Development of Infrastructure for Testing Candidate Convection Schemes

Abstract

Convective parameterization continues to be a critical physical component in the ACME model. Candidate convection schemes need to be rigorously tested to inform the adoption of an optimal scheme for the ACME’s next-stage production simulations. While the implementations of candidate schemes are underway, the infrastructure for testing candidate convection schemes has been established. It will use a suite of modeling frameworks, which are based on evolving ACME model but with different emphasis, to assess the performance of the schemes in capturing observed multi-scale spatiotemporal variability of key physical quantities. The key testing platform is the Cloud Associated Parameterization Testbed (CAPT) that can be effectively used to evaluate the model physics operating under realistic atmospheric conditions with full dynamical feedback. The workflow of standard CAPT has been streamlined for ACME operation, including automation, efficient execution in super-computing environment, and utilities for enhanced post-processing and diagnostics. A list of critical metrics and diagnostics has been identified specifically for the evaluation of convection, to complement the Tier 1a and Tier 1b metrics developed from other task teams. To optimally utilize well-established field campaign cases and to gauge the model performance at target ACME model resolution without incurring heavy computational burden, the convection testing also heavily involves simulations using single column model (SCM) and regionally-refined model (RRM) configurations. Finally, AMIP-type simulations will be routinely used to assess the ability in reproducing the broader spectrum of variability in space and time that are important to the global climate.