Identifying Systematic Origin–Destination-Level Mismatches Between Household Travel Surveys and Mobile Phone Data: A Case Study of Seoul

Abstract

This study examines mismatches between Household Travel Survey (HTS) and mobile phone-based LTE/5G Cellular Signaling (LCS) data in representing origin–destination (OD) trip patterns in Seoul, South Korea. By comparing OD flows across subgroups defined by travel type, age group, gender, and arrival time, we quantify distributional mismatches using standardized root mean square error (SRMSE) and the relative zero-cell ratio. The results reveal larger mismatches for children, young adults, and noncommuting trips. A spatial regression analysis, using origin-level SRMSE as the dependent variable, shows that areas with larger numbers of households, higher average land prices, and larger tertiary sector business size exhibit lower mismatches, whereas areas with concentrated residential land use and high OD pattern diversity show higher mismatches. The findings suggest that observed mismatches between HTS and LCS are associated with subgroup characteristics and spatial context. These insights underscore the need for tailored data integration and sampling strategies that account for behavioral and spatial biases when combining survey and passively collected mobility data for travel behavior analysis.