Font size:
Print
INCOIS’s New Product to Forecast El Niño and La Niña
Context:
Recently, INCOIS (Indian National Centre for Ocean Information Services) has introduced a new forecasting product called the Bayesian Convolutional Neural Network (BCNN) to predict El Niño and La Niña conditions.
More On News :
- This product leverages advanced technologies such as Artificial Intelligence (AI), deep learning, and machine learning (ML) to enhance the accuracy and lead time of forecasts related to the phases of the El Niño Southern Oscillation (ENSO).
BCNN Working:
- BCNN predicts ENSO phases (El Niño, La Niña, neutral), crucial for agriculture, fisheries, and disaster management.
- It uses SST anomalies in the Niño3.4 region (5°N to 5°S, 170°W to 120°W) to calculate the Niño 3.4 index.
- By integrating Bayesian inference with CNNs, BCNN handles prediction uncertainty and leverages spatial SST data relationships.
- It provides early forecasts by analysing oceanic variations and atmospheric couplings, offering sufficient lead time for decision-making.
BCNN Comparison with Existing Weather Forecasting Models:
- BCNN integrates dynamic weather models with AI, enhancing ENSO phase forecasts up to 15 months ahead—outperforming traditional models by 6 to 9 months.
- It utilises historical simulations from CMIP5 and CMIP6 datasets (1850-2014) for robust training, overcoming oceanic data limitations.
Significant challenges
- Limited Oceanic Data: Scarce long-term sea surface temperature (SST) records in the central Pacific hindered BCNN training.
- Short Observation Period: Sparse SST data since 1871 limited BCNN’s accuracy, especially for rare El Niño or La Niña events.
- Complexity of ENSO Phenomenon: The intricate interactions between ocean and atmosphere in El Niño and La Niña events posed challenges for BCNN, as these phenomena involve nonlinear dynamics and complex feedback loops.
- Operational Deployment: Real-time data integration and user-friendly interfaces posed challenges in BCNN’s operational use.
What is El Niño–Southern Oscillation (ENSO) ?
- ENSO is a recurring climate pattern where temperatures in the central and eastern tropical Pacific Ocean fluctuate every 3 to 7 years. Surface waters across a large area warm or cool by 1°C to 3°C, affecting rainfall in the tropics and influencing the global weather system.
- Though ENSO is a single climate phenomenon, it has 3 phases : El Niño , La Niña and ENSO-neutral.
- El Niño :
-
- It causes warming of the ocean surface, resulting in above-average sea surface temperatures (SST) in the central and eastern tropical Pacific Ocean.
- This leads to reduced rainfall over Indonesia and increased rainfall across the central and eastern tropical Pacific Ocean.
- The usual easterly winds along the equator weaken or may reverse, occasionally becoming westerly.
- The warmer the ocean temperature anomalies, the stronger the El Niño (and vice-versa).
- La Niña :
-
- This entails cooling of the ocean surface, causing below-average sea surface temperatures (SST) in the central and eastern tropical Pacific Ocean.
- Over Indonesia, rainfall tends to increase while decreasing over the central and eastern tropical Pacific Ocean.
- The normal easterly winds along the equator intensify further.
- The cooler the ocean temperature anomalies, the stronger the La Niña (and vice-versa).
- Neutral Phase:
-
- This phase occurs when neither El Niño nor La Niña is dominant. Sea surface temperatures (SSTs) in the tropical Pacific are generally near average.
- El Niño, La Niña and Indian Monsoon Relationship :
-
- El Niño weakens trade winds across the Pacific, reducing moisture-laden monsoon winds over India, leading to reduced monsoon rainfall. Historically, at least half of the El Niño years were monsoon droughts (below -10% departure from the long-term average).
- La Niña, the cool phase of ENSO, generally enhances the strength of the trade winds, potentially increasing the moisture-laden monsoon winds over India, often leading to increased rainfall.
|
The Indian Ocean Dipole (IOD) :
Phases : Positive IOD:
Negative IOD:
|
|
ESSO-INCOIS :
|