Reinforcement Learning to Support Monthly Precipitation Prediction. Case Study: Department of Boyacá - Colombia

Zea Gutiérrez, Jimmy Alejandro; Suárez Barón, Marco Javier; González Sanabria, Juan Sebastián

Reinforcement Learning to Support Monthly Precipitation Prediction. Case Study: Department of Boyacá - Colombia

dc.creator	Zea Gutiérrez, Jimmy Alejandro
dc.creator	Suárez Barón, Marco Javier
dc.creator	González Sanabria, Juan Sebastián
dc.date	2024-06-27
dc.date.accessioned	2025-10-01T23:53:12Z
dc.description	The emission of greenhouse gases, directly or indirectly attributed to human activity, is the main cause of global climate change. Among the gases emitted, carbon dioxide (CO2) is the most important contributor to the spatio-temporal variation of physical quantities such as relative humidity, atmospheric pressure, ambient temperature and, most significantly, precipitation. The objective of the research was to present an analysis of the prediction of monthly precipitation in the department of Boyacá using models based on reinforced learning (RL). The methodology used consisted of extracting data from CHIRPS 2.0 (Climate Hazards Group InfraRed Precipitation with Station data, version 2.0) with a spatial resolution of 0.05 ° that were subsequently preprocessed for the implementation of Monte Carlo simulation and deep reinforced learning (DRL) approaches to provide monthly precipitation predictions. The results obtained showed that Monte Carlo simulation such as DRL generate meaningful predictions of monthly precipitation. It is essential to recognize that conventional models based on Deep Learning, such as Short-Term Memory (LSTM) or Short-Term Convolutional Networks (ConvLSTM), can outperform Monte Carlo and DRL simulation approaches in terms of prediction accuracy. It is concluded that the implementation of reinforcement learning techniques in monthly precipitation prediction models detects information patterns that can be used to support strategies aimed at mitigating economic and social risks derived from climate phenomena.	en-US
dc.description	La emisión de gases de efecto invernadero, atribuida directa o indirectamente a la actividad humana, es la principal causa del cambio climático a nivel mundial. Entre los gases emitidos, el dióxido de carbono (CO2) es el que más contribuye a la variación espacio temporal de magnitudes físicas como la humedad relativa, la presión atmosférica, la temperatura ambiente y, de manera más significativa, la precipitación. El objetivo de la investigación fue presentar un análisis de la predicción de la precipitación mensual en el departamento de Boyacá mediante el uso de modelos basados en aprendizaje reforzado (RL, por sus siglas en inglés). La metodología empleada consistió en extraer datos desde CHIRPS 2,0 (Climate Hazards Group InfraRed Precipitation with Station data, versión 2,0) con una resolución espacial de 0,05° que posteriormente fueron preprocesados para la implementación de enfoques basados en una simulación Montecarlo y aprendizaje reforzado profundo (DRL, por sus siglas en inglés) para proporcionar predicciones de la precipitación mensual. Los resultados obtenidos demostraron que la simulación Montecarlo como el DRL generan predicciones significativas de la precipitación mensual. Es esencial reconocer que los modelos convencionales basados en Aprendizaje profundo, como Memoria a Corto Plazo (LSTM) o Redes Convolucionales a Corto Plazo (ConvLSTM), pueden superar a los enfoques de simulación Montecarlo y DRL en términos de precisión de predicción. Se concluye que la implementación de técnicas de aprendizaje por refuerzo en modelos de predicción de la precipitación mensual detecta patrones de información que pueden ser usados como soporte a estrategias dirigidas a mitigar los riesgos económicos y sociales derivados de fenómenos climáticos.	es-ES
dc.format	application/pdf
dc.format	text/xml
dc.format	application/zip
dc.format	text/html
dc.identifier	https://revistas.itm.edu.co/index.php/tecnologicas/article/view/3017
dc.identifier	10.22430/22565337.3017
dc.identifier.uri	https://hdl.handle.net/20.500.12622/7904
dc.language	spa
dc.publisher	Instituto Tecnológico Metropolitano (ITM)	es-ES
dc.relation	https://revistas.itm.edu.co/index.php/tecnologicas/article/view/3017/3286
dc.relation	https://revistas.itm.edu.co/index.php/tecnologicas/article/view/3017/3426
dc.relation	https://revistas.itm.edu.co/index.php/tecnologicas/article/view/3017/3532
dc.relation	https://revistas.itm.edu.co/index.php/tecnologicas/article/view/3017/3533
dc.relation	/ref/Esri. “Análisis y predicción de meteorología.” arcgis.com. Accessed: Feb. 19, 2024. [Online]. Available: https://learn.arcgis.com/es/paths/meteorology/
dc.relation	/ref/E. Mesgari, P. Mahmoudi, Y. Kord Tamandani, T. Tavousi, and S. M. Amir Jahanshahi, “A comparative assessment of the ability of different types of machine learning in short-term predictions of nocturnal frosts,” Acta Geophysica, vol. 72, no. 4, pp. 2955–2973, Aug. 2024. https://doi.org/10.1007/S11600-023-01276-1
dc.relation	/ref/IDEAM. “Predicción-IDEAM.” ideam.gov.co. Accessed: Feb. 19, 2024. [Online]. Available: http://ideam.gov.co/web/siac/prediccion
dc.relation	/ref/S. Zi-yi, and B. Wen-chao, “Machine learning model combined with CEEMDAN algorithm for monthly precipitation prediction,” Earth Sci. Inform., vol. 16, no. 2, pp. 1821–1833, Jun. 2023. https://doi.org/10.1007/S12145-023-01011-W
dc.relation	/ref/A. Thamm GmbH. “Become a data.musketeer.” alexanderthamm.com. Accessed: Feb. 19, 2024. [Online]. Available: https://www.alexanderthamm.com/de/karriere/
dc.relation	/ref/M. Akbarian, B. Saghafian, and S. Golian, “Monthly streamflow forecasting by machine learning methods using dynamic weather prediction model outputs over Iran,” Journal of Hydrology, vol. 620, p. 129480, May. 2023. https://doi.org/10.1016/J.JHYDROL.2023.129480
dc.relation	/ref/J. Nithyashri, R. K. Poluru, S. Balakrishnan, M. Ashok Kumar, P. Prabu, and S. Nandhini, “IOT based prediction of rainfall forecast in coastal regions using deep reinforcement model,” Measurement: Sensors, vol. 29, p. 100877, Oct. 2023. https://doi.org/10.1016/J.MEASEN.2023.100877
dc.relation	/ref/Y. Jun-He, C. Ching-Hsue, and C. Chia-Pan, “A time-series water level forecasting model based on imputation and variable selection method,” Comput. Intell. Neurosci., vol. 2017, pp. 1-11, Nov. 2017. https://doi.org/10.1155/2017/8734214
dc.relation	/ref/R. Yang, H. Liu, N. Nikitas, Z. Duan, Y. Li, and Y. Li, “Short-term wind speed forecasting using deep reinforcement learning with improved multiple error correction approach,” Energy, vol. 239, p. 122128, Jan. 2022. https://doi.org/10.1016/J.ENERGY.2021.122128
dc.relation	/ref/M. Zhang, Z. Xu, Y. Wang, S. Zeng, and X. Dong, “Evaluation of uncertain signals’ impact on deep reinforcement learning-based real-time control strategy of urban drainage systems,” J. Environ. Manage., vol. 324, p. 116448, Dec. 2022. https://doi.org/10.1016/J.JENVMAN.2022.116448
dc.relation	/ref/K. Skarlatos, E. S. Bekri, D. Georgakellos, P. Economou, and S. Bersimis, “Projecting Annual Rainfall Timeseries Using Machine Learning Techniques,” Energies 2023, Vol. 16, Page 1459, vol. 16, no. 3, p. 1459, Feb. 2023, https://doi.org/10.3390/EN16031459
dc.relation	/ref/P. D. Nolasco Ramírez, “Aplicación de Machine Learning para pronóstico de desplazamiento de lluvias usando imágenes del radar de lluvias de UDEP,” Trabajo de Pregrado, Universidad de Piura, Piura, Perú, 2023. [Online]. Available: https://hdl.handle.net/11042/6007
dc.relation	/ref/R. He, L. Zhang, and A. W. Z. Chew, “Data-driven multi-step prediction and analysis of monthly rainfall using explainable deep learning,” Expert Systems with Applications, vol. 235, p. 121160, Jan. 2024. https://doi.org/10.1016/J.ESWA.2023.121160
dc.relation	/ref/A. U. G. Senocak, M. T. Yilmaz, S. Kalkan, I. Yucel, and M. Amjad, “An explainable two-stage machine learning approach for precipitation forecast,” Journal of Hydrology, vol. 627, p. 130375, Dec. 2023. https://doi.org/10.1016/J.JHYDROL.2023.130375
dc.relation	/ref/C. Wang, J. Xu, G. Tang, Y. Yang, and Y. Hong, “Infrared Precipitation Estimation Using Convolutional Neural Network,” IEEE Transactions on Geoscience and Remote Sensing, vol. 58, no. 12, pp. 8612–8625, Dec. 2020. https://doi.org/10.1109/TGRS.2020.2989183
dc.relation	/ref/Q. Yuan et al., “Deep learning in environmental remote sensing: Achievements and challenges” Remote Sensing of Environment, vol. 241, p. 111716, 2020. https://doi.org/10.1016/j.rse.2020.111716
dc.relation	/ref/T. Nan et al., “Evaluation of shallow groundwater dynamics after water supplement in North China Plain based on attention-GRU model,” Journal of Hydrology, vol. 625, p. 130085, Oct. 2023. https://doi.org/10.1016/J.JHYDROL.2023.130085
dc.relation	/ref/E. E. León Baque, V. D. Vásquez Granda, and M. D. Valderrama Chávez, “Cambios en patrones de precipitación y temperatura en el Ecuador: regiones sierra y oriente,” Dilemas contemporáneos: educación, política y valores, vol. 8, no. SPE, Mar. 2021. https://doi.org/10.46377/DILEMAS.V8I.2608
dc.relation	/ref/C. J. Fernández Pérez, G. E. Cely Reyes, and P. A. Ramírez, “Cuantificación de la captura de carbono y análisis de las propiedades del suelo en coberturas naturales y una plantación de pino en el páramo de Rabanal, Colombia,” Cuad. Geogr. Rev. Colomb. Geogr., vol. 28, no. 1, pp. 121–133, Jan. 2019. https://doi.org/10.15446/RCDG.V28N1.66152
dc.relation	/ref/E. Morales Rojas, E. A. Díaz Ortiz, L. García, and M. E. Milla Pino, “Pronóstico de precipitaciones mensuales: Un estudio de caso en las comunidades nativas del Perú,” Revista Científica Pakamuros, vol. 9, no. 3, pp. 71–85, Sep. 2021. https://doi.org/10.37787/13azmg02
dc.relation	/ref/D. M. Herrera Posada, and E. Aristizábal, “Modelo de inteligencia artificial y aprendizaje automático para la predicción espacial y temporal de eventos de sequía en el departamento del Magdalena, Colombia,” INGE CUC, vol. 18, no. 2, pp. 249–265, Nov. 2022. https://doi.org/10.17981/INGECUC.18.2.2022.20
dc.relation	/ref/E. Duarte. “Rainfall data for the department of Boyacá.” kaggle.com. Accessed: Feb. 19, 2024. [Online]. Available: https://www.kaggle.com/datasets/estebanduarte/rainfall-data-for-the-department-of-boyaca
dc.relation	/ref/M. A. Saleh, and H. M. Rasel, “Performance evaluation of Machine Learning based regression models for rainfall forecasting,” Research Square, vol. 25, Jan. 2024. https://doi.org/10.21203/RS.3.RS-3856741/V1
dc.relation	/ref/S. D. Latif, and A. N. Ahmed, “A review of deep learning and machine learning techniques for hydrological inflow forecasting,” Environment, Development and Sustainability, vol. 25, no. 11, pp. 12189–12216, Mar. 2023. https://doi.org/10.1007/S10668-023-03131-1
dc.relation	/ref/S. Eddamiri, F. Z. Bassine, V. Ongoma, T. Epule Epule, and A. Chehbouni, “An automatic ensemble machine learning for wheat yield prediction in Africa,” Multimed. Tools Appl., Jan. 2024. https://doi.org/10.1007/S11042-024-18142-X
dc.relation	/ref/D. Elavarasan, and P. M. Durairaj Vincent, “Crop Yield Prediction Using Deep Reinforcement Learning Model for Sustainable Agrarian Applications,” IEEE Access, vol. 8, pp. 86886–86901, May. 2020. https://doi.org/10.1109/ACCESS.2020.2992480
dc.relation	/ref/J. B. Valencia, V. V. Guryanov, J. Mesa-Diez, N. Diaz, D. Escobar-Carbonari, and A. V. Gusarov, “Predictive Assessment of Climate Change Impact on Water Yield in the Meta River Basin, Colombia: An InVEST Model Application,” Hydrology, vol. 11, no. 2, p. 25, Feb. 2024. https://doi.org/10.3390/HYDROLOGY11020025
dc.relation	/ref/G. Sharma, A. Singh, and S. Jain, “A hybrid deep neural network approach to estimate reference evapotranspiration using limited climate data,” Neural Comput. Appl., vol. 34, no. 5, pp. 4013–4032, Mar. 2022. https://doi.org/10.1007/S00521-021-06661-9
dc.relation	/ref/S. Duarte, G. A. Corzo Perez, G. Santos, and D. P. Solomatine, “Application of Natural Language Processing to Identify Extreme Hydrometeorological Events in Digital News Media: Case of the Magdalena River Basin, Colombia,” in Advanced Hydroinformatics: Machine Learning and Optimization for Water Resources, G. A. Corzo Perez, D. P. Solomatine, Ed., New York, NY, USA: Wiley, 2024, pp. 283–318. https://doi.org/10.1002/9781119639268.CH10
dc.relation	/ref/Z. Gao et al., “PreDiff: Precipitation Nowcasting with Latent Diffusion Models,” 2023, arXiv: 2307.10422. https://arxiv.org/abs/2307.10422v2
dc.relation	/ref/
dc.rights	Derechos de autor 2024 TecnoLógicas	es-ES
dc.rights	https://creativecommons.org/licenses/by-nc-sa/4.0	es-ES
dc.source	TecnoLógicas; Vol. 27 No. 60 (2024); e3017	en-US
dc.source	TecnoLógicas; Vol. 27 Núm. 60 (2024); e3017	es-ES
dc.source	2256-5337
dc.source	0123-7799
dc.subject	Aprendizaje automático	es-ES
dc.subject	aprendizaje reforzado	es-ES
dc.subject	modulación CHIRPS	es-ES
dc.subject	simulación Montecarlo	es-ES
dc.subject	Machine Learning	en-US
dc.subject	reinforcement learning	en-US
dc.subject	CHIRPS modulation	en-US
dc.subject	Monte Carlo simulation	en-US
dc.title	Reinforcement Learning to Support Monthly Precipitation Prediction. Case Study: Department of Boyacá - Colombia	en-US
dc.title	Aprendizaje por refuerzo como soporte a la predicción de la precipitación mensual. Caso de estudio: Departamento de Boyacá - Colombia	es-ES
dc.type	info:eu-repo/semantics/article
dc.type	info:eu-repo/semantics/publishedVersion
dc.type	Research Papers	en-US
dc.type	Artículos de investigación	es-ES