A hybrid big-data framework for tourist and high-speed rail mobility modelling

This paper introduces a novel hybrid methodology to address the critical challenge of accurately forecasting demand for significant infrastructure projects, such as High-Speed Rail (HSR), where traditional estimates often suffer from significant overestimation. The study has three main objectives: (1) to identify and quantify biases in current long-distance transport demand forecasts; (2) to understand how these biases affect the highly variable and unpredictable demand generated by tourists; and (3) to introduce an integrated approach that reduces such biases and improves transport planning by combining large-scale mobile phone signalling data with traditional data sources (e.g., traffic counts, travel surveys). The novelty of the study lies in the integration of behavioural big data into transport modelling and in its ability to capture same-day visitors—a group typically underrepresented in conventional tourism statistics yet highly relevant for long-distance mobility forecasts. Using the planned Prague–Brno HSR corridor in the Czech Republic as a real-world case study, we analyse both the advantages and drawbacks of using signalling phone data for transport planning. This approach is crucial because multi-billion-euro HSR investments depend on accurate forecasts; overestimations risk misallocating vast public funds. Our method provides a more granular understanding of travel dynamics, particularly by capturing the highly volatile and often underreported impact of tourist flows on long-distance travel. Demand is estimated through scenarios to reduce uncertainty, reflecting alternative assumptions about behavioural responses and modal shifts. The analysis estimates a potential annual demand of 4.5–10.5 million passengers for the Prague–Brno HSR, a figure significantly lower than official projections. This discrepancy highlights the risk of optimism bias in conventional forecasting. For policymakers, the primary conclusion is the urgent need to institutionalise multi-source, data-driven methods for evidence-based decision-making. By providing a more realistic picture of current demand and revealing nuanced behaviours such as regional differences in the willingness to shift from cars to HSR, this hybrid approach supports fiscally responsible planning and the development of targeted strategies to promote sustainable mobility.