Lufft CHM 15k User Manual Download Page 56

Page: 56 / 73

Operating Manual CHM 15k

R13 / 07-2019

Data Evaluation / Sky Condition Algorithm (SCA)

Data Evaluation / Sky Condition Algorithm (SCA)

The  CHM 15k  ceilometer  is  a  laser  remote  sensing  device  with  embedded  algorithm  for  determining
layers of particles and droplets in the atmosphere. The embedded algorithm is collectively referred to as
the Sky Condition Algorithm (SCA). Ceilometers determine the cloud base and provide information on
the penetration depth into cloud. In case another cloud or aerosol layer can be measured above the lower
cloud  the  penetration  depth  can  be  interpreted  as  cloud  thickness.  In  addition,  the  degree  of  cloud
coverage  is  determined  in  terms  of  eighths  of  the  sky.  For  visibilities  below  2  km  the  vertical  visibility
(VOR) is calculated and output in addition. An aerosol algorithm based on a wavelet algorithm detects
different aerosol layers and transmits those detected within the atmospheric boundary layer. Fog / haze
and precipitation are detected and transmitted in the Sky Condition Index (SCI) parameter.

9.1 Laser remote sensing

A pulsed near infrared laser probes the sky vertically from top of the instrument up to 15 km altitude.
Targets like aerosol layers and clouds show up as echoes with certain backscatter intensity and signal
extinction. Molecular absorption as well as Rayleigh scattering by air molecules is negligible at a laser
wavelength of 1064 nm. The distance of the scattering particles to the instrument is calculated from the
travelling time of the laser pulses.

9.2 Preparation of the measured data

Data pre-processing is an important task before the different steps in the SCA algorithm begin. The main
reason for this is to harmonize / normalize the data sets between different CHM 15k systems to get similar
results, e. g. for cloud bases, even if the sensitivity between instruments varies.
Each single measurement is normalized by determining the detection sensitivity with a reference light
pulse p

calc

. Differences between different devices are compensated by a scaling factor c

, which is

determined by a comparison measurement with a reference device. Figure 25 shows the profiles of two
different devices after normalization and calibration.

Figure 25 Normalized backscatter signals P(r) for reference unit (blue) and a test unit (red). A horizontal path is used with a hard
target in 9.4 km distance for this method. At 16 km distance the reference light pulse is visible.

The following formula is used to obtain the normalized backscatter signal:

𝑃(𝑟) =

𝑃

𝑟𝑎𝑤

(𝑟) − 𝑏

𝑐

𝑠

∙ 𝑂(𝑟)

∙

𝑝

𝑐𝑎𝑙𝑐

Here,

raw

corresponds to the raw backscatter signal,

to the baseline and

O(r)

is the overlap function.

calc

and c

are the normalization and calibration factor, respectively. The normalized backscatter signal

P(r)

is multiplied with

and stored in the variable beta_raw in the NetCDF files.

A further processing step is performed to determine cloud heights and aerosol layers. To compensate for
the reducing signal-to-noise ratio at higher altitudes, the signal is averaged with a height-depending
averaging time as shown in Figure 26. At different altitudes, the time averaging varies from 15 seconds
below 3 km to 300 seconds above 6 km.