Abstract

The plans for the ACME model are to ultimately develop a credible, fully coupled system fit for purpose. We show a set of experiments that illustrate the potential impact we may expect to see in the fully coupled system in response to a number of near surface change sensitivities in the atmosphere model. 

The introduction of turbulent mountain stress, a missing stress process keying off sub-grid orographic scales between small and resolved scales, demonstrated many improvements in north-Atlantic and southern ocean surface stress biases in CAM5. However, the sensitivity to the free parameters in the system was never systematically examined. With such a significant potential to influence the coupled system we examine sensitivities in v0.3 AMIP simulations and identify the key parameters responsible for regulating surface stresses over the ocean (Hannay et al, 2015). With these most impactful parameters we further investigate the sensitivity in the fully coupled system using the v0.3 coupled system.

Convective gustiness was previously shown to have a beneficial impact on surface latent heat fluxes in equatorial regions, with improvements in the Asian monsoon precipitation ‘hole’ currently seen (Neale et al, 2015). Because of the dependence of the scheme on surface convective precipitation, and the different patterns seen in coupled experiments compared to AMIP, coupling could prove to moderate the simulation markedly and we investigate this potential.

Finally, with the plan to increase vertical resolution from 30 to a possible 72 layers in the vertical (Ma et al, 2015), the potential impacts on the coupled simulations are obviously significant. A simple systematic increase in vertical resolution has previously been shown to exacerbate the existing monsoon ‘hole’: a problem that intensifies even at vertical levels numbers as high as 120. Perhaps more importantly the thinning of the atmosphere level nearest the earth’s surface has a greater potential to impact the coupled simulation. This could be both due to the decreased, and more responsive, mass of the layer and the interaction with aspects of the atmosphere parameterizations that are dependent on just the existence of the lowest model layer and not its thickness: this is certainly the case for turbulent mountain stress.

In order to illustrate these potentially important surface coupling sensitivities we will show results from a range of AMIP-type simulations and illustrate their impact more directly in a limited set of coupled integrations.

References: 

Neale et al. (2015), Deep Convection Modifications for Gustiness, Entrainment and Timescale. ACME All-Hand Meeting
(https://acme-climate.atlassian.net/wiki/download/attachments/23200211/Neale_ACME_poster_May2015.pdf)

Ma et al. (2015). Vertical Resolution, ACME All-Hand Meeting
(https://acme-climate.atlassian.net/wiki/download/attachments/23200211/ACME_Meeting_c150505_Vertical_Resolution.pdf)

 Hannay and Neale (2015) Sensitivity to Turbulent Mountain Stress
(https://acme-climate.atlassian.net/wiki/display/ATM/Sensitivity+to+Turbulent+Mountain+Stress)