Title: Probabilistic temperature forecasting using $d$-vine copula regression
Authors: Annette Moeller - Clausthal University of Technology (Germany) [presenting]
Claudia Czado - Technische Universitaet Muenchen (Germany)
Daniel Kraus - Technische Universitaet Muenchen (Germany)
Ludovica Spazzini - Partners4Innovation (Italy)
Abstract: To account for forecast uncertainty in numerical weather prediction (NWP) models it has become common practice to employ ensemble prediction systems generating probabilistic forecast ensembles by multiple runs of the NWP model, each time with variations in the details of the numerical model and/or initial and boundary conditions. However, forecast ensembles typically exhibit biases and dispersion errors as they are not able to fully represent uncertainty in NWP models. Therefore, it is common practise to employ statistical postprocessing models to correct ensembles for biases and dispersion errors in conjunction with recently observed forecast errors. We propose a novel postprocessing approach for temperature forecasts based on $d$-vine copula quantile regression. The $d$-vine copula regression model is a multivariate regression approach predicting quantiles of a response (temperature observations) based on a set of predictor variables (ensemble forecasts). It exploits the dependence of observation and predictors, accounting for non-Gaussian dependencies in a flexible way. In a comparative study with temperature forecasts of different forecast horizons from the European Center for Medium Range Weather Forecast (ECMWF) the $d$-vine postprocessing approach shows to be highly competitive to the state-of-the-art EMOS model, clearly improving over standard EMOS for larger forecast horizons.