Estimating Predictability of Tropical Trans-basin Climate Variability using the CESM2-based Multi-year Prediction System

The Tropical Trans-Basin climate variability (TBV) has been demonstrated as an essential source of multi-year predictability for climate variability in many tropical and extratropical regions. This study aims to better estimating the multi-year predictability of tropical trans-basin climate variability and its global teleconnection by using a multi-year Earth System Prediction system based on Community Earth System Model Version 2 (CESM2). The prediction system consists of 100-member historical simulations, 30-member ocean assimilations, and 30-member 5-year hindcasts initiated from every January 1^st from 1960 to 2021. The two TBI indices based on sea level pressure (SLP) and sea surface temperature (SST) are defined as the zonal gradient of SLP and SST anomalies averaged and standardized in each region between the tropical central Pacific (15S-15N, 180W-150W) and the tropical Atlantic-Indian Ocean (15^oS-15^oN, 40^oW-60^oE). Comparison between the mean of 100-member historical simulations and observation indicates that external forcings explain about 10% of the observed nonfiltered TBV variability for the last 62 years. Compared with the historical simulations, ocean assimilations exhibit a more realistic TBV pattern and the associated global teleconnection. The TBV is predictable and reasonably predicted for up to 5 years in the multi-year Earth System Prediction system. However, the model’s mean biases in the tropical oceans degrade the forecast skill at longer leads. We discuss how the model’s intrinsic mode of TBV differs from the observation and how the atmosphere-ocean coupled processes extend the predictability of TBV and its global teleconnection.