Exploring the Pantanal's Hydrological Past: An Artificial Intelligence Approach for Modeling of EWH with In Situ River Flow Data

• Grant number: 24/03093-4
• Support Opportunities: Scholarships abroad - Research Internship - Master's degree
• Start date: September 01, 2024
• End date: February 28, 2025
• Field of knowledge: Physical Sciences and Mathematics - Geosciences - Geology
• Principal Investigator: Didier Gastmans
• Grantee: Bruno Henrique Valdambrini Vieira
• Supervisor: Alexandra Richey Mclarty
• Host Institution: Centro de Estudos Ambientais (CEA). Universidade Estadual Paulista (UNESP). Campus de Rio Claro. Rio Claro , SP, Brazil
• Institution abroad: Washington State University (WSU), United States
• Associated to the scholarship: 23/10474-1 - Impacts of Seasonal Variation in Groundwater Availability on the Hydrological Regime of the Pantanal Basin, BP.MS

Abstract

The hydrological dynamics of the Pantanal, vital for ecosystems and local communities, has had significant impacts due to climate change and economic exploitation, especially by agribusiness. This research internship project abroad proposes the application of the artificial intelligence technique, Long Short-Term Memory (LSTM), to analyze the relationship between river flow and variations in the total column of water stored in the subsurface, using data from GRACE and going back to the 1970s. Preliminary results indicate a significant reduction in the last six years for these two variables. Given this scenario, we intend to apply LSTM to fill gaps in the data and improve the understanding of the relationship between water column variation signals and In Situ river flow data. This aims to cover earlier periods where GRACE data is not available, but we rely on in situ information. The intention is to deepen the understanding of hydrological evolution over the last five decades, explore the recent drop in water indices, identify interdecadal patterns, and quantify possible future trends.