Short-Term ForecastInland Cruise Passenger Demand in the Republic of Serbia

Abstract

Forecasting cruise traffic is vital for industry stakeholders seeking to manage port infrastructure, marketing strategies, and service offerings effectively. Unlike other transportation modes, cruise tourism presents unique forecasting challenges due to its dependence on seasonal patterns, destination attractiveness, economic conditions, and geopolitical influences. In this study, we apply seasonal ARIMA and ARIMAX models to forecast cruise passenger demand in Serbia. Models were fitted using the Box–Jenkins methodology, with a Box–Cox transformation applied to stabilize the variance, and comprehensive residual diagnostics were used to ensure robust model specification. The results show that incorporating Germany’s GDP substantially improves both model fit and forecast accuracy, while temperature‑based regressors introduce noise rather than explanatory or predictive value. These findings indicate that the proposed methodology is effective and that seasonal ARIMA and ARIMAX models are useful predictors of cruise tourism demand.