Prediction skill at subseasonal-to-seasonal (S2S) time scale is generally low, particularly for precipitation. Unified modeling systems spanning from weather to climate are believed to have the greatest potential for fully realizing potential predictability at S2S scales. However, systematic model biases due to inadequate representations of key physical processes (such as convection, aerosol-cloud interactions, boundary layer dynamics, and atmosphere, ocean, sea ice and land coupling) likely limit realizing potential predictability from sources of climate variability such as ENSO, MJO, QBO, SSW, diabatic heating over ocean, and land-surface and Arctic sea ice feedbacks.

We welcome contributions that advance the diagnostics and metrics of Earth System models for improving S2S prediction. Studies applying observations to evaluate model performance and guide model development at the process level are particularly encouraged. We also seek showcases of community software that illustrate model deficiencies in key processes that have a significant impact on S2S prediction skill.

• AMS: American Meteorological Society -
• EGU: European Geosciences Union -
• WCRP: World Climate Research Programme -