Lidar Measurements: Atmospheric Constituents ...
225
wavelength and is normally on the same order of magnitude as the molecular contribution for
A1 - Ao wavelength differences of a few nanometers.
B corrects for differences in backscatter. Here gradients, not the magnitude of atmospheric
backscatter itself cause noticeable effects. Again the molecular part is not critical, but the
aerosol contribution can be so important that no reliable concentration data can be deduced
from the lidar signals at the onset of a thick cloud with sharp boundary.
G is the gas interference term. It is obvious that if other gases i than the one to be measured show sizeable differential absorption at wavelengths A1 and Ao and can be present in
concentrations high enough so the product of particle density and differential cross section gets
comparable to that of the gas of interest, the effect must be taken into account. The term G can
usually be neglected in measurements of nitrogen dioxide, water vapor and a few other gases in
the visible and near infrared, but plays a role in ultraviolet DAS lidar and in middle-infrared
DAS lidar around 10 jtm. If and only if A1 < Ao then A and B are negative. G is negative
for those interfering gases for which O"(Ad > o"(Ao), as is the case by definition for the gas of
interest.
10.2.3 Raman lidar
Prompt inelastic scattering of light which is characterized by a shift of the scattered radiation
with respect to the primary radiation is called Raman scattering. The shift originates from the
capability of the scattering molecule to absorb part of the energy of the incoming photon
(10.20)
leaving behind a scattered photon of energy
(10.21)
Here Ep and ER are the energies of the primary and Raman-scattered photons, Ap and AR
the corresponding wavelengths, and hand c Planck's constant and the speed of light. 6.E is
the amount of energy absorbed by the molecule. Clearly, 6.E cannot take any value but is
determined by the vibrational and rotational properties of the molecule.
In this context only vibration-rotation transitions and only so-called Stokes scattering, in which
6.E is positive, are considered. It may be mentioned, however, that purely rotational Raman
scattering and anti-Stokes transitions, in which the molecule adds energy to the photon, can
both be used in other types of lidar which aim, e.g., at the measurement of atmospheric temperature profiles (Vaughan et al., 1993; Zeyn et al., 1994).
For Raman lidar only one laser is needed. The radiation collected by the receiver optics contains
a mixture of wavelengths (d. Fig. 10.7) that must be separated by a spectrometer or monochromator. Usually a number of narrow wavelength regions are "watched" simultaneously, the
corresponding optics being called a "channel" and the whole instrument "polychromator".
The signal from distance x that is observed at the Raman wavelength AR if the primary (laser)
wavelength is Ap is given by
P( ' .'
) _ c!::"tR A1)O(x)N( )dO"(7r,AP,AR)
{_ r[ (I:)
( ))dl:}
"p, "R, X - 2 0 x2
x
dl!
exp
Jo Cip ~ + CiR e ~ . (10.22)
225
wavelength and is normally on the same order of magnitude as the molecular contribution for
A1 - Ao wavelength differences of a few nanometers.
B corrects for differences in backscatter. Here gradients, not the magnitude of atmospheric
backscatter itself cause noticeable effects. Again the molecular part is not critical, but the
aerosol contribution can be so important that no reliable concentration data can be deduced
from the lidar signals at the onset of a thick cloud with sharp boundary.
G is the gas interference term. It is obvious that if other gases i than the one to be measured show sizeable differential absorption at wavelengths A1 and Ao and can be present in
concentrations high enough so the product of particle density and differential cross section gets
comparable to that of the gas of interest, the effect must be taken into account. The term G can
usually be neglected in measurements of nitrogen dioxide, water vapor and a few other gases in
the visible and near infrared, but plays a role in ultraviolet DAS lidar and in middle-infrared
DAS lidar around 10 jtm. If and only if A1 < Ao then A and B are negative. G is negative
for those interfering gases for which O"(Ad > o"(Ao), as is the case by definition for the gas of
interest.
10.2.3 Raman lidar
Prompt inelastic scattering of light which is characterized by a shift of the scattered radiation
with respect to the primary radiation is called Raman scattering. The shift originates from the
capability of the scattering molecule to absorb part of the energy of the incoming photon
(10.20)
leaving behind a scattered photon of energy
(10.21)
Here Ep and ER are the energies of the primary and Raman-scattered photons, Ap and AR
the corresponding wavelengths, and hand c Planck's constant and the speed of light. 6.E is
the amount of energy absorbed by the molecule. Clearly, 6.E cannot take any value but is
determined by the vibrational and rotational properties of the molecule.
In this context only vibration-rotation transitions and only so-called Stokes scattering, in which
6.E is positive, are considered. It may be mentioned, however, that purely rotational Raman
scattering and anti-Stokes transitions, in which the molecule adds energy to the photon, can
both be used in other types of lidar which aim, e.g., at the measurement of atmospheric temperature profiles (Vaughan et al., 1993; Zeyn et al., 1994).
For Raman lidar only one laser is needed. The radiation collected by the receiver optics contains
a mixture of wavelengths (d. Fig. 10.7) that must be separated by a spectrometer or monochromator. Usually a number of narrow wavelength regions are "watched" simultaneously, the
corresponding optics being called a "channel" and the whole instrument "polychromator".
The signal from distance x that is observed at the Raman wavelength AR if the primary (laser)
wavelength is Ap is given by
P( ' .'
) _ c!::"tR A1)O(x)N( )dO"(7r,AP,AR)
{_ r[ (I:)
( ))dl:}
"p, "R, X - 2 0 x2
x
dl!
exp
Jo Cip ~ + CiR e ~ . (10.22)