITH DIAL) • Low detection sensitivity at long ranges due to the low Raman cross
sections that ...• limit the method to the detection of species present in high concentrations
(e.g. smoke stacks in industrial plants, 100-1000 ppm, 30-100 m).• In contrast, measurements are always range resolved (RR) and there is no
1) does not dependon the excitationwavelength λ0 and,
2) it is specific of thechemical species
A) Laser needs not be tunable
B) The Ramanspectrum ischaracteristic of eachmolecule
C) Conveystemperature info.
Overview of the lidar backscatter signals for 532-nm laser excitation wavelength.
RAMAN SPECTRUM CHARACTERISTICS
Fig. SOURCE: Behrendt, A., et al., “Combined Raman lidar for the measurement of atmospheric temperature, water vapor, particle extinction coefficient, and particle backscatter coefficient”, Appl. Opt. 41 (36), 7657-7666, (2002).
Stokes linesAnti-Stokes
Rayleigh and Mie Scattering
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
6
LID
AR
REM
OTE
SEN
SIN
G
RAMAN LIDARRAMAN LIDAR
Fig. SOURCE: Inaba, H. Detection of Atoms and Molecules by Raman Scattering and Resonance Fluorescence. In Laser Monitoring of the Atmosphere, Hinkley, E. D., Ed.; Springer-Verlag: New York, 1976; Chap. 5, p.162.
[ ]1
0Rcm,11 −κκ−
λ=
λ
KEY CONCEPTS1) Raman components
• Stokes lines– molecule gains energy from the
radiation field– scattered radiation is at λR> λ0
• Anti-Stokes lines (λR<λ0)
2) Motivation for the “wavenumber”concept (with κ, the Raman shift):
3) Raman cross-sections• dependency ∝ λ-4
Common Raman shifts:N 2 2 3 3 1 cm -1 H 2O 3 6 5 4 cm -1
O 2 1 5 5 6 cm -1
( ) ( )Ray
4,3
Raman dd10
dd
Ωπσ
≈Ωπσ −−
4
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
7
LID
AR
REM
OTE
SEN
SIN
GRAMAN LIDAR LINKRAMAN LIDAR LINK--BUDGETBUDGET
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
8
LID
AR
REM
OTE
SEN
SIN
G
TEMPERATURE MEASUREMENT (I)TEMPERATURE MEASUREMENT (I)
Fig. SOURCE: Inaba, H. Detection of Atoms and Molecules by Raman Scattering and Resonance Fluorescence. In Laser Monitoring of the Atmosphere, Hinkley, E. D., Ed.; Springer-Verlag: New York, 1976; Chap. 5, 153-236.
KEYRaman signatures are direct measures of the relative populations among the internal molecular modes
• (In termal equilibrium) →fundamental def. of temperature
METHODS1) Rotational Raman (RR)
• Comparison of the envelope shape of all the lines
• Intensity ratio of selected spectral regions of the band
Suitable for atmospheric profiling
2) Vibrational Raman (VR)• +• Intensity ratio between Stokes
and anti-Stokes components• Width of a specific Q-branch
Suitable for high-temperature diagnostics (e.g. flames)
Fig. SOURCE: Behrendt, A., et al., “Combined Raman lidar for the measurement of atmospheric temperature, water vapor, particle extinction coefficient, and particle backscatter coefficient”, Appl. Opt. 41 (36), 7657-7666, (2002).
RASC (GKSS) lidar1) T (temperature)2) WV (water vapor mixing ratio)3) α, β, SM
4) RH (humidity)
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
10
LID
AR
REM
OTE
SEN
SIN
G
UPC POLYCROMATOR TEST LAYOUTUPC POLYCROMATOR TEST LAYOUTH
Tab. SOURCE: Behrendt, A., et al., “Combined Raman lidar for the measurement of atmospheric temperature, water vapor, particle extinction coefficient, and particle backscatter coefficient”, Appl. Opt.41 (36), 7657-7666, (2002).
DISCUSSION PARAMETERS (RASC lidar LAYOUT)
Note:ND filters are used to cope with saturation effects in the RR channels in the lower troposfere (correction of photon-counter receiver dead-time effects).
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
12
LID
AR
REM
OTE
SEN
SIN
G
TEMPERATURE MEASUREMENT (IV)TEMPERATURE MEASUREMENT (IV)
Fig. SOURCE: Behrendt, A., et al., “Combined Raman lidar for the measurement of atmospheric temperature, water vapor, particle extinction coefficient, and particle backscatter coefficient”, Appl. Opt.41 (36), 7657-7666, (2002).
Specific RR Temperature approaches:
Use of two RR channels with opposite temperature dependency
• + 3rd RR channel (isosbestic point) as reference or,
• combine them to obtain a temperature-indep. reference
• calibrate Q(T) with a radiosonde– find c for minimum temp. variation
Fig. SOURCE: Behrendt, A., et al., “Combined Raman lidar for the measurement of atmospheric temperature, water vapor, particle extinction coefficient, and particle backscatter coefficient”, Appl. Opt. 41 (36), 7657-7666, (2002).
Problem: Elastic cross-talk with the RR channelAction: Calibrate on a cirrus cloud using
( ) ( )( ) ( )zNTN
TNTQElRR
RR
ε−=
1
2
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
14
LID
AR
REM
OTE
SEN
SIN
G
MOLECULAR SPECIES (GAS) DETECTION (I)MOLECULAR SPECIES (GAS) DETECTION (I)
CONCEPTS:1) The absolute concentration of each molecular species can be performed by comparing the Raman backscattered intensity with that of the Raman line from N2 which occupy the same volume.
( ) ( ) ( ) ( ) ( ) ( ) ( )[ ]{ }∫ ξξα+ξα−×⎥⎦
⎤⎢⎣
⎡Ωπσ
λΔ=λλ
λλλ
R tottotRRR d
dd
RNTFR
ROKRPR
RRR 02 0
exp,
1A) Raman-backscattered signal:
1B) (Oversimplified) Gas-to-N2 normalised ratio
( )( )
( )( )
( )( )
( )( )
( )( )
( )( ) ( )
( ) ( ) ( )[ ]{ }∫ ξξα−ξα−=λλτΔ
λλτΔλξλξ
⎥⎥⎦
⎤
⎢⎢⎣
⎡
ΩπσΩπσ
λΔλΔ
=
λλ
λ
λ
λ
λ
R tottotNX
NXN
X
N
X
NN
XX
N
X
dRwhere
Rdddd
RNRN
TFTF
RORO
RPRP
NX
N
X
N
X
0exp,,
,,,,,
N2-normalised cross section
solve for Nx(R) known (US-std model, radiosonde)
8
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
15
LID
AR
REM
OTE
SEN
SIN
GMOLECULAR SPECIES (GAS) DETECTION (II)MOLECULAR SPECIES (GAS) DETECTION (II)
2) The (VR) spectrum is preferred to the (RR)
• RR lines of major atmospheric constituents overlap,
• large Rayleigh-Miecross-talk.
• In contrast, VR cross-sections are usually lower than RR ones.
ERROR SOURCES AND UNCERTAINTIES• Water-Vapor Mixing Ratio Error
• k*: Calibration factor → Can be considered to be small– Calibrated using a radiosonde (e.g. VAISALA RS80-A) or a MW radiometer
• Rw: Signal-induced statistical error dominates the error budget
• = Differential Transmission (DT).– λN and λH experience different amounts of attenuation on their return trips,
caused mainly by Rayleigh scattering. The DT can be calibrated by using1) A radiosonde estimate the molecular number density (i.e, T(z), P(z))2) For hazy atmospheres (DT < 0.9 for τPBL > 2), the N2-Raman channel is
used to estimate the aerosol extinction (typ., λ-1 dependence)See Sec. Inversion of Optical Parameters / Extinction inversion,Note: WV absorbs weakly at λH=660 nm ⇒ λL=355 nm preferred to 532 nm
( )( )RPRPR
RRkw N
Hw
w
R
w
Rkw ww =σ
≈τΔ
σ+
σ+
σ=
σ τΔ ;2
2
2
2
2
2
2*
2
2
2*
aeraerNH λλ αα ,
( )RHN ,,λλτΔ
Fig. SOURCE: Inaba, H. Detection of Atoms and Molecules by Raman Scattering and Resonance Fluorescence. In Laser Monitoring of the Atmosphere, Hinkley, E. D., Ed.; Springer-Verlag: New York, 1976; Chap. 5, p.162.
KEY• Water-vapor mixing ratio (wH2O) + Temperature profile ⇒ RH• Derivation of the RH profile emerges from specific physical refs.1,2
( ) ( )( )zezezRH
w
=
( ) ( ) ( )( ) ( ) ( )[ ]
( )[ ]⎭⎬⎫
⎩⎨⎧
−+−
=+
=273
273exp107.6,622.0
2
2
zTMzTMze
zwzwzP
zeB
Aw
OH
OH
where e(z) is the WV pressure, and ew(z) is the saturation pressure,
– MA=17.84, 17.08 and MB=245.4, 234.2 for T < and > 273 K, respectively.
REFERENCES:1) R.R. Rogers and M.K. Yau, A Short Course in Cloud Physics (Pergamon, New York ,1988).2) R.J. List, ed., Smithsonian Meteorological Tables (Smithsonian Institution, Washington, D.C., 1951).
12
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
23
LID
AR
REM
OTE
SEN
SIN
GA COMPLETE RAMAN SYSTEMA COMPLETE RAMAN SYSTEM
Fig. SOURCE: Matthis, I, Ansmann, A et al., “Relative-humidity profiling in the troposphere with a Raman lidar”, Appl. Opt. 41 (30), 6451-6462, (2002).
LAYOUT
• 3 unshifted returns (1064, 532, 355 nm), NO polarization• 4 returns (Stokes and anti-Stokes portions) of the N2 RR spectrum• 3 vibrational Raman returns (N2 at 387, 607 nm and H2O at 407 nm)• 2 returns from the parallel and cross-polarized unshifted 532 nm
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
24
LID
AR
REM
OTE
SEN
SIN
G
A COMPLETE RAMAN SYSTEMA COMPLETE RAMAN SYSTEM
Fig. SOURCE: Matthis, I, Ansmann, A et al., “Relative-humidity profiling in the troposphere with a Raman lidar”, Appl. Opt. 41 (30), 6451-6462, (2002).
COMPOSITE OUTPUTS
Radiosounding calibration
13
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
25
LID
AR
REM
OTE
SEN
SIN
GINVERSION OF OPTICAL PARAMETERS (I)INVERSION OF OPTICAL PARAMETERS (I)
PRINCIPLE1) Two receiving channels: Besides the ELASTIC receiver, which is onlysensitive to the elastic return, another receiver -i.e., the RAMAN receiver-is spectrally tuned to the Raman-shifted wavelength (Q-branch) of anyabundant species of known relative concentration (usually N2).
2) From:• radiosoundings or• ground-level measurements of pressure and temperature +
assumption of a standard atmosphere,the N2 concentration -as a function of the range to the lidar- is inferred.
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
26
LID
AR
REM
OTE
SEN
SIN
G
INVERSION OF OPTICAL PARAMETERS (II)INVERSION OF OPTICAL PARAMETERS (II)
[1] Clifford, S. T.; Kaimal, J. C.; Lataitis, R. J.; Strauch, R. G. Ground-Based Remote Profiling in Atmospheric Studies: An Overview. Proc. IEEE 1994, 82 (3), 313-355.
[2] Henderson, S. W.; Hale, C. P. Tunable single-longitudinal-mode diode laser-pumped Tm,Ho:YAGlaser. Appl. Opt. 1991, 29 (12), 1716-1718.
[3] Huffaker, R. M.; Hardesty, R. M. Remote Sensing of Atmospheric Wind Velocities Using Solid-State and CO2 Coherent Laser Systems. Proc. IEEE 1996, 84 (2), 181-204.
[4] Roux, R. Cooperative ventures monitor atmospheric conditions. Laser Focus World 1994, 30 (8), S7-S9.
2
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
3
LID
AR
REM
OTE
SEN
SIN
GWIND LIDAR SUMMARYWIND LIDAR SUMMARY
Uses airborne particles&molecules as “tracers” along with –usually- the Doppler principle to invert the wind radial component
• (1992) First commercial system. Specs.: 30-3000 m range, 1-m/s resolution, 150-m spatial resolution and 5-min integration time.
• (Today) Wind sensors: LAWS (NASA) and ALADIN (ESA), ...(NOAA).• A few systems rely on correlation techniques instead
– Aerosol and molecular motioninside the scattering volume
– Rayleigh and Mie peaks
3
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
5
LID
AR
REM
OTE
SEN
SIN
GCOHERENT DOPPLER LIDARCOHERENT DOPPLER LIDAR
DESIGN POINTS OF CONCERN:1. Photomixer efficiency requires:
• Precise alignment between local oscillator and return signalbeams– Tool: BPLO (back-propagated
oscilator)• Diffraction-limited optics• Emission laser with good phase
coherence– Longitudinal-mode operation– Long pulses (large coherence
length)• Return signal spot must be
coherent across most of itstransversal section– Van Cittert-Zernicke theorem
2. Refractive turbulence effects• Phase and amplitude distortion• Degrades signal coherence
SOURCE: With contributions from A. Rodríguez and A. Belmonte (UPC)
FIG: SOURCE: B. J. Rye and R. G. Frehlich, "Optimal truncation and optical efficiency of an apertured coherent lidar focused on an incoherent backscatter target," Appl. Opt. 31, 2891- (1992).
Assume the target area is illuminated from both thetransmitter and the BPLO (“Feuilleté model”)
ReceivingArea (Ar)
dΩrA
d2λ
≈Ω
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
6
LID
AR
REM
OTE
SEN
SIN
G
COHERENT DOPPLER LIDARCOHERENT DOPPLER LIDAR
ad
sc
λ≈
θλ
=ρ 61.022.1
1.22 0.61cs
da
λ λρθ
≈ =
SOURCE: With contributions from A. Rodríguez and A. Belmonte (UPC)
DESIGN POINTS OF CONCERN:1. Photomixer efficiency requires:
• Precise alignment between local oscillator and return signalbeams– Tool: BPLO (back-propagated
oscilator)• Diffraction-limited optics• Emission laser with good phase
coherence– Longitudinal-mode operation– Long pulses (large coherence
length)• Return signal spot must be
coherent across most of itstransversal section– Van Cittert-Zernicke theorem
2. Refractive turbulence effects• Phase and amplitude distortion• Degrades signal coherence
Incoherently illuminated pinholeSpatially coherent radiation is obtained within radius ρ,
The coherent transverse radius of a backscattered signal spot from a roughtarget illuminated by a Gaussian beam ofdiameter D=2a is
DR
cλ
=ρ
4
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
7
LID
AR
REM
OTE
SEN
SIN
GCOHERENT DOPPLER LIDARCOHERENT DOPPLER LIDAR
Turbulence/Amplitude effects
Turbulence/Phase effects
SOURCE: With contributions from A. Rodríguez and A. Belmonte (UPC)
DESIGN POINTS OF CONCERN:1. Photomixer efficiency requires:
• Precise alignment between local oscillator and return signalbeams– Tool: BPLO (back-propagated
oscilator)• Diffraction-limited optics• Emission laser with good phase
coherence– Longitudinal-mode operation– Long pulses (large coherence
length)• Return signal spot must be
coherent across most of itstransversal section– Van Cittert-Zernicke theorem
2. Refractive turbulence effects• Phase and amplitude distortion• Degrades signal coherence
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
8
LID
AR
REM
OTE
SEN
SIN
G
Conical scanning [θ,φ(t)], with zenith angle, θ, constant.Radial wind speed component along thelidar LOS at any time t is given by
( )[ ] ( )( )[ ] θ+φ−φθ=
=⋅=φcoscossin 0 wtv
strtv
H
rrr
VAD representation:Plot vs. the azimuth angle, φ,
⎪⎩
⎪⎨
⎧
θ=θφ=θφ=
seccscsincsccos
0
0
OSAwAvAu
( ) OSAAVAD +φ−φ= 0cos
VAD (VERTICAL AZIMUTH DISPLAY)VAD (VERTICAL AZIMUTH DISPLAY)
Clifford, S. T.; Kaimal, J. C.; Lataitis, R. J.; Strauch, R. G. Ground-Based Remote Profiling in Atmospheric Studies: An Overview. Proc. IEEE 1994, 82 (3), 313-355.
( )[ ]tvr φ
Wind components derived from amplitude, A, and offset, Aos
MAIN LIMITATION• Temporal coherence (i.e., the time
that the atmospheric scatterers take touncorrelate themselves from and initialstate of “known” phase)
Doviak and Zrnic’s estimation
where σv is the velocity std (1 m/s).CONSEQUENCE
• Uncorrelated returns (τc << PRT)• fd canNOT be estimated as the rate
of change of the phase betweensuccessive return pulses,
vc σ
λ≈τ 1.0
( )dt
tdfdφ
π=
21
SIGPRO GOAL– Range-resolved vr estimate
METHODSSpatial windowing
• Estimate the Doppler shift withineach range “gate” (τwin).
Digital spectral-peak estimators• Basics:
Periodogram/autocorrelation• Periodogram, AR time series,
Capon estimator, Poly-pulse pairErrorbars: Cramer-Rao bound for
covariance estimatorsKey trade-off:
2;
4maxτ
=Δλ
<ΔΔcRcvR r
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
10
LID
AR
REM
OTE
SEN
SIN
G
Wind measurement example using a Doppler lidar at Eldorado Canyon during a mesofront invasion.
SOURCE: Courtesy of NOAA (National Oceanics and Atmospherics Administration).
COHERENT DOPPLER LIDARCOHERENT DOPPLER LIDAR
6
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
11
LID
AR
REM
OTE
SEN
SIN
GINCOHERENT DOPPLER LIDARINCOHERENT DOPPLER LIDAR
(or DIRECT DETECTION DOPPLER LIDAR)
EDGE TECHNIQUE (ET)• A Fabry-Perot or etalon used as frequency-to-amplitude transducer• fd is estimated by measuring the transmission change• Velocity accuracy → sprectrumsharpness → Mie’s peak is used
DOUBLE-EDGE (DET)• Two etalons symmetrically located around the REF laser line• Separation of aerosol/molecular Doppler returns
1)Shape-invariant pattern with time2)Uniform speed
METHOD 1: αº-width scanningAt sweep times t1, t2, we measure
( ) ( )2211 , tvrSStvrSS rrrr −=−=
Pattern-matching method:
( ) ( )∫ Δ+=ΔASS dArrSrSrR rrrr
21)(21
is maximum for ( )12 ttvropt −=Δ rr
Limitations: Decorr./ Low-scan speedsE.g. Aerosol patterns (mean ∅= 50m) advected at v=10 m/s would exist forapprox. D/v = 5 sScan (90º at 1º/step, 1s/step) = 90s!
METHOD 2: 3-azimuth scanTime-space lagging of
LOS range, Ri
φ1 φ2
( )nm tRSi
,φ
FIG. SOURCE: S. Tomás et al., “A wind speedand fluctuationsimulator forcharacterizingthe wind lidarcorrelationmethod,” Proc. IEEE IGARSS 2007, in press.
1
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
Chap.5. Other Laser Radar Chap.5. Other Laser Radar SystemsSystems
Francesc Rocadenbosch
Remote Sensing Lab. (RSLAB)Universitat Politècnica de Catalunya
OPERACIONAL PRINCIPLE• DIAL (Differential Absorption Lidar)• Uses two (or more) tuning wavelengths, one of which is absorbed by the
atmospheric species of interest, and another one that is not.
( )( )( )RPRP
RN
aaa
λ′
λ
σ−σ′≈ ln
21
where:Na is the molecule concentration, are the molecule absorption cross-sections at and, are the backscattered return powers at , normalised to the transmitted ones.
aa σ ′σ ,λ ′λ ,
λ ′λ PP ,λ ′λ ,
Fig. Contours of NO2 concentration (ppm) in the vicinity of a chemical plant, as measured by differential absorption lidar. (SOURCE: K. W. ROTHE et al. 1974. Appl. Phys. 4, 181 (1974)).
2
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
3
LID
AR
REM
OTE
SEN
SIN
GDIAL SUMMARYDIAL SUMMARY
APLICATIONS1) Concentration of chemical species in the atmosphere, car exhausts, refineries,...Measurement types:
HCl, vapor H2O, NO, N2O, SF6Typ. Resolutions: ppb to ppm. Typ. Ranges: a few kms.
2) Temperature and humidity
Fig. SOURCE: Whiteman, D. N.; Melfi, S. H. Cloud liquid water, mean droplet radius and number density measurements using a Raman lidar. J. Geophys. Res. 1999, 104 (D24), 31411-31419
Fig. SOURCE: INERIS - DRC-AIRE-00-23443-n° 535 Annexes A-C in www.lcsqa.org/rapport/rap/prog2000/ ineris/annexec_lidar_evaluation.pdf(Accessed June 2004).
Fig.1 Selection of λon and λoffwavelengths for toluene measurement.
Fig.2 Pollutant gas measurement sensitivity
Fig. 1
Fig. 2
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
10
LID
AR
REM
OTE
SEN
SIN
G
DIALDIAL
Fig. SOURCE: Well Test Flare Plume Monitoring–Literature Review. Report CCT-P 016.01, Carbon and Energy Management. Alberta Research Council Inc., Alberta (Canada). Dec. 2001.
SOME MEASUREMENT EXAMPLES
Fig. 1. NO2 horizontal emission in Geneva (Ref. Elight, GmbH)
Fig. 2-3. Methane Plume 130 m Downwind of Ship Loading Vent (CH4 concentration from 0 to 17 ppmv, Ref. Spectrasyne Ltd.)
• Two operation modes– (Mode 1) 2 of the 35 channels are calibrated for ambient sea-water fluorescence– (Mode 2) System is switched to the classification mode (full OMA activated)
• Classification method (3 main groups)– (Mode 1) Peak + fluorescence conversion efficiency at λcal1,2
Fig. SOURCE: Measures (1992); R.M. Measures, "Laser Remote Sensing. Fundamentals and Applications". John Wiley & Sons, 1984. (Reprint de 1992, Krieger Publishing Company).
9
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
17
LID
AR
REM
OTE
SEN
SIN
GFLUORESCENCE LIDARFLUORESCENCE LIDAR
Fig. SOURCE: Canada Center for Remote Sensing (CCRS) & Measures (1992).
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
18
LID
AR
REM
OTE
SEN
SIN
G
FLUORESCENCE LIDARFLUORESCENCE LIDAR
Fig. SOURCE: Measures (1992); R.M. Measures, "Laser RemoteSensing. Fundamentals and Applications". John Wiley & Sons, 1984. (Reprint de 1992, Krieger Publishing Company).
BETTER CLASSIFICATION APPROACH:
Fluoresc. Decay Spectroscopy (FDS)MOTIVATION
• Modest classification: 3 types• More channels: Cost↑ and
SNR↓MEASUREMENT KEY
• Fluoresc. decay time as a function of λ
• is a spectral fingerprint of materials that allows fine discrimination.
10
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
19
LID
AR
REM
OTE
SEN
SIN
GOTHER LASER RADAR SYSTEMS: PART 1OTHER LASER RADAR SYSTEMS: PART 1
Fig. SOURCE: Measures (1992); R.M. Measures, "Laser Remote Sensing. Fundamentals and Applications". John Wiley& Sons, 1984. (Reprint de 1992, Krieger Publishing Company).
BATHYMETRYIt’s hydrographic lidar.MOTIVATION
• IR and MW radiation have negligible penetration in water
• Uses the blue-green “window on sea”
KEYS• Sounding Depth and bathym.
lidar equation depends on– αabs/αsca
– two-way (i.e., air-water and water- air) transmission factor
– beam spreading and diffusion on water medium → multipathattenuation (αmp)
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
20
LID
AR
REM
OTE
SEN
SIN
G
OTHER LASER RADAR SYSTEMS: PART 1OTHER LASER RADAR SYSTEMS: PART 1
THE SHOALS (Scanning Hydrographic Operational Airborne Lidar Survey) PROJ.
• Airborne– 400 Hz system– Collects 400 soundings/s and
every 4 m (variable spot)– Equals 16 km2/h
• Ground-based processing system
– depth-extraction algorithm (NOAA)
Sect. SOURCE: J.L. Irish, J.K. McClung, W.J. Lillycrop, “The SHOALS System”, Joint Airborne Lidar Bathymetry Technical Center of Expertise, US Army Engineers District in http://shoals.sam.usace.army.mil/pdfFig. SOURCE: Measures (1992).
11
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
21
LID
AR
REM
OTE
SEN
SIN
GOTHER LASER RADAR SYSTEMS: PART 1OTHER LASER RADAR SYSTEMS: PART 1
SHOALS APPLICATIONS(Fig. 1) Tidal inlet in Lake Worth (Flda).
• Quantify channel dredgingrequirements and nearshoreconditions
• Volumetry– Calculate sediment volumes for
navigation and nourishment projects
(Fig. 2) Tidal inlet in Long Island (NY).• Reveal the depth and extent of the
scour hole• Comparison with historical data
3-h SHOALS survey = Sveral days with a single-beam acoustic system!
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
22
LID
AR
REM
OTE
SEN
SIN
G
OTHER LASER RADAR SYSTEMS: PART 1OTHER LASER RADAR SYSTEMS: PART 1
SHOALS - Coastal Mapping - Port Huron (Lake Huron, Michigan)
• Navigation charts
MORE APPLICATIONS...• Sediment processes• Shoaling and dredging at
the port• Flux modelling
12
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
23
LID
AR
REM
OTE
SEN
SIN
GOTHER LASER RADAR SYSTEMS: PART 1OTHER LASER RADAR SYSTEMS: PART 1
SHOALS - Coastal Mapping - Solander Island, New Zealand• Very fine resolution as compared to acoustic survey vessels
• Need to update outdated navigation charts
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
24
LID
AR
REM
OTE
SEN
SIN
G
OTHER LASER RADAR SYSTEMS: PART 2OTHER LASER RADAR SYSTEMS: PART 2
OTHER LASER RADAR SYSTEMS: PART 3OTHER LASER RADAR SYSTEMS: PART 3
ACTIVE IMAGING (III):CMS (Cavity Monitoring System)
Volume computation (VCMS) and TOF imaging
15
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
29
LID
AR
REM
OTE
SEN
SIN
GFREEFREE--SPACE OPTICAL COMMUNICATION LINKSSPACE OPTICAL COMMUNICATION LINKS
ADVANTAGES:• Lower mass, weight and volume of TXT/RTX systems• Laser beams: narrower ⇒ higher power densities• No restrictions in the use of frequencies / bandwidths
IGARSS 07, (C) F. Rocadenbosch (RSLAB)
DE
P. O
F S
IGN
AL
THE
OR
Y A
ND
CO
MM
UN
ICA
TIO
NS
30
LID
AR
REM
OTE
SEN
SIN
G
FUTURE TRENDS IN LASER RADARFUTURE TRENDS IN LASER RADAR
CONCERNING:LIDAR SYSTEMS ARCHITECTURE
• System simplification and reliability• Operation in autonomous automated routine regime
TECHNOLOGYCAL TRENDS• Semiconductor diode laser technology• If the application allows it, use of eyesafe lasers (λ > 1,5 µm) and/or
low enough power levels
OTHERS• Multisensor data fusion• Efforts in the methodology of data interpretation