DHI (Danish Hydraulic Institute) and HyMOLab (Hydrodynamics and Met-Ocean Laboratory of the Dept. of Engineering and Architecture of the University of Trieste) have undertaken a joint applied research project with the aim to develop a state-of-art wind-wave forecast service at mid resolution for the Mediterranean Sea and at very high resolution for the Adriatic Sea. Weather routing, civil protection, coastal engineering, oil&gas and renewable energy fields, the planning of operations at sea, ... are just few among the multiple potential applications of this service. The meteorological model used in this study is WRF-ARW, one of the most widely used state-of-the-art open-source non-hydrostatic model. Global Forecast System (GFS) dataset provides the boundary and initial conditions. MIKE21-Spectral Waves is used as wave model with resolution ranging from 0.1 to 0.03 approximately. The use of a local area meteorological model guarantees higher levels of resolution and accuracy in an area such as the Mediterranean Sea where the complex orography and coastline induce short-time/small-space weather scales. The model chain runs daily (or twice a day on demand) on the High Performance Computing (HPC) infrastructure of HyMOLab. The validation of the entire model chain and specifically the forecast data obtained for the sea state is continuously updated according to new available data from satellites and buoys. Anyway, a major verification of the performance of the model chain against historic data (hindcast) is almost mandatory. For this aim, we performed a multi-decade test obtaining very good statistical parameters for the entire model chain performance. In this context the hindcast dataset developed by DHI and HyMOLab consists of 35 years of hourly data for the period 1979-2013, with the same model chain. The CFSR d093.0 hourly dataset with a spatial resolution of 0.5 provides the boundary and initial conditions. The atmospheric and wave models performance is checked against six satellite datasets, missions Envisat, ERS-2, Geosat FO, Jason-1, Jason-2, Topex-Poseidon, using a moving window technique procedure. Wave data close to coast are compared with available data from more than 20 buoys. The paper describes the validation procedure adopted for the hindcasted data. Furthermore the forecast service is described too, with specific emphasis to the very high resolution adopted in the Adriatic Sea.
A high resolution wind&wave forecast model chain for the Mediterranean and Adriatic Sea
DONATINI, LUCA;LUPIERI, GUIDO;CONTENTO, GIORGIO;FEUDALE, LAURA;
2015-01-01
Abstract
DHI (Danish Hydraulic Institute) and HyMOLab (Hydrodynamics and Met-Ocean Laboratory of the Dept. of Engineering and Architecture of the University of Trieste) have undertaken a joint applied research project with the aim to develop a state-of-art wind-wave forecast service at mid resolution for the Mediterranean Sea and at very high resolution for the Adriatic Sea. Weather routing, civil protection, coastal engineering, oil&gas and renewable energy fields, the planning of operations at sea, ... are just few among the multiple potential applications of this service. The meteorological model used in this study is WRF-ARW, one of the most widely used state-of-the-art open-source non-hydrostatic model. Global Forecast System (GFS) dataset provides the boundary and initial conditions. MIKE21-Spectral Waves is used as wave model with resolution ranging from 0.1 to 0.03 approximately. The use of a local area meteorological model guarantees higher levels of resolution and accuracy in an area such as the Mediterranean Sea where the complex orography and coastline induce short-time/small-space weather scales. The model chain runs daily (or twice a day on demand) on the High Performance Computing (HPC) infrastructure of HyMOLab. The validation of the entire model chain and specifically the forecast data obtained for the sea state is continuously updated according to new available data from satellites and buoys. Anyway, a major verification of the performance of the model chain against historic data (hindcast) is almost mandatory. For this aim, we performed a multi-decade test obtaining very good statistical parameters for the entire model chain performance. In this context the hindcast dataset developed by DHI and HyMOLab consists of 35 years of hourly data for the period 1979-2013, with the same model chain. The CFSR d093.0 hourly dataset with a spatial resolution of 0.5 provides the boundary and initial conditions. The atmospheric and wave models performance is checked against six satellite datasets, missions Envisat, ERS-2, Geosat FO, Jason-1, Jason-2, Topex-Poseidon, using a moving window technique procedure. Wave data close to coast are compared with available data from more than 20 buoys. The paper describes the validation procedure adopted for the hindcasted data. Furthermore the forecast service is described too, with specific emphasis to the very high resolution adopted in the Adriatic Sea.File | Dimensione | Formato | |
---|---|---|---|
IMAM_2015.pdf
accesso aperto
Tipologia:
Bozza finale post-referaggio (post-print)
Licenza:
Digital Rights Management non definito
Dimensione
5.58 MB
Formato
Adobe PDF
|
5.58 MB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.