Log‐normal distribution based Ensemble Model Output Statistics models for probabilistic wind‐speed forecasting. (26th March 2015)

• Record Type: Journal Article
• Authors: Baran, Sándor; Lerch, Sebastian
• Is Part Of: Quarterly journal of the Royal Meteorological Society. Volume 141:Number 691(2015:Jul.)
• Page Start: 2289
• Page End: 2299
• Publication Date: 2015-03-26
• Subjects: Meteorology -- Periodicals; 551.5
• DOI: 10.1002/qj.2521
• Languages: English
• ISSNs: 0035-9009

Abstract

<abstract abstract-type="main" id="qj2521-abs-0001"> <title> <x xml:space="preserve">Abstract</x> </title> <p id="qj2521-para-0001">Ensembles of forecasts are obtained from multiple runs of numerical weather forecasting models with different initial conditions and typically employed to account for forecast uncertainties. However, biases and dispersion errors often occur in forecast ensembles: they are usually underdispersive and uncalibrated and require statistical post‐processing. We present an Ensemble Model Output Statistics (EMOS) method for calibration of wind‐speed forecasts based on the log‐normal (LN) distribution and we also show a regime‐switching extension of the model, which combines the previously studied truncated normal (TN) distribution with the LN.</p> <p id="qj2521-para-0002">Both models are applied to wind‐speed forecasts of the eight‐member University of Washington mesoscale ensemble, the 50 member European Centre for Medium‐Range Weather Forecasts (ECMWF) ensemble and the 11 member Aire Limitée Adaptation dynamique Développement International‐Hungary Ensemble Prediction System (ALADIN‐HUNEPS) ensemble of the Hungarian Meteorological Service; their predictive performance is compared with that of the TN and general extreme value (GEV) distribution based EMOS methods and the TN–GEV mixture model. The results indicate improved calibration of probabilistic forecasts and accuracy of point forecasts in comparison with the raw ensemble and climatological forecasts. Further, the TN–LN mixture model outperforms the traditional TN method and its predictive performance is able to keep up with models utilizing the GEV distribution without assigning mass to negative values.</p> </abstract>