Lidar Measurements: Atmospheric Constituents ...
239
better suited for horizontal than for vertical profiling. As aerosol has a tendency to change
with moisture conditions, backscatter coefficients also change with moisture, and usually do
so differently for different laser lines. CO2 laser pulses show a long (1 fJ-s) tail that is hard
to correct for if good depth resolution is required. Finally line strengths are temperaturedependent, making a simultaneous measurement of the temperature profile desirable.
Therefore the visible and near-visible infrared wavelength range was also investigated. Browell et al. (1979) used a ruby-laser-pumped dye laser along with the ruby fundamental. Later
Nd:YAG-laser-pumped dye lasers (Browell et al., 1981; Cahen et al., 1982; Ehret and Renger,
1988; Ehret et al., 1993), excimer-pumped dye lasers (Bosenberg et al., 1990), as well as alexandrite (Higdon et al., 1993) and titanium-sapphire lasers (Wulfmeyer et al., 1995) were all
used for the purpose.
The temperature and pressure dependence, Le., the height dependence of parameters such as
the wavelength and cross section of the on-resonance line, the Rayleigh Doppler broadening
of the backscatter signal, possible shift and broadening of laser emission lines as well as laser
spectral impurity all had to be investigated (Schotland, 1974; Measures, 1977; Zuevet al., 1983;
Ansmann, 1984; Brown, 1985; Zuev et al., 1985; Bosenberg, 1985; Ansmann and Bosenberg,
1987; Ansmann, 1989) and have made DAS lidar moisture measurements a challenging venture
up to our days.
10.4.2 Raman lidars
It was therefore templing to try the Raman lidar concept for moisture determination. First
measurements were carried out with frequency-doubled ruby (Melfi et al., 1969) and frequencyquadrupled Nd:YAG lasers (Renaut et al., 1980). Frequency-doubled and frequency-tripled
Nd:YAG lasers were then used because their emission wavelengths provide better sensitivity or
range (Bukin et al., 1985; Melfi and Whiteman, 1985; Vaughan et al., 1988). A big step forward
was made when excimer lasers came into existence; KrF (Cooney et al., 1985; Eichinger et al.,
1994) and XeCl (Weitkamp et al., 1992; Bisson and Goldsmith, 1993) have remained the most
popular lasers until now.
Today double-laser Raman lidar systems (McGee et al., 1993; Reichardt et al., 1995) provide
the best results because of their capability to measure ozone, temperature, extinction and
backscatter in addition to the water vapor mixing ratio and because they can apply the novel
principle of Raman DIAL, a differential-absorption technique not based on Rayleigh-Mie, but
on Raman scattering in the atmosphere.
Clearly, Raman cross sections which are::; 10- 33 m 2 sr- 1 require powerful lasers, high pulse
repetition rates, large receiver telescopes and sensitive detectors. Skylight interference makes
the use of narrow fields of view and of appropriate filters or other wavelength-selective and
wavelength-blocking devices mandatory. The geometric adjustment is time-consuming, particularly for the low-intensity water vapor channel. The best choice of the wavelength (cf.
Table 10.7) is difficult and depends somewhat on the purpose of the measurement: At long
wavelengths the ),-4 dependence of Raman scattering results in small signals superimposed by
intense skylight, at short wavelengths tropospheric ozone reduces the range so drastically that
no reasonable range can be obtained. It appears to be today generally accepted that except for
studies of the lowest part of the troposphere, wavelengths between 308 and 355 nm are best
suited for moisture profiling with Raman lidar.
Apart from these difficulties, Raman moisture !idar offers a number of advantages over other
lidar schemes. For Raman lidar only one laser is needed. Its wavelength, width, and stability are
uncritical. The geometric alignment is not critical once the polychromator is properly adjusted.
239
better suited for horizontal than for vertical profiling. As aerosol has a tendency to change
with moisture conditions, backscatter coefficients also change with moisture, and usually do
so differently for different laser lines. CO2 laser pulses show a long (1 fJ-s) tail that is hard
to correct for if good depth resolution is required. Finally line strengths are temperaturedependent, making a simultaneous measurement of the temperature profile desirable.
Therefore the visible and near-visible infrared wavelength range was also investigated. Browell et al. (1979) used a ruby-laser-pumped dye laser along with the ruby fundamental. Later
Nd:YAG-laser-pumped dye lasers (Browell et al., 1981; Cahen et al., 1982; Ehret and Renger,
1988; Ehret et al., 1993), excimer-pumped dye lasers (Bosenberg et al., 1990), as well as alexandrite (Higdon et al., 1993) and titanium-sapphire lasers (Wulfmeyer et al., 1995) were all
used for the purpose.
The temperature and pressure dependence, Le., the height dependence of parameters such as
the wavelength and cross section of the on-resonance line, the Rayleigh Doppler broadening
of the backscatter signal, possible shift and broadening of laser emission lines as well as laser
spectral impurity all had to be investigated (Schotland, 1974; Measures, 1977; Zuevet al., 1983;
Ansmann, 1984; Brown, 1985; Zuev et al., 1985; Bosenberg, 1985; Ansmann and Bosenberg,
1987; Ansmann, 1989) and have made DAS lidar moisture measurements a challenging venture
up to our days.
10.4.2 Raman lidars
It was therefore templing to try the Raman lidar concept for moisture determination. First
measurements were carried out with frequency-doubled ruby (Melfi et al., 1969) and frequencyquadrupled Nd:YAG lasers (Renaut et al., 1980). Frequency-doubled and frequency-tripled
Nd:YAG lasers were then used because their emission wavelengths provide better sensitivity or
range (Bukin et al., 1985; Melfi and Whiteman, 1985; Vaughan et al., 1988). A big step forward
was made when excimer lasers came into existence; KrF (Cooney et al., 1985; Eichinger et al.,
1994) and XeCl (Weitkamp et al., 1992; Bisson and Goldsmith, 1993) have remained the most
popular lasers until now.
Today double-laser Raman lidar systems (McGee et al., 1993; Reichardt et al., 1995) provide
the best results because of their capability to measure ozone, temperature, extinction and
backscatter in addition to the water vapor mixing ratio and because they can apply the novel
principle of Raman DIAL, a differential-absorption technique not based on Rayleigh-Mie, but
on Raman scattering in the atmosphere.
Clearly, Raman cross sections which are::; 10- 33 m 2 sr- 1 require powerful lasers, high pulse
repetition rates, large receiver telescopes and sensitive detectors. Skylight interference makes
the use of narrow fields of view and of appropriate filters or other wavelength-selective and
wavelength-blocking devices mandatory. The geometric adjustment is time-consuming, particularly for the low-intensity water vapor channel. The best choice of the wavelength (cf.
Table 10.7) is difficult and depends somewhat on the purpose of the measurement: At long
wavelengths the ),-4 dependence of Raman scattering results in small signals superimposed by
intense skylight, at short wavelengths tropospheric ozone reduces the range so drastically that
no reasonable range can be obtained. It appears to be today generally accepted that except for
studies of the lowest part of the troposphere, wavelengths between 308 and 355 nm are best
suited for moisture profiling with Raman lidar.
Apart from these difficulties, Raman moisture !idar offers a number of advantages over other
lidar schemes. For Raman lidar only one laser is needed. Its wavelength, width, and stability are
uncritical. The geometric alignment is not critical once the polychromator is properly adjusted.