Presentation of my master project, that was the implementation of a flexible framework facilitating the characterisation of meteorological data.
Here is the abstract of my Master thesis. More information can be found on the project's page.
Cometa is a flexible framework which facilitates the characterisation of meteorological data by means of metadata and supports the consistent combination of very different kinds of data. Its goal is to develop and validate integrated approaches in the context of the meteorological forecasting class called nowcasting. Cometa is conceived extensible, allowing to easily incorporate further or more refined data characteristics and combination methods. XML files are used to describe data types and processing information, while a set of Java classes composing the framework provides functionalities to verify data consistency and to manage data combination. In addition, it allows to improve existing algorithms through parameter perturbation functionalities.
Its capabilities have been tested through a concrete use case. A synergetic application, which uses both radar and satellite observations has been implemented and exploited in case of thunderstorms developments over the Alpine region. High quality instantaneous precipitation fields, retrieved by radar, have been forecasted according to dynamical features, retrieved by satellite.
Cometa proved to be a valid tool for setting up and automating large R&D sets of experiments with little effort; first quality assessments shed a light on potential meteorological developments of this nowcasting technique.
Cometa is the fruit of a collaboration between the Federal Office of Meteorology and Climatology MeteoSwiss and the DIVA Group at the University of Fribourg.