Colas et al. (2026) Evaluation of snow conditions modelling for various LCZ types

Identification

• Journal: Urban Climate
• Year: 2026
• Date: 2026-01-10
• Authors: Gabriel Colas, Minttu Havu
• DOI: 10.1016/j.uclim.2025.102767

Research Groups

• Météo-France
• CNRS
• Univ. Toulouse
• CNRM

Short Summary

This study evaluates the impact of improved snow parameterizations, including a multilayer snow model and snow removal, within the Town Energy Balance (TEB) urban canopy model across 9 cold urban sites. It finds that these enhancements significantly improve the simulation of town albedo and energy fluxes, particularly in low-density urban areas, while highlighting remaining challenges for dense city centers.

Objective

• To determine the relevance of including improved snow parameterizations in urban canopy models.
• To identify the snow-related processes that have the greatest impact on model performance.
• To assess whether the urban canopy model accurately represents the evolution of albedo, which could have positive implications for weather forecasts and climate change studies.

Study Configuration

• Spatial Scale: Nine cold winter climate urban sites across America, Europe, and Asia, representing various Local Climate Zone (LCZ) types (LCZ2, LCZ3, LCZ5, LCZ6, LCZ8). The model simulates the local environment around flux towers as a single vertical profile.
• Temporal Scale: Continuous simulations including a one-year spin-up, followed by a validation period with 30-minute or hourly outputs. Albedo and flux performances were analyzed for subsets ranging from the first to the last snow flake of the season.

Methodology and Data

• Models used:
  • Town Energy Balance (TEB) urban canopy model.
  • Three configurations:
    • Reference (Ref): One-layer snow model (Masson, 2000).
    • Explicit Snow (ES): Multilayer snow model on road surfaces (Decharme et al., 2016).
    • ES + removal: Explicit Snow model combined with a snow removal parameterization (Colas et al., 2025), where 75% of the road is cleared of snow every 6 hours of continuous snow cover, with snow piled on the remaining 25%.
  • The original spectral albedo calculation in ES for natural surfaces (Decharme et al., 2016) was used.
• Data sources:
  • Urban-PLUMBER intercomparison project dataset, providing high-quality flux tower measurements.
  • Model forcing data: Downward shortwave (SW↓) and longwave radiation (LW↓), air temperature, specific humidity, air pressure, wind direction and speed, rainfall, and snowfall rates.
  • Validation data: Non-gap-filled measurements of upwelling radiation fluxes (SW↑, LW↑), sensible (Qh), and latent (Qle) heat fluxes from flux towers.
  • Land-use type fractions for each site.

Main Results

• The multilayer snow model (ES) alone marginally improved albedo simulations, primarily reducing albedo during snowmelt periods.
• The addition of the snow removal parameterization (ES + removal) further reduced simulated albedo and significantly improved upwelling solar fluxes (SW↑) for all LCZ types, confirming the importance of anthropogenic impacts.
• Despite improvements, simulated albedo remained higher than observed for most LCZ types, especially in dense city centers (LCZ2, e.g., FI-Torni and PL-Narutowicza), suggesting missing processes like pitched roof snow shedding or enhanced snow darkening.
• The model performed best for low-rise and large low-rise building setups (LCZ3, LCZ6, LCZ8).
• A faint improvement was observed for sensible heat fluxes (Qh) with ES and ES + removal.
• A performance degradation was noted for latent heat fluxes (Qle) with the new parameterizations, particularly at sites with substantial snowfall.
• While ES alone decreased upwelling infrared radiation fluxes (LW↑), ES + removal yielded similar LW↑ scores to the reference model, though the ES + removal configuration was deemed more physically sound for other surface variables.

Contributions

• Provides a comprehensive evaluation of different snow parameterization complexities (simple to multilayer with removal) in urban environments using a multi-site, multi-LCZ approach.
• Demonstrates that incorporating a multilayer snow model and a snow removal parameterization significantly enhances the accuracy of urban winter condition simulations, particularly for town albedo and associated energy fluxes.
• Identifies critical limitations of current urban canopy models in dense urban areas (LCZ2), specifically highlighting the need to account for pitched roof snow shedding and improved snow darkening parameterizations.
• Reinforces the importance of anthropogenic activities (snow removal) in accurately modeling urban climate.
• Offers insights for improving weather forecasts and climate change studies in cold cities by addressing current model deficiencies in snow representation.

Funding

Not explicitly mentioned in the provided text.

Citation

@article{Colas2026Evaluation,
  author = {Colas, Gabriel and Havu, Minttu},
  title = {Evaluation of snow conditions modelling for various LCZ types},
  journal = {Urban Climate},
  year = {2026},
  doi = {10.1016/j.uclim.2025.102767},
  url = {https://doi.org/10.1016/j.uclim.2025.102767}
}