Assessing Spatial Accessibility Uncertainty with Dempster–Shafer Theory: A Comparison of Potential and Revealed Accessibility

This study introduces a framework for comparing and integrating revealed and potential accessibility maps, using the Dempster–Shafer theory to identify regions with varying spatial accessibility while accounting for uncertainty. It presents a method for determining revealed accessibility from individuals’ trajectory data, weighting accessibility inversely to the square of uncertainty. This dual approach aids urban planners in making more reliable decisions. The methodology is applied to supply centers, including shops, restaurants, and sports centers, using data from the Mobile Data Challenge (MDC) in Vaud, Switzerland. The results show good access to shops in the northwestern and southeastern regions and good access to restaurants in the eastern regions. The final maps indicate that areas with low access to sports centers form the highest proportion (62.7%) of regions with low access, while those with low access to shopping centers form the lowest (9.3%). The findings suggest the need for more sports centers in Nyon and Jura-Nord Vaudois and more accessible restaurants in Nyon and southern Aigle. Additionally, the analysis reveals that lower station densities correlate with smaller discrepancies between real and expected accessibilities, while higher population densities are linked to lower uncertainty, underscoring the importance of considering density in spatial accessibility assessments.