GMDH-based wave prediction model for one-week nearshore waves using one-week forecasted global wave data

Wave forecasting up to 168 h (one-week) ahead is crucial the operational management of shipping and construction in marine work site. The present study proposes a method areal-time One-week Forecast Nearshore Waves (OWFNW) at 13 stations on Japanese coast. For each station, two Group Method Data Handling (GMDH) based wave prediction models for height period are developed transform global data into nearshore waves. GMDH-based model, four trained with different training sets using one-week forecast released real-time by Japan Meteorological Agency (JMA), National Oceanic Atmospheric Administration (NOAA), European center Medium-Range Weather Forecasts (ECMWF). conducts series testing examines accuracy predictions. best performing model selected among compared data. It was found that performance varies locations. On average, able improve predictions 38% lead time. improves error 65% forecasts. methodology get OWFNW can be applied other seas where corresponding observed one available.