Proceedings of Topical Meeting on Optoinformatics (pdf-format, 1.21 ...

56 OPTOINFORMATICS’05
LIDAR SIGNAL PROCESSING
G. J. Ciuciu, D.N. Nicolae, C. Talianu, M. Ciobanu, V. Babin
National Institute <strong>of</strong> R&D for Optoelectronics, 1 Atomistilor Str., Bucharest – Magurele,
P.O. Box MG-5, RO-077125, Romania
Tel/Fax: 40-21-457 45 22, E-mail: jeni@inoe.inoe.ro, URL: http://inoe.inoe.ro
This paper presents LiSA method for Lidar signal processing. It is a combined
method, based on the Fernald-Klett algorithm and optical atmospheric model.
The purpose is to provide some atmospheric parameters: backscattering and
extinction coefficients.
LIDAR technique is an active remote sensing method and is based on the emission <strong>of</strong> short
laser pulses (ns or fs duration) to the atmosphere under study. The laser photons
backscattered by the atmospheric volume under study are collected by a receiving optical
telescope. The signals are acquired and digitized in the analog and/or the photon counting
mode by fast transient recorders and subsequently transferred to a personal computer for
further analysis and storing.
Recently, at the National Institute <strong>of</strong> R&D for Optoelectronics – Romania, a compact
backscatter lidar was installed. LiSA system can work separately or simultaneous on two
wavelengths. It is made to detect from distance (max. 10 Km) micronic aerosols, with a
spatial resolution <strong>of</strong> 6 m, using as sounding radiation the beam <strong>of</strong> a Nd:YAG laser with
second harmonic.
To analyze the return signal in laser remote sensing means to find solutions for the
equation which relates the characteristics <strong>of</strong> the received and emitted signal, and the
propagation medium. The form <strong>of</strong> the equation depends <strong>of</strong> the interaction type [1] . For those
applications which involves scattering (elastic or inelastic), the form <strong>of</strong> equation is quite
simple:
Z
β ( Z ) ⋅ exp[ − 2∫
α( z)
dz]
Z0
S ( Z ) = C ( Z ) ⋅
+ S
(1)
S
Z
2
bg
where Z is the distance to the scattering point, S(Z) is the Lidar signal (power), C S (Z) is the
so-called system function, β(Z) is the backscattering atmospheric coefficient at distance Z,
α(Z) is the extinction atmospheric coefficient at distance Z and S bg is the background
signal (power).
In writing this equation, the multiple scattering was neglected. The main problem is that
we have two unknown parameters - β(Z) and α(Z) – and one equation, so that is necessary
to postulate a relation between the two. For this, the Lidar ratio S a ( Z ) = α( Z ) β ( Z ) is used.
This parameter is dependent <strong>of</strong> the scatterer dimension and, if is known, the equation can
be solved. To know Sa values over entire investigation distance is mostly impossible.
In 1972, Fred Fernald [2] realised that Lidar equation is a Bernoulli equation on first rang
and obtained its solution in ‘forward’ form, choosing for calibration the closest point Z 0 in
the investigation interval. This method works well if the backscattering coefficient in Z 0
can be provided by complementary measurements. In 1981, Klett [3] proved that this
solution becomes unstable if atmospheric extinction is important and in that case it
diverges with increases <strong>of</strong> the distance. He suggested an ‘inversion’ <strong>of</strong> solution, that means
to choose the references point Z ∞ at the end <strong>of</strong> the investigation interval. Rearrangement <strong>of</strong>