APPLICATION OF THE SPEEDY MODEL FOR THE SIMULATION OF EL NIÑO EVENTS BASED ON CATEGORIES AND ENSEMBLE
ENOS; Climate Simplified; Ensemble forecast; Score; ROC
Several factors control the Earth's climate and regulate its variations. To study some of these factors, numerical modeling is an indispensable tool. Many atmospheric teleconnections, associated with disturbances that occur in a specific area of the globe and cause climate effects in other distant regions, are linked to Sea Surface Temperature (SST) anomalies. One of the most well-known and widely studied events is the El Niño-Southern Oscillation (ENSO), which has positive and negative phases, both related to changes in SST. El Niño causes droughts in the equatorial part of South America, as well as excessive rainfall in southern Brazil, northern Argentina, and Uruguay during spring and summer. To simulate these events, the SPEEDY model will be used, applying sub-seasonal forecasts and ensemble predictions with 100 members for each forecast set. Three different regions of Brazil—North, Northeast, and South—which are considered regions with high predictability for models, will be the focus of the study. For this purpose, three distinct El Niño events were selected, one for each intensity category based on the Oceanic Niño Index (ONI). To verify the simulation results, three atmospheric variables will be used: precipitation data from NOAA's Climate Prediction Center (CPC), air temperature, and wind components from ECMWF (ERA-5) data. As metrics for verifying the skill of the ensemble forecasts, the Brier Score (BS), Mean Continuous Ranked Probability Score (CRPS), and Relative Operating Characteristic (ROC) diagram and Score will be used.