Wind-Wave Probabilistic Forecasting based on Ensemble Predictions | Maxime Fortin
| | Abstract | Wind and wave forecasts are of a crucial importance for a number of decision-making problems. Nowadays, considering all potential uncertainty sources in weather prediction, ensemble forecasting provides the most complete information about future weather conditions. However, ensemble forecasts tend to be biased and underdispersive, and are therefore uncalibrated. Calibration methods were developed to solve this issue. So far, these methods are usually applied on univariate weather forecasts and do not take any possible correlation into account. Since wind and wave forecasts have to be jointly taken into account in some decision-making problems, e.g. offshore wind farm maintenance, we propose in this thesis a bivariate approach, generalizing existing univariate calibration methods to jointly calibrated ensemble forecasts. A other method using the EPS-prescribed correlation in order to recover the dependence lost during the marginal calibration is also proposed. Even if the univariate performance of the marginal calibration is preserved, results confirm the need for bivariate approaches. Contrary to the univariate approach, the bivariate calibration method generates correlated bivariate forecasts, though it appears to be too sensitive to outliers when estimating necessary model parameters. Jointly calibrated distributions are too wide and therefore overdispersive. The different calibration methods are tested on ECMWF ensemble predictions over the offshore platform FINO1 located in the North Sea close to the German shore. | | Type | Master's thesis [Academic thesis] | | Year | 2012 | | Publisher | Technical University of Denmark, DTU Informatics, E-mail: reception@imm.dtu.dk | | Address | Asmussens Alle, Building 305, DK-2800 Kgs. Lyngby, Denmark | | Series | IMM-M.Sc.-2012-86 | | Note | DTU supervisor: Pierre Pinson, pp@imm.dtu.dk, DTU Informatics | | Electronic version(s) | [pdf] | | Publication link | http://www.imm.dtu.dk/English.aspx | | BibTeX data | [bibtex] | | IMM Group(s) | Mathematical Statistics |
|