2026-11-03 パシフィック・ノースウェスト国立研究所（PNNL）
Figure 2 compares urban fraction fields from default Weather Research and Forecasting (WRF) model, U-Surf, and adapted U-Surf-WRF with satellite imagery in four cities. The adapted fields show more fine-scale variation and less built-up suburban areas in most cities. Figure 2 from Jiang et al. (2026), used under the article’s Creative Commons Attribution license.
＜関連情報＞
- https://www.pnnl.gov/publications/neighborhood-details-enhance-regional-modeling-urban-heat-and-heatwaves
- https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2026MS005775
空間的に連続した都市表面特性が夏季および熱波シミュレーションに与える影響：複数都市分析 Impact of Spatially Continuous Urban Surface Properties on Summertime and Heatwave Simulations: A Multi-City Analysis
Timothy Jiang, TC Chakraborty, Yifan Cheng, Rohit Mukherjee, Lei Zhao, Alberto Martilli
Journal of Advances in Modeling Earth Systems Published: 13 October 2026
DOI:https://doi.org/10.1029/2026MS005775
Abstract
Urban areas are unique in form, function, and fabric, and representing them in process-based models requires, among other things, prescribing facet-level morphological and radiative properties, which vary widely. Most urban canopy models (UCMs) prescribe these by density class or local climate zone (LCZ), assigning identical values across broad regions or worldwide. Global km-scale, facet-level urban property data sets have recently emerged, but are seldom applied in regional modeling. Here, we incorporate one such data set, U-Surf, into the Weather Research and Forecasting (WRF) model, modifying it for WRF’s multi-layer UCM and releasing it as U-Surf-WRF. Considering 13 US cities, U-Surf-WRF parameters vary more between LCZs than default WRF parameters, with consistently lower impervious fraction. To determine the effects of using U-Surf-WRF, we conduct high-resolution WRF summertime seasonal and recent heatwave simulations for these 13 cities using default and U-Surf-WRF parameters. Either prescribed by LCZ or for each grid point, using U-Surf-WRF usually yields more accurate surface temperatures. It generally decreases modeled urban air temperature and increases modeled urban humidity, yielding lower heat and dry islands. Decomposing the impact of each U-Surf-WRF variable, we find that while albedo can impact daytime simulations, especially for air temperature, morphology and impervious fraction are most relevant, especially for surface temperature. This study demonstrates the impact of city-specific, facet-level urban properties in urban-resolving numerical weather simulations, with remaining gaps against observations demonstrating other challenges in representing cities in process-based models. Conversely, in WRF simulations with poorly constrained parameters, we suggest caution interpreting the magnitude and spatial distribution of urban signals.
