Abstract
<jats:p>Urban nature-based solutions (NbS), including trees and vegetated ground, modify urban energy and water exchanges through shading, evapotranspiration, and soil-vegetation interactions. However, these processes are either neglected or only implicitly represented in most urban canopy models used in regional and global land modeling systems. Here we develop a physics-based Urban-NbS scheme within a single-layer urban canopy model (SLUCM), which is coupled with the Noah-MP land surface model within the urbanized High-Resolution Land Data Assimilation System (u-HRLDAS), resulting in a new offline modeling system, u-HRLDAS-NbS. The Urban-NbS scheme explicitly represents within-canyon vegetation features and their coupled radiative, hydrological, and turbulent exchanges with other urban facets. Model performance is first evaluated using observations from 20 global urban flux towers under three alternative vegetation configurations. The proposed NbS configuration provides the best overall representation of urban surface energy exchanges while avoiding the geometric inconsistencies associated with treating the entire grid cell as a vegetated street canyon. Performance improvements are greatest for environment with intermediate urban and vegetation fractions, highlighting the importance of physically consistent subgrid urban vegetation partitioning. Regional 500-m modeling over New York City further demonstrates that, relative to the default u-HRLDAS, u-HRLDAS-NbS reduces the mean warm bias in 2-m air temperature from 0.59&deg;C to 0.10&deg;C across street-level sites and produces a domain-average daytime cooling of 1.31&deg;C. These results establish a scalable approach for representing urban NbS in land surface models and support future coupled urban weather and climate simulations and urban greening strategies.</jats:p>