All plots are for July 1850 (7 months into run)

Based on reduce the absolute difference in the dynamical core fields, I created the following run (u-bn750).

The biggest mismatch between Snr and Jnr is for fluxes_r_surf_cnv.

Still overwriting the cloud moisture fields

• Stash 392: CLD LIQ MIXING RATIO (mcl) AFTER TS
• Stash 393: CLD ICE MIXING RATIO (mcf) AFTER TS

but not overwriting

• Stash 391: VAPOUR MIXING RATIO (mv) AFTER TS
• Stash 394: RAIN MIXING RATIO (mr) AFTER TS

This is u-bo405.

This greatly improves

• the mismatch for fluxes_r_surf_cnv, and
• the err_in_energy_budget

This is u-bo420.

• The magnitude for err_in_energy_budget is now less for Jnr than Snr.

However, there are still large discrepancies for SW and LW fluxes between Snr and Jnr.