Forecasting number of foreign tourist arrivals using time series intervention analysis based on SARIMA and log-linear poisson autoregressive model

Forecasting the influx of foreign tourists is imperative for effective tourism management and planning. This quantitative study employs two time-series models-SARIMA and Log-linear Poisson Autoregressive Intervention Analysis-to provide a robust forecasting model for count data on foreign tourist arrivals at Indonesia’s Ngurah Rai Bali Airport. These statistical models were chosen for their proficiency in handling count data, enabling the evaluation of interventions, including tourism policy changes and global events like the COVID-19 pandemic, on foreign tourism. Assessment based on the lowest RMSE, MAD, and MAPE values indicates the best model. The SARIMA intervention model, identified as the best performer, predicts a total of 3,631,933 arrivals from May to October 2024. Furthermore, this research presents estimates of the financial repercussions of the COVID-19 pandemic on the Indonesian tourism industry, with potential losses approximated at USD 12,150,579,685 at Ngurah Rai Airport. The outcomes underscore the severe economic consequences of global emergencies on tourism, emphasizing the need for sustainable recovery strategies to address challenges arising from future unknown emergencies.