A Novel Downscaling Approach for Urban Climate: Land-Surface-Physics-Based Downscaling

The substantial computational resources required for dynamical downscaling (D-DS) have limited extensive exploration of urban climate prediction, particularly for long-term and high-resolution scenarios. To bridge this gap, we propose a novel downscaling approach: Land-Surface-Physics-Based Downscaling (LSP-DS). This innovative approach aims to perform high-resolution, long-term urban-specific simulations with significantly reduced computational demands.

LSP-DS integrates the widely used HRLDAS/Noah-MP land surface model with urban canopy-process physics (SLUCM) and is driven by coarse-resolution reanalysis or numerical modeling data. Our evaluation of LSP-DS focuses on simulating the interactions between the urban heat island (UHI) effect and heat waves (HWs), using the Tokyo area as a case study. The analysis, which incorporates observational data with numerous LSP-DS simulations, confirms that the impact of UHI intensifies during HW periods.

To further assess the performance of LSP-DS, we conducted a comparative analysis with conventional direct D-DS (WRF) over the past decade in the Tokyo and Singapore regions. The results reveal that the WRF model overestimates nighttime urban temperatures, leading to an overestimation of the UHI effect, while HRLDAS/Noah-MP demonstrates accurate UHI effect estimates for both daytime and nighttime. Additionally, it was found that the urban-rural 1st atmospheric layer temperature profiles of HRLDAS/Noah-MP are too similar, but the SLUCM proved to be highly effective.

This research underscores the potential of LSP-DS in urban climate prediction, offering a less resource-intensive yet accurate alternative to conventional D-DS methods, with significant implications for urban climate studies and policy-making.