Assessing urban slab model parameterization in the global ICON-NWP
Abstract
Urban areas can strongly influence mesoscale atmospheric processes in its surroundings. Despite over 55% of the world’s population living in urban areas, the integration of urban parameterizations into national meteorological and hydrological service models has only recently begun. The incorporation of an urban representation within numerical weather prediction (NWP) models is assumed to have the potential to enhance the forecast accuracy and to provide an improved data basis for urban planning. In this study, We investigate the applicability of the urban slab model TERRA_URB to the global operational ICON-NWP at the German national meteorological service with 13 km resolution by simulating two five-day periods in July 2021 and June 2022. The ICON model is a widely used modeling framework but in its operational global configuration, urban areas are represented solely as a distinct soil type. The June 2022 simulation period is characterized by synoptical conditions favorable for a strong urban heat island development, while July 2021 is characterized by more overcast conditions, including several precipitation events. A comparison with 2 m air temperature observations from the urban climate station in Berlin, Alexanderplatz, revealed reduced root mean square errors when including TERRA_URB in ICON-NWP by 26% (from 1.9 to 1.4 for a five-day lead time in July 2021, and from 4.2 to 3.1 in June 2022). The implementation of TERRA_URB thus results in a quantifiable improvement in the global ICON-NWP runs, with results that are closer to the theory of urban climate. The study finds that urban weather forecasts are less accurate under strong than weak canopy urban heat island conditions, indicating room for improvement.
Related
- Assessing the potential to scale-up urban agriculture in the Global North
- Global monsoon in ICON: the scale-dependent response of Northern Hemisphere monsoons
- Assessing demand-side policy impacts on US urban passenger transport
- TEPI: A transferable street-level framework for assessing urban thermal equity under data constraints
- A remote sensing-based surface thermal stress index for assessing urban heat dynamics in Greater Cairo
BibTeX
@article{bohmann2026assessing,
title = {{Assessing urban slab model parameterization in the global ICON-NWP}},
author = {May Bohmann and Arianna Valmassoi and Patrick Ludwig and Joaquim G. Pinto and Leonie Esters},
journal = {Meteorologische Zeitschrift},
year = {2026},
month = aug,
doi = {10.1127/metz/202519},
url = {https://doi.org/10.1127/metz/202519},
note = {CC BY-NC},
}
RIS
TY - JOUR
TI - Assessing urban slab model parameterization in the global ICON-NWP
AU - May Bohmann
AU - Arianna Valmassoi
AU - Patrick Ludwig
AU - Joaquim G. Pinto
AU - Leonie Esters
PY - 2026
DA - 2026/08/01
JO - Meteorologische Zeitschrift
DO - 10.1127/metz/202519
UR - https://doi.org/10.1127/metz/202519
AB - Urban areas can strongly influence mesoscale atmospheric processes in its surroundings. Despite over 55% of the world’s population living in urban areas, the integration of urban parameterizations into national meteorological and hydrological service models has only recently begun. The incorporation of an urban representation within numerical weather prediction (NWP) models is assumed to have the potential to enhance the forecast accuracy and to provide an improved data basis for urban planning. In this study, We investigate the applicability of the urban slab model TERRA_URB to the global operational ICON-NWP at the German national meteorological service with 13 km resolution by simulating two five-day periods in July 2021 and June 2022. The ICON model is a widely used modeling framework but in its operational global configuration, urban areas are represented solely as a distinct soil type. The June 2022 simulation period is characterized by synoptical conditions favorable for a strong urban heat island development, while July 2021 is characterized by more overcast conditions, including several precipitation events. A comparison with 2 m air temperature observations from the urban climate station in Berlin, Alexanderplatz, revealed reduced root mean square errors when including TERRA_URB in ICON-NWP by 26% (from 1.9 to 1.4 for a five-day lead time in July 2021, and from 4.2 to 3.1 in June 2022). The implementation of TERRA_URB thus results in a quantifiable improvement in the global ICON-NWP runs, with results that are closer to the theory of urban climate. The study finds that urban weather forecasts are less accurate under strong than weak canopy urban heat island conditions, indicating room for improvement.
SN - 0941-2948
SN - 1610-1227
ER -