Original scientific paper
https://doi.org/10.65776/ep.20.4.4
Propagation of weather forecast uncertainties through attainable ship speed prediction models
Marijana Marjanović
; University of Rijeka, Faculty of engineering, Vukovarska 58, Rijeka, 51000, Croatia
Jasna Prpić-Oršić
; University of Rijeka, Faculty of Engineering, Vukovarska 58, Rijeka, 51000, Croatia
Marko Valčić
; University of Rijeka, Faculty of Engineering, Vukovarska 58, Rijeka, 51000, Croatia; University of Zadar, Maritime Department, Ruđera Boškovića 5, Zadar, 23000, Croatia
Abstract
Accurate prediction of weather-induced ship speed loss requires understanding how meteorological forecast uncertainties propagate through vessel performance models. This study presents a comparative assessment of three computational approaches for attainable speed estimation: the Wärtsilä NTPRO 5000 navigation simulator with JONSWAP and Pierson-Moskowitz spectral implementations, and HydroComp NavCad hydrodynamic software. Over 2,000 simulations were conducted for a 28,050 DWT bulk carrier across varying wave heights (0-12 m), encounter angles (0°-180°), and operational speeds (12.0-14.5 knots). Correlation analysis between meteorological predictor variables and ship speed response uncertainties revealed that significant wave height errors exhibit the strongest coupling with speed prediction errors (r = 0.65-0.97), while encounter angle geometry substantially modulates uncertainty propagation patterns. The Pierson-Moskowitz implementation demonstrated the most balanced error characteristics, whereas JONSWAP produced more polarised correlations at extended forecast horizons. Results indicate that uncertainty quantification approaches must account for both model-specific sensitivities and spatially varying forecast skill degradation.
Keywords
Weather forecast degradation; Attainable speed modelling; Uncertainty quantification; Correlation analysis
Hrčak ID:
348287
URI
Publication date:
26.6.2026.
Visits: 0 *