EUVIMEDEUVIMED entdecken
Crossref

XGBoost-Based Corrections for Road Surface Temperature Forecasts

Virve Karsisto, Leila Hieta, Matti Kämäräinen

Journal of Applied Meteorology and Climatology · 2026-06 · 65 (6) · 881-897

DOI: 10.1175/jamc-d-25-0189.1 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Applied Meteorology and Climatology, 2026-06. Verfasst von Virve Karsisto und 2 weiteren Autor:innen. Die Arbeit nennt 64 Literaturangaben. Gefördert von: Business Finland, European Union, Interrge Nord.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate road surface temperature forecasts are essential for planning road salting and keeping roads free of ice. However, these forecasts often contain systematic errors due to various factors. This study presents a postprocessing methodology aimed at improving forecast accuracy, especially for temperatures near 0°C, where even small errors can be critical for road safety. The methodology was tested to correct systematic errors in the Finnish Meteorological Institute’s road surface temperature forecasts. An extreme gradient boosting (XGBoost) model was trained using data from several years and approximately 400 road weather stations to learn patterns between forecasted weather variables and road surface temperature errors. A distinctive approach in this study was to apply sample-specific weights during model training, which gave more emphasis to cases where the surface temperature was near 0°C. Another XGBoost model without weights was trained to perform well across all temperatures. The models were evaluated using data from September 2024 to May 2025. Based on scores calculated over the entire period, the unweighted model reduced the mean absolute error (MAE) by 3%–23% compared to original road weather model results, depending on the forecast lead time that extended up to 62 h. For cases where the observed surface temperature was between −3° and 3°C and the lead time was 12 h or less, the weighted model achieved a MAE reduction of 17%–34%. In conclusion, both models performed well, with the weighted model showing particular effectiveness in improving forecast accuracy at near 0° temperatures. Significance Statement This study aims to improve the accuracy of road surface temperature forecasts using machine learning. Accurate forecasts are important because they help in timing road salting operations effectively, keeping roads free of ice and preventing unnecessary salting. The machine learning model developed for forecasting near 0°C temperatures was able to reduce forecast errors by 17%–34% in that temperature range. These models have strong potential to enhance future road weather forecasts in Finland, and the same methodology can be applied in other countries as well.

Förderung

  • Business Finland
  • European Union, Interrge Nord

Zitieren

APA

Karsisto, V., Hieta, L., & Kämäräinen, M. (2026). XGBoost-Based Corrections for Road Surface Temperature Forecasts. Journal of Applied Meteorology and Climatology, 65(6), 881-897. https://doi.org/10.1175/jamc-d-25-0189.1

Vancouver

Karsisto V, Hieta L, Kämäräinen M. XGBoost-Based Corrections for Road Surface Temperature Forecasts. Journal of Applied Meteorology and Climatology. 2026;65(6):881-897. doi:10.1175/jamc-d-25-0189.1

BibTeX
@article{Karsisto2026,
  author = {Karsisto, Virve and Hieta, Leila and Kämäräinen, Matti},
  title = {XGBoost-Based Corrections for Road Surface Temperature Forecasts},
  journal = {Journal of Applied Meteorology and Climatology},
  year = {2026},
  volume = {65},
  number = {6},
  pages = {881--897},
  doi = {10.1175/jamc-d-25-0189.1}
}
Kostenlos starten →