Barriers to incorporating dichotomy-based impervious surface datasets into high-resolution urban climate studies

Source abstract

Urban impervious surfaces fundamentally alter urban thermal, hydrological, and ecological processes. However, their accurate spatiotemporal characterization at intra-urban scales remains a major challenge for urban climate research. Most existing global impervious surface area (ISA) datasets rely on impervious–pervious dichotomy frameworks, and their application in urban models remains limited, with their suitability for representing intra-urban impervious surface structures insufficiently evaluated. Here, we systematically evaluate ten high-resolution (10–30 m) global ISA datasets using Urban-PLUMBER site observations, multi-scale spatial analyses, and the CoLM-UBCM urban canopy simulations. Results show that most dichotomy-based ISA products exhibit substantial biases, with deviations of up to 52% relative to reference impervious fractions and limited ability to capture realistic characteristics of urban impervious surfaces. Increasing spatial resolution from 30 m to 10 m improves spatial detail but does not fundamentally resolve these inconsistencies. Urban canopy simulations further reveal that ISA uncertainties can directly propagate into urban energy and hydrological processes, producing maximum hourly deviations in summer-mean diurnal cycles of up to 12.3 K for ground surface temperature, 57.4 W m −2 for upward longwave radiation, 46.8 W m −2 for sensible heat flux, and 5 mm h −1 for peak hourly surface runoff. The results highlight critical limitations of dichotomy-based ISA products for high-resolution urban climate applications. We emphasize that accurate impervious surface data with continuous fractional representations are essential for improving urban climate and Earth system modeling under rapid urbanization and climate change.