The results of this work, published in the Journal of Ocean & Coastal Management, highlight an innovative use of artificial intelligence to predict sea level changes along the Gulf of Texas coast, a region where some of the country’s largest ports and communities are located, increasingly vulnerable to flooding, land subsidence and rising sea levels.
"The need to make predictions for the coming weeks and months is growing in importance as the frequency of floods along our coast increases," says Philippe Tissot, co-lead researcher at AI2ES (Trustworthy AI in Weather, Climate, and Coastal Oceanography) and Professor of Coastal AI at the Conrad Blucher Institute of Topography and Science (CBI) at Texas A&M-Corpus Christi University TAMU-CC. "Traditional forecasting methods are usually designed for short-term conditions, such as tides or storms, covering periods of only a few hours to a few days. This work focuses on extending these deadlines to months and even years, providing a complementary perspective for improving preparedness," the scientist adds.
To create the AI model, research teams combined decades of tidal and satellite data with an advanced regional grouping statistical method developed by Veronica Nieves, director of the AI4OCEANS group at the Image Processing Laboratory (LPI), located in the scientific-academic area of the University of Valencia Science Park (PCUV), and co-principal investigator of the project, and Cristina Radin, then PhD student in the UV. "Capturing a shared regional signal was essential to understanding how large-scale ocean patterns affect all coastal stations," says Veronica Nieves. "This approach allowed us to identify the broader processes shaping sea level variability along the Texas coast".
"Capturing a shared regional signal was essential to understanding how large-scale ocean patterns affect all coastal stations. This approach allowed us to identify the broader processes shaping sea level variability along the Texas coast", LPI researcher Veronica Nieves
Cristina Radin and Marina Vicens-Miquel, then a doctoral researcher at CBI, trained the AI to recognize patterns on the Texas coast that influence rising and falling sea levels over time. "The AI model was able to learn patterns that can help predict sea levels months and even years in advance," says Vicens-Miquel. "This can provide coastal regions with earlier indicators and help guide planning for the coming years, especially in areas where data is limited," adds Radin.
According to the study, the AI model, which is extrapolable to other coastal regions, generated forecasts with greater accuracy and reliability than existing predictive methods in the region. This provides valuable time for coastal planners and resource managers to anticipate periods of higher water levels and support longer-term management decisions.
Beyond its scientific results, the project highlights the importance of international collaboration in addressing complex environmental challenges. Both AI2ES and AI4OCEANS are globally recognized for promoting the use of AI in ocean and climate research. At a time of growing challenges for research funding, the success of this collaboration demonstrates how intercontinental cooperation can drive innovation and strengthen coastal resilience around the world.
Supported by an initiative of the National Science Foundation (NSF), the project involves the AI4OCEANS group from the University of Valencia (UV) in Spain and a team of researchers from the NSF AI2ES Institute (AI Institute for Research on Trustworthy AI In Weather, Climate, and Coastal Oceanography), affiliated with Texas A&M-Corpus Christi University (TAMU-CC).
Source: UV News