Quantifying urban function accessibility and its effect on population mobility based on function-associated population mobility network

• Function-associated population mobility network is proposed and constructed. • The k-step function accessibility with population mobility is defined to avoid underestimating transferability. • The function-driven index is proposed to quantify the driving effects of urban functions on population mobility. Due to rapid urbanization and globalization, urban functions are increasingly segregated in cities comprising centers of population aggregation and economic activity. Urban development yields intertwining and interdependent functional areas for residence, commerce, and education, which leads to complex but regular population mobility patterns. Population mobility spaces can effectively represent the actual service scopes of urban functions. Therefore, in this study, we propose a function-associated population mobility network (FPMN) to dynamically delineate the service capacities of urban functions, and to quantify the driving effects of urban functions on population mobility under spatial interaction conditions. By integrating the cumulative opportunities of geographic nodes with the potential opportunities reflected by population mobility, k-step function accessibility with population mobility (k-step PFA) is defined in FPMN to characterize the accessibility relationships between geographic nodes and function nodes. Based on k-step PFA, considering the complementarity and intervention opportunities, the function-driven index (FDI) is proposed to quantify the driving effects of urban functions on population mobility. Our experimental results indicated that FPMN could effectively represent the actual service scopes and capabilities of urban functions. Compared with general location-based accessibility indicators, k-step PFA has a higher numerical distribution and less spatial disparity, and it aligns more closely with the actual service scopes of urban functions. In addition, compared with intervening opportunities models, FDI delineates the functional driving effects of population mobility with greater precision, as well as revealing the travel propensity traits of a population relative to various urban functions through propensity parameters.