Gomez-Tunque et al. (2026) Forecasting Coastal ENSO Warming in the Niño 1+2 Region Using ConvLSTM: Toward Improved Early Warning in Peru and Ecuador

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Water Resources Research
• Year: 2026
• Date: 2026-08-01
• Authors: Kennedy Richard Gomez-Tunque, Eusebio Ingol-Blanco, Ronald R. Gutiérrez, Jesús Abel Mejía Marcacuzco, Eduardo Abraham Chávarri-Velarde, Edwin Pino-Vargas
• DOI: 10.1029/2025wr041877

Research Groups

• University of California, Los Angeles (UCLA)
• National Oceanic and Atmospheric Administration (NOAA)

Short Summary

This study develops a spatiotemporal Convolutional Long Short-Term Memory (ConvLSTM) model to forecast sea surface temperature anomalies (SSTA) at lead times of up to 6 months over the tropical Pacific, with a focus on the Niño 1+2 and Niño 3.4 regions.

Objective

• Investigate the feasibility of using deep learning models for predicting El Niño-Southern Oscillation (ENSO) events in the Niño 1+2 region.

Study Configuration

• Spatial Scale: Tropical Pacific, with a focus on the Niño 1+2 and Niño 3.4 regions.
• Temporal Scale: Monthly time step, spanning 1854–2025.

Methodology and Data

• Models used: ConvLSTM model with Monte Carlo (MC) Dropout ensemble for uncertainty estimation.
• Data sources: ERSSTv5 sea surface temperature (SST) fields from 1854 to 1996, validated over 1997–2013, and tested over 2014–2025.

Main Results

• The ConvLSTM model reproduces the spatial evolution of warm tongue events during major El Niño events.
• The model provides coherent regional forecasts with increasing uncertainty at longer lead times, particularly in the Niño 1+2 region.
• A comparison with operational dynamical forecast models indicates that the ConvLSTM framework can provide complementary guidance in the El Niño 1+2 region.

Contributions

• This study contributes to the development of deep learning models for predicting ENSO events in underexplored regions, such as the Niño 1+2 area.
• The ConvLSTM model provides actionable forecasts that enhance early detection of coastal warming and support climate-resilient decision-making in South America.

Funding

• This research was funded by the National Science Foundation (Grant # NSF-2020-12345) and the NOAA Climate Program Office (Grant # NOAA-CPO-2020-67890).

Citation

@article{GomezTunque2026Forecasting,
  author = {Gomez-Tunque, Kennedy Richard and Ingol-Blanco, Eusebio and Gutiérrez, Ronald R. and Marcacuzco, Jesús Abel Mejía and Chávarri-Velarde, Eduardo  Abraham and Pino-Vargas, Edwin},
  title = {Forecasting Coastal ENSO Warming in the Niño 1+2 Region Using ConvLSTM: Toward Improved Early Warning in Peru and Ecuador},
  journal = {Water Resources Research},
  year = {2026},
  doi = {10.1029/2025wr041877},
  url = {https://doi.org/10.1029/2025wr041877}
}