What about the last 10-15 years? GCM community was starting to pay serious attention to PBL clouds In general GCMs did not have acceptable Sc (e.g. ERA15,UKMO) (a) cloud cover too low (in spite of Slingo87 and prog. clouds) (b) PBL too shallow (due to local Ri-number with dry variables) Cu boundary layers were a serious problem: vertical structure/cloud cover and water. A consistent and well organized effort (e.g. GCSS-WG1) between people from LES, parameterization and observation communities Noticeable progress from LES simulation of Sc, to a better conceptual understanding, to a improved parameterizations of Sc PBL (e.g. UKMO)
2) What are some the main issues right now? Although there has been progress, we are still far away from where we want to be : The problem of too many patched parameterizations and switches Still models (mesoscale/global) that do not have anything close to reasonable Sc and Cu. Major general parameterization issues: microphysics/aerosols and mesocale organization Transition from Sc to Cu: Is this still a theoretical (conceptual) challenge? Or should current parameterizations be able to do this transition well? Stratocumulus: Do we still trust LES for Sc? Is the level of agreement in top-entrainment parameterizations good enough? How can we improve this? Moist conserved variables: should GCM dynamics be written in conserved variables? What about data-assimilation? Do we focus too much in the subtropics? Is this a problem? There is more to PBL clouds than sub-tropical PBL clouds: (i) Shallow convection over land, (ii) wintertime stratus, (iii) Arctic stratus, (iv) radiation and advection fog. Social-cultural issues: Funding time-scale (3 years) – bad for parameterization development The typical time-scale for operational implementations is too long
3) What about the future? The successful collaboration between OBS/LES/SCM/GCM (a la GCSS-WG1) should definitely continue The major undertaking of trying to parameterize microphysics/aerosols needs this important collaboration The community should try to explore the enormous wealth of data-sets in a more efficient manner Cloud parameterization: pdf-based cloud schemes provide a promising framework for a general and successful cloud parameterization Vertical sub-grid mixing: ideally unified approaches are desirable. But regimes-based approaches provide a particularly successful short-term (?) strategy. Is nature organized in distinct regimes as well? Mesoscale organization: provides a major challenge to our typical parameterization framework. Does it imply the need for stochastic approaches? Or major theoretical shifts on the way we do parameterization? A more consistent and long-term perspective (funding and otherwise) is necessary from the major agencies/institutions. Mostly US or also Europe (EU funding)?