Representative researches covering ENSO prediction methods, the spring predictability barrier, and localized impacts of El Niño on global climate networks, combining complexity metrics, climate networks and physics-informed machine learning.
ENSO is one of the most influential modes of interannual climate variability. It affects the tropical Pacific and, through teleconnections, modulates rainfall, extremes, and climate risks across the globe.
The group approaches ENSO from two complementary directions: improving long-lead prediction across the spring predictability barrier, and identifying how El Nino impacts are organized across the global climate network. Complexity indicators and machine learning support earlier forecasts, while climate-network analysis helps explain where and how ENSO impacts are organized.
Normalized System Sample Entropy Linear Forecast Model
A NorSysSampEn-based linear forecast method for the 2026 El Nino peak ONI, combining ERA5/ONI data, preprocessing, parameter search, model construction, and the corresponding forecast results with uncertainty expression.
Enhancing the Predictability Limits of ENSO with Physics-Guided Deep Echo State Networks
An ENSO real-time forecasting method based on ORAS5 reanalysis data and physics-guided deep echo state networks (DESN), combining climate-mode index construction, seasonal-cycle and trend removal, ensemble training, member selection, and Niño3.4 forecast results for 2026-2027.
Tropical monsoon rainfall can be predicted with lead times up to 10 months
This study verifies that tropical monsoon rainfall has a predictability lead time of up to 10 months, extending the theoretical limit of interannual monsoon climate prediction and providing critical support for long-range cross-seasonal climate forecasting.
Complexity-based approach for El Niño magnitude forecasting before the spring predictability barrier
Complexity metrics are adopted to characterize the system state over the Niño3.4 region and predict the intensity of El Niño events prior to the spring predictability barrier.
