An urban module coupled with the Variable Infiltration Capacity model to improve hydrothermal simulations in urban systems

Abstract

Global urban expansion has altered surface aerodynamics and hydrothermal dynamics, aggravating environmental challenges such as urban heat and urban dry islands. To identify such environmental responses, various physical models, including urban canyon models (UCMs) and land surface models (LSMs), have been developed to represent surface hydrothermal processes. However, UCMs often treat a city as a unified entity and overlook subcity heterogeneity. LSMs are generally designed for natural land cover types and lack the capability to capture urban characteristics. To address these limitations, the aim of this study is to couple an urban module with a sophisticated LSM, i.e. the Variable Infiltration Capacity (VIC) model. This coupled model, i.e. the VIC-urban model, is characterized by its ability to coordinate certain critical urban features, including urban geometry, radiative interactions, and human impacts. Adopting Beijing as an evaluation site, the VIC-urban model shows higher performance than the original version, with excellent accuracy in simulating sensible heat, latent heat, runoff, and land surface temperature (LST). The absolute error is smaller than 25 % for the sensible heat and latent heat and smaller than 12 % and 30 % for the LST and runoff, respectively, which indicates that VIC-urban can effectively simulate hydrological and thermal fluxes in urban systems. Sensitivity analysis reveals that the roof emissivity and interception capacity exert the greatest impact on the roof temperature and evaporation and the height-to-width ratio has the greatest influence on the canyon. Our work introduces a reliable option for large-scale land surface simulations that accounts for urban environments and is among the first attempts to establish a systematic urban modelling framework of the VIC model. The VIC-urban model enables the analysis of urbanization-induced environmental changes and quantification of environmental variations among different urban configurations. The proposed model can thus offer invaluable insights for urban planners and landscape designers.