• Home
  • Research
  • Research in progress

Research in progress

Currently, the following proposed thesis topics are being developed in the doctoral program:

Spectral Wave Data Assimilation Scheme For Operational Forecasting

  • There are several challenges in wave spectral data assimilation (DA). The first is related to the fundamental structure of the variable to deal with, which is a 2D matrix containing the wave energy distribution in frequency and direction. This distribution is characterized by the existence of energy clusters associated to different meteorological genesis, i.e., wave systems. Consistent wave DA must consider these individual wave systems separately because in bi-modal or multi-modal sea states integral parameters of the whole spectrum lose significantly their meaning. Another challenge is associated to the inherent variability contained in data in general and spectral wave data in particular. The uncertainties associated to the directional distribution of measured spectra are large, if at all known. In general also, a one to one match between model and observed wave systems is unlikely so a robust cross-assignment algorithm is a fundamental component of the whole system. Moreover, DA in general consists of spreading point measured information into the model spatial domain. In any particular area every overlapping wave system has its own spatial domain plus other specific characteristics, which need to be characterized in order to determine the associated background errors, the other fundamental element of the DA system. From the model point of view the dissection and the later recomposition of its main variable, the wave spectrum, is a promise of numerical instabilities. For most of these challenges a sound technique is the use of spectral partitioning and long-term wave spectral statistics, which allows characterizing wave systems, carry out a consistent cross-assignment between observations and model systems, and specifying background errors. For the optimization step, several strategies need to be tested (e.g., Optimal Interpolation, 3DVAR, 4DVAR) and assessed in terms of effectiveness and efficiency. The modelling platform is the WaveWatch III model, and the data to be used are any type of spectral wave data, including directional buoys or satellite SAR spectra..
  • Synthesis of unconventional materials for environmental applications