Identifying key convection-sensitive oceanic regions to weaken the ENSO spring predictability barrier

Area of Science:

• Climate Science
• Oceanography
• Atmospheric Science

Background:

• The El Niño-Southern Oscillation (ENSO) shows weakest predictability during boreal spring due to the Spring Predictability Barrier (SPB).
• The SPB is caused by weak air-sea coupling, limiting ENSO signal development and persistence.
• Accurate springtime ENSO prediction necessitates identifying oceanic regions crucial for convection variability.

Purpose of the Study:

• To introduce a novel Sea Surface Temperature Range Index (SRI) for quantifying convection-favorable sea surface temperatures.
• To identify critical oceanic regions in the Pacific and Atlantic that initiate persistent convection during spring.
• To enhance ENSO predictability by understanding the role of convection-sensitive areas in modulating the Walker circulation.

Main Methods:

• Development and application of the Sea Surface Temperature Range Index (SRI).
• Analysis of sea surface temperatures exceeding 26°C (east-central Pacific) and 28.5°C (eastern Atlantic) during spring.
• Implementation of a Long Short-Term Memory (LSTM) deep learning model incorporating the SRI.

Main Results:

• The SRI effectively quantifies spatial extents of convection-sensitive sea surface temperatures.
• Expansion of these critical areas strengthens the Bjerknes feedback and modulates the Walker circulation.
• The LSTM model with SRI demonstrated superior predictive skill compared to traditional models, particularly for multiyear La Niña events.

Conclusions:

• Convection-sensitive oceanic regions play a pivotal role in overcoming the Spring Predictability Barrier.
• The SRI provides an effective predictor for ENSO evolution, especially during the critical spring transition.
• Integrating SRI into deep learning models offers a promising approach for improving seasonal climate predictions.