Geoscience Reference
In-Depth Information
Fig. 1 An overview of the different current methods for atmospheric sounding of CO
2
and CH
4
.
The FTIR spectrometer has a viewing geometry similar to (nadir-viewing) satellites, sampling the
whole atmosphere.
In situ instruments and tall
towers can only sample their
immediate
surroundings and the boundary layer, respectively
The standard TCCON retrieval is done by a least-square-
tting method, which
scales an a priori pro
le of the target
gas) in order to minimize a cost function. This approach yields the vertical scaling
factor (VSF) of the target gas, which is multiplied with the viewing geometry
dependent airmass and the a priori column to yield the retrieved gas column.
Simultaneously (i.e. from the same measurement/spectrum) O
2
is retrieved in the
same way and because the atmospheric content of O
2
is uniform and well known
(20.95 %), we can normalise the retrieved target gas by the retrieved O
2
value to
obtain the column-average dry-air mole fraction, e.g. for CO
2
:
le (i.e. an initial guess of the atmospheric pro
column
CO
2
column
O
2
xCO
2
¼
0
2095
ð
1
Þ
:
3 Spectra from the Bremen Instrument Setup
After characterisation and performance tests of the newly obtained InGaAs detector,
it was installed in the Bremen FTIR instrument and spectra were taken on several
consecutive nights around the full moon in July 2013.
Figure
2
shows two example spectra from July 21, 2013. The solar spectrum
(green) was taken using the standard TCCON setup, the lunar spectrum (blue
overlay) was taken the following night. The y-scale is arbitrary and dependent on
preampli
er settings of the detectors and thus not comparable. However, one can
identify the lower signal-to-noise (S/N) ratio of the lunar spectrum. The difference
in the 4,000
5,500 cm
−
1
region originates from the different sensitivity ranges of
the two detectors. The solar InGaAs detector has a lower cut cut-on wavenumber.
-
Search WWH ::
Custom Search