EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Machine Learning na Previsão de Níveis Hídricos Urbanos

Gustavo Antunes Vieira Alves, Lucio Silva de Souza

Ensaios de Geografia · 04.11.2026 · 13 (26) · e132617

DOI: 10.22409/njsfs591 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Ensaios de Geografia, online erschienen am 04.11.2026. Verfasst von Gustavo Antunes Vieira Alves und 1 weiteren Autor:innen. Die Arbeit nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The intensification of extreme hydrometeorological events and accelerated urbanization increase flood vulnerability in the city of Rio de Janeiro, particularly in fast-response watersheds. This study evaluates the application of the XGBoost algorithm to predict water level in the Acari River (RJ) across forecast horizons of 30 to 300 minutes, using streamflow and rainfall data from 2010 to 2016 standardized to a 5-minute temporal resolution. The modeling was structured around predicting the level variation (Δlevel), incorporating time lags, cumulative precipitation and cyclical features, with hyperparameter tuning performed via the Optuna library. Results showed high performance at very short horizons, with R² of 0.986 and NSE of 0.985 for the 30-minute horizon, remaining satisfactory up to approximately 120 minutes, with progressive degradation and a tendency to underestimate peaks at longer horizons. It is concluded that XGBoost shows strong potential for short-term hydrological forecasting in urban basins, representing a promising tool for early warning systems, provided its limitations at longer horizons are taken into account.

Zitieren

APA

Alves, G. A. V., & Souza, L. S. D. (2026). Machine Learning na Previsão de Níveis Hídricos Urbanos. Ensaios de Geografia, 13(26), e132617. https://doi.org/10.22409/njsfs591

Vancouver

Alves GAV, Souza LSD. Machine Learning na Previsão de Níveis Hídricos Urbanos. Ensaios de Geografia. 2026;13(26):e132617. doi:10.22409/njsfs591

BibTeX
@article{Alves2026,
  author = {Alves, Gustavo Antunes Vieira and Souza, Lucio Silva de},
  title = {Machine Learning na Previsão de Níveis Hídricos Urbanos},
  journal = {Ensaios de Geografia},
  year = {2026},
  volume = {13},
  number = {26},
  pages = {e132617},
  doi = {10.22409/njsfs591}
}
Kostenlos starten →