hide
Free keywords:
-
Abstract:
Clouds in the atmosphere have different macroscopic shapes and microphysical characteristics, strongly influencing the photon trajectories of the incoming solar radiation. Reflection at the cloud top leads to a shortening of the photon path lengths while multiple scattering inside the cloud and reflections between different cloud clusters can significantly increase the path lengths. All these modifications affect the short-wave energy deposition of the atmosphere.
Within this thesis, photon path length distributions of reflected sunlight and other cloud properties are retrieved from high spectral resolution satellite measurements of the oxygen A-band by performing comparative radiative transfer simulations. The measured radiances originate from the Fourier Transform Spectrometer TANSO-FTS onboard the GOSAT satellite. Radiative transfer simulations of different cloud scenarios are performed with the Monte Carlo model McArtim and compared to the measurements. The simulation output provides direct access to the scattering events, allowing the calculation of photon path length distributions. The comparison between measurement and simulation is achieved by means of an optimization procedure, which was developed within this thesis.
The investigation is applied to selected measurements in presence of single layer clouds and one cloud system with multiple layers. For those cases, also collocated lidar measurements of CALIOP (CALIPSO) and radar measurements of CPR (CloudSat) are available. The cloud properties of the simulations, which agree best to the measurement, are compared to the collocated cloud profiles from CALIOP and CPR as well as to the cloud optical depths from TANSO-CAI (GOSAT).
In general, the results show a reasonable agreement with the independent collocated measurements and the compared cloud optical depths agree well. From the comparison with the collocated cloud profiles, a systematic overestimation of the retrieved cloud top heights was found for the single layer cloud cases and an indication for an overestimation of the simulated O2 absorption in the order of 5-10%. In order to resolve this discrepancy, different sensitivity studies have been performed, evaluating the horizontal homogeneity of the cloud system and simulation input parameters like the O2 absorption cross sections and the asymmetry parameter. So far, no explanation for the observed discrepancy has been found. However, it is unlikely that it is caused solely by the simulation uncertainties. Interestingly, a similar overestimation for the O2 absorption has been observed for a clear sky case over a bright surface. A clear-sky case over a dark surface, however, shows the opposite effect, suggesting an underestimation of the O2 absorption. It is interesting to note that the corresponding radiance spectrum shows many negative values, indicating possible calibration problems. Here, more clear-sky cases of different surface albedos should be investigated for a better understanding of the performance of the instrument. The case study of the multiple layer cloud system leads to a better comparison to the collocated cloud profiles. Besides the observed overestimation of the O2 absorption, the developed investigation method works well and can be applied to satellite measurements on a global scale.
