Predictive Model Prototype for Flooding in Caxias do Sul

• Eduardo Eberhardt Pereira UCS
• Daniel Luis Notari UCS

DOI: https://doi.org/10.5753/eramiars.2025.16310

Resumo

Heavy rainfall events that cause floods are natural and unavoidable, yet result in devastating damage to society and, therefore, must be studied to mitigate their impacts. This study proposes a predictive artificial intelligence model capable of classifying the occurrence of flood events using machine learning techniques, specifically, MLP neural networks. This research will be conducted in the city of Caxias do Sul, Brazil, using meteorological and hydrological data collected by Instituto Nacional de Meteorologia (INMET). The performance of the model will be evaluated using metrics such as accuracy and confusion matrix to explore recall and precision, and the explainability will be investigated using the Shapley Additive Explanations (SHAP) method.

Referências

CNM (2022). Prejuízos causados pelas chuvas no brasil entre 2017 e 2022 ultrapassam 55,5 bilhões, revela cnm.

de Sene, E. and Moreira, J. C. (2012). Geografia geral e do Brasil: espaço geográfico e globalização, volume 1. Editora Scipione, São Paulo.

FLOODsite (2005). Language of risk.

Geron, A. (2019). Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems. O’Reilly Media, Sebastopol, 2 edition.

IPCC (2023). Climate change 2023: Synthesis report.

Klemeš, V. (1983). Conceptualization and scale in hydrology. Journal of Hydrology, 65:1–23.

Marengo, J. A., Dolif, G., Cuartas, A., Camarinha, P., Gonçalves, D., Luiz, R., Silva, L., Alvava, R. C. S., Seluchii, M. E., Moraes, O. L., Soares, W. R., and Nobre, C. A. (2024). O maior desastre climático do brasil: chuvas e inundações no estado do rio grande do sul em abril-maio 2024. Estudos Avançados.

Mosavi, A., Ozturk, P., and wing Chau, K. (2018). Flood prediction using machine learning models: Literature review. Water, 10(11):1536.

Munich Re Group (2004). Topics geo annual review: Natural catastrophes 2004.

Razavi, S., Gober, P., Maier, H. R., Brouwer, R., and Wheater, H. (2020). In Anthropocene flooding: challenges for science and society, volume 34, pages 1996–2000. Hydrological Processes.

Russel, S. J. and Norvig, P. (2022). Inteligência Artificial: Uma Abordagem Moderna. gen LTC, 4 edition.

Seo, D.-J. and Breidenbach, J. P. (2002). Real-time correction of spatially nonuniform bias in radar rainfall data using rain gauge measurements. Journal of Hydrometeorology, 3(2):93–111.

Sharda, R., Delen, D., and Turban, E. (2019). Business intelligence e análise de dados para gestão do negócio. Bookman, 4 edition.

Sharma, K., Anand, D., Sabharwal, M., Tiwari, P. K., Cheikhrouhou, O., and Frikha, T. (2021). A disaster management framework using internet of things-based interconnected devices. Mathematical Problems in Engineering, 2021(2629):1–21.

Todini, E. (1988). Rainfall-runoff modeling – past, present and future. Journal of Hydrology, 100:341–352.

Tucci, C. E. M. (2004). In Hidrologia: Ciência e Aplicação. Editora da UFRGS, Porto Alegre, 4 edition.

Ukhurebor, K. E., Azi, S. O., Aigbe, U. O., Onyancha, R. B., and Emegha, J. O. (2020). Analyzing the uncertainties between reanalysis meteorological data and ground measured meteorological data. Measurement, 165:108110.

Viessman, W. J. and Lewis, G. L. (2003). Introduction to Hydrology. Prentice Hall, Upper Saddle River, New Jersey, 4 edition.

World Weather Attribution (2024). When risks become reality: Extreme weather in 2024.