Glossary - Surface Based Remote Sensing of the Atmospheric Boundary Layer - page 158

frequency of the backscattered signal and can easily be analysed. This is one of two
principal techniques used to analyse the Doppler shift of backscattered light. See
also “incoherent”.

Continuous-wave Often RADARs are operated as continuous-wave RADARs
(CW-RADAR) as an alternative to pulsed RADARs. CW-RADARs continuously
emit signals. The advantage of CW-RADARs is that they need considerably less
power and shielding efforts. The disadvantage of CW-RADARs is that they do not
allow for an easy range determination via the delay between emitted and received
signal.“frequency-modulated continuous-wave”.
Differential absorption Trace gas-specific active optical remote sensing can be
performed by emitting two signals with neighbouring wavelengths. One wavelength
is chosen so that it coincides with the centre of an absorption line of this trace gas,
the other close-by wavelength is chosen besides this absorption line. The difference
of the backscattered signal depends only on the trace gas concentration, because
backscatter from aerosols is practically identical for both wavelengths.

Doppler beam swinging method The Doppler beam swinging (DBS) method is
used to measure all three spatial components of the wind vector. Here, three to five
pulses are emitted one after the other into three to five different directions and the
Doppler shift of the backscattered signal is analyzed. One of the three or five direc-
tions is usually the vertical direction, the other directions are tilted by 15 ◦ to 20 ◦
from the vertical. The azimuthal directions of the tilted beams usually differ by 90 ◦ .
The angle between the slanted beams should be as low as possible to receive the
three - or five - different pieces of information, which are later used to compute
the three orthogonal wind components, from air volumes close together. If the angle
becomes too small, the computation of horizontal wind components from nearly
vertical radial velocity components along the beams is too uncertain.
Because the next pulse cannot be emitted before the backscatter of the previously
emitted pulse has been received in order to avoid disturbances, one measurement
cycle lasts six to ten times the time the pulse needs to travel up to the maxi-
mum vertical range of the instrument. A related technique is conical scanning (see
“scanning”).
Doppler broadening Caused by the thermal motion of gas molecules spectral lines
are not ideally thin but have a certain width due to the many small Doppler shifts
(“Doppler effect”) from the moving molecules. The shape of a Doppler broadened
line is Gaussian. The intensity of the thermal motion depends on the temperature
of the gas. Thus, Doppler broadening is temperature-dependent and the determi-
nation of the line width may be used for temperature measurements. This requires
spectrally high resolved measurements.

Doppler effect The Doppler frequency shift f (named after the Austrian physicist

C. Doppler (1803-1853)) is proportional to the emitted frequency f 0 and to the radial
speed of the scattering object, v where

f ≈ 2 f 0 v/c . The relative frequency shift,

f/f 0 , depends only on the ratio between the velocity of the scattering object and the
propagation speed, c , of the emitted signal. The received frequency is increased (the