ORIGINAL RESEARCH article
Front. Water
Sec. Water and Artificial Intelligence
This article is part of the Research TopicArtificial Intelligence Applications to Water Quality ModelingView all 3 articles
Lake Titicaca Water Level Forecasting using Data Augmentation and Recurrent Neural Networks
Provisionally accepted
- Universidad Nacional de Moquegua, Perane, Peru
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Lake Titicaca, located on the border between Peru and Bolivia, is the highest navigable lake in the world and holds great environmental, cultural, social, and economic importance, both for Peru and Bolivia as well as for the Andean region in general. In this study, various artificial intelligence models based on recurrent neural networks, including LSTM, BiLSTM, GRU, and BiGRU with data augmentation, are analyzed to predict the water levels of Lake Titicaca. Data augmentation enriches the historical records and enhances the quality of the model predictions. The accuracy of the forecasts is crucial, as it contributes to proper water resource management, community safety, and ecosystem preservation. The experimental results show that all the implemented models benefit significantly from data augmentation, outperforming models reported in the literature, with GRU achieving the highest accuracy in its predictions, obtaining its best RMSE of 0.0182 m and an R² of 0.9986, surpassing LSTM, BiLSTM, and BiGRU.
Keywords: Data Augmentation4, Forecasting3, Lake Titicaca1, recurrent neural networks5, Water Leve2
Received: 19 Aug 2025; Accepted: 04 Feb 2026.
Copyright: © 2026 Flores, Guzman-Valdivia, Morales-Gonzales, Tito-Chura and Cuentas-Toledo. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Anibal Flores
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.