[A-CG32] Climate Variability and Predictability on Subseasonal to Centennial Timescales

[ACG32-10] Enhanced Skill and Signal-to-Noise in an Eddy-Resolving Decadal Prediction System

★Invited Papers

*Stephen G Yeager¹, Ping Chang², Gokhan Danabasoglu¹, Lixin Wu³, Nan Rosenbloom¹, Qiuying Zhang², Frederic Castruccio¹, Abishek Gopal², Cameron Rencurrel² (1.National Center for Atmospheric Research, 2.Texas A&M University, 3.Ocean University of China)

Keywords:decadal prediction, high-resolution climate modelling, mesoscale air-sea interaction

The sensitivity of decadal prediction system performance to model resolution is examined by comparing results from low- and high-resolution (LR and HR) predictions conducted with the Community Earth System Model (CESM). The primary difference between the two systems is the horizontal grid spacing of the ocean and atmosphere models (1° for both in LR; 0.1° and 0.25°, respectively, in HR), permitting a direct comparison of how skill and signal-to-noise characteristics change when moving to the ocean eddy-resolved modeling regime. HR exhibits significantly increased overall skill at hindcasting pentadal anomalies in the ocean and atmosphere compared to LR, although results vary from region to region. Examination of signal-to-noise characteristics for atmospheric fields reveals substantial improvement in HR. This result lends support to the hypothesis that mesoscale atmosphere-ocean interaction, which is present in HR but absent in LR, is an important mechanism involved in the transmission of predictable signals from the ocean to the atmosphere.