Publication Date:
2019-02-19
Description:
The present work describes the determination of the cloud liquid water path from passive measurements of the thermal radiation of the atmosphere. The data used came from expedition cruises aboard the research vessel Polarstern within the project OCEANET. The whole process of remote sensing, from the preparation and execution of the measurements, the processing of the data, including the retrieval creation process, to validation and evaluation of the results has been reviewed and improvements have been found. A new retrieval has been created based on reanalysis data, that breaks away from the up to now usual geographical restriction and thus allows seamless remote sensing on the entire trip across the Atlantic. This is achieved while reducing the error to 20 g/m^2 wich is comparable to current retrievals for shore stations and better than error values of previously used retrievals for remote sensing at sea. Due to the successful use of reanalysis data in the training process, the creation of retrievals was freed from the restriction on radiosonde data sets. The data obtained was extensively compared with climatology and model results, and despite the small number of expeditions, and thus statistically independent information, only minor differences were found. Especially in the subtropics, where shallow boundary layer clouds dominates, the observed differences are in the range of the retrieval error. But it's particularly evident in this area, that, compared with the climatology, the obtained data is still contaminated due to outliers on synoptic scales. It is expected that the matches further improves, when more expedition data is included in the data base. Finally, the gathered data is used as a reference for the validation of satellite remote sensing. The CM-SAF uses Meteosat-SEVIRI radiances for the determination of the liquid water path. Here too, the comparison shows a good to very good agreement. However in some cases, the temporal mean from ground measurements and the spatial mean from the satellite can't be compared strait forward. So to assess the quality of these comparisons, one needs information about the type and distribution of the clouds.
Type:
Thesis
,
NonPeerReviewed
Format:
text
Permalink