We’re working on a long-term ICON-Med AMIP simulation with three two-way nested atmospheric domains: global ~40 km, Mediterranean ~20 km, and an inner Mediterranean domain at ~10 km.
I’d appreciate your advice on the most appropriate treatment of convection at ~10 km, particularly for long-term climate simulations.
Would you recommend retaining deep-convection parameterization, using shallow-convection-only, or adopting a scale-aware approach?
I’m also interested in whether different convection settings across two-way nested domains could introduce consistency or feedback issues.
Are there relevant experiments or sensitivity studies that could help guide this choice?
Many thanks