Abstract Urban Land Surface Models (ULSMs) simulate energy and water exchanges between the urban surface and atmosphere. However, earlier systematic ULSM comparison projects assessed the energy balance but ignored the water balance, which is coupled to the energy balance. Here, we analyze the water balance representation in 19 ULSMs participating in the Urban‐PLUMBER project using results for 20 sites spread across a range of climates and urban form characteristics. As observations for most water fluxes are unavailable, we examine the water balance closure, flux timing, and magnitude with a score derived from seven indicators expecting better scoring models to capture the latent heat flux more accurately. We find that the water budget is only closed in 57% of the model‐site combinations assuming closure when annual total incoming fluxes (precipitation and irrigation) fluxes are within 3% of the outgoing (all other) fluxes. Results show the timing is better captured than magnitude. No ULSM has passed all water balance indicators for any site. Models passing more indicators do not capture the latent heat flux more accurately refuting our hypothesis. While output reporting inconsistencies may have negatively affected model performance, our results indicate models could be improved by explicitly verifying water balance closure and revising runoff parameterizations. By expanding ULSM evaluation to the water balance and related to latent heat flux performance, we demonstrate the benefits of evaluating processes with direct feedback mechanisms to the processes of interest.

Abstract Urban Land Surface Models (ULSMs) simulate energy and water exchanges between the urban surface and atmosphere. However, earlier systematic ULSM comparison projects assessed the energy balance but ignored the water balance, which is coupled to the energy balance. Here, we analyze the water balance representation in 19 ULSMs participating in the Urban‐PLUMBER project using results for 20 sites spread across a range of climates and urban form characteristics. As observations for most water fluxes are unavailable, we examine the water balance closure, flux timing, and magnitude with a score derived from seven indicators expecting better scoring models to capture the latent heat flux more accurately. We find that the water budget is only closed in 57% of the model‐site combinations assuming closure when annual total incoming fluxes (precipitation and irrigation) fluxes are within 3% of the outgoing (all other) fluxes. Results show the timing is better captured than magnitude. No ULSM has passed all water balance indicators for any site. Models passing more indicators do not capture the latent heat flux more accurately refuting our hypothesis. While output reporting inconsistencies may have negatively affected model performance, our results indicate models could be improved by explicitly verifying water balance closure and revising runoff parameterizations. By expanding ULSM evaluation to the water balance and related to latent heat flux performance, we demonstrate the benefits of evaluating processes with direct feedback mechanisms to the processes of interest.

The accuracy of PM 2.5 forecasts in Seoul from 2019 to 2023 was assessed using multiple methods. Daily short-term PM 2.5 forecasts, which were provided as four categories ( good , moderate , bad , and very bad ), were directly compared with the corresponding PM 2.5 observation data. Although the probability of detection for days with high PM 2.5 concentrations increased, a simultaneous rise in the false alarm rate resulted in no improvement in the total accuracy and F1-score. To analyze these trends in more detail, the forecast accuracy was further examined based on the PM 2.5 categories. The results showed an annual improvement of 3.65% in the accuracy for the bad category. An analysis based on the announcement time also indicated an increase of over 20% in the accuracy for the bad category for next-day and day-after forecasts. The confusion matrices of forecasted and observed PM 2.5 categories confi rmed this improvement, which was primarily due to a reduction in the number of instances where the moderate category was forecasted as bad . However, the accuracy for the good category showed no signifi cant change and that for the moderate category even declined. These fi ndings highlight the importance of category-specifi c evaluation in air quality forecasting and improving the forecast accuracy, particularly for the good and moderate categories. The reliability of forecasts and their policy relevance may be improved by utilizing these insights and addressing temporal and spatial limitations.

Abstract Enhanced surface runoff in urban environments reduces water availability and limits the evaporative cooling potential. We evaluate surface runoff in 18 urban land surface models (ULSM) in Urban‐PLUMBER for 6,570 rainfall events across 21 urban sites. Surface runoff occurs when rainfall exceeds the infiltration, saturation, or interception capacity. Ten models omit at least one of these processes, while seven fail to increase runoff with increasing imperviousness. Surprisingly, some models lack any runoff during intense (>) or prolonged (>20 mm) rainfall. Urban land surface models (ULSMs) turn 0%–86% rainfall into runoff. Most models produce runoff in agreement with an empirical comparison offered by the CN method, especially for high imperviousness. However, ULSM runoff exceeds CN runoff estimates for low impervious fraction, particularly by models with incomplete process description and for low‐intensity rainfall. The large discrepancy between ULSMs calls for advancing the urban hydrology representation in ULSMs, which is essential for correct simulation of evaporative cooling in cities.

We developed a microscale airflow modeling system with detailed building and terrain data to better understand the urban microclimate. Building shapes and heights, and terrain elevation data were integrated to construct a high-resolution urban surface geometry. The system, based on computational fluid dynamics using OpenFOAM, can resolve complex flow structures around built environments. Inflow boundary conditions were generated using logarithmic wind profiles derived from Automatic Weather System (AWS) observations under neutral stability. After validation with wind-tunnel data for a single block, the system was applied to airflow modeling around a university campus in Seoul using AWS data from four nearby stations. The results demonstrated that the system captured key flow characteristics such as channeling, wake, and recirculation induced by complex terrain and building configurations. In particular, easterly inflow cases with high-rise buildings on the leeward side of a mountain exhibited intensified wakes and internal recirculations, with elevated centers influenced by tall structures. This modeling framework, with further development, could support diverse urban applications for microclimate and air quality, facilitating urban resilience.

Understanding the characteristics of wind at pedestrian level is important to ensure pedestrian safety and comfort in urban environments. Pedestrian-level wind characteristics are mainly affected by the average characteristics of surrounding buildings such as plan and frontal area densities and mean building height, and the shape of each building also has a big impact on the wind characteristics. In particular, the various shapes of buildings, including curved surfaces, complicate the wind environment in urban areas. Currently, various computational fluid dynamics (CFD) models are being utilized to study the complex flows in urban areas. Many CFD models use cartesian grids to simulate flows in urban space by converting the urban space made up of buildings into a set of cuboids. The OpenFOAM (Open Field Operation And Manipulation) model, an open source CFD model, has the advantage of being able to represent the curved shape of building to the greatest extent possible by generating unstructured meshes using the mesh generation utility called snappyHexMesh.In this study, we simulate the campus space of University of Seoul located in Seoul, South Korea using structured (cartesian grid) and unstructured meshes for OpenFOAM simulations, and compare the results over the two meshes. To implement realistic flow conditions, a mesoscale weather model (WRF) are used as the initial and lateral boundary conditions. In this study, we analyze the differences in the generation of turbulent eddies specially around the curved building surfaces according to the two types of meshes and compare the simulation results with observations. We expect that using the unstructured mesh will allow us to more accurately simulate flow separation, wakes around buildings, and turbulence statistics at pedestrian level. This study can be used to identify detailed features of pedestrian-level wind (e.g., gustiness), thereby providing a basis for pedestrian environmental impact assessment.

We investigated the structures of the wintertime lower troposphere in Seoul, South Korea on 17 and 18 January 2017 by performing 1 h interval radiosonde observation and numerical simulations. In the daytime on 17 January, the height of the convective boundary layer (CBL) sharply and quickly increased when the residual layer became a part of the CBL. From the afternoon on 17 January, moist air with clouds began to substantially intrude in the lower troposphere in Seoul, and radiative heating/cooling weakened. As a result, the mixing of air in the lower troposphere was inhibited and the vertical gradients of potential temperature and water vapor mixing ratio changed little on 18 January. We evaluated the performance of four planetary boundary layer (PBL) parameterization schemes (the Yonsei University (YSU), Mellor–Yamada–Janjić (MYJ), Mellor–Yamada–Nakanishi–Niino (MYNN), and Asymmetric Convective Model version 2 (ACM2) schemes) coupled with the Weather Research and Forecasting model in simulating the structures of the lower troposphere against 1 h interval radiosonde observation. The general tendencies of the air temperature and wind speed in the lower troposphere were well-reproduced in the four simulations. However, the sharp increase in the CBL height did not appear in the four simulations, implying that the process of the residual layer becoming a part of the CBL in the daytime is not well-parameterized. Additionally, the simulated water vapor mixing ratio near the surface was smaller compared with the observation. We found that small-scale turbulence in the CBL, which mixes advected air and pre-existing air, was not reproduced well by the PBL parameterization schemes. Compared with the other simulations, the most accurate air temperature and wind speed were reproduced in the simulation with the MYJ scheme, while the CBL development and moisture advection were reproduced relatively well in the simulation with the MYNN scheme.

In this study, the PM2.5, prediction performances of long short-term memory (LSTM), gated recurrentunit (GRU), and bidirectional LSTM (Bi-LSTM) were compared using data from Seoul, Daejeon, andBusan, which are representative cities in Korea. The data analysis period was from 9:00 on May 16,2014, to 23:00 on December 31, 2021, based on data at 1 h intervals. The causal factors affecting thechange in PM2.5 of three cities in Korea, and five major cities in China were determined. The analysisrevealed that the three models showed similarly high performances in short-term prediction within24 h (R2 0.9). The Bi-LSTM model using both past and future time information showed high predictionaccuracy for long-term prediction (R2 0.6). Using the PM2.5 data of the five major Chinese cities, it wasconfirmed that the accuracy of the PM2.5 prediction model for Seoul, Daejeon, and Busan improved. Thedeep learning model showed a high accuracy even when the Fine Dust Act measures were implemented. This study can facilitate governments to prepare measures against air pollution with a high regional predictionperformance by identifying the causal factors affecting PM2.5, specific to the city, and designingdifferent models for each forecasting period.