Model Formulation of a GCEP Transient 1860-2000 All-Forcings Run

At the third GCEP meeting (19 Feb 2007) we discussed starting a set of ensembles from an all-forcings run. There already exists such a run on the cluster (xaaqc on gorgon). However, I am concerned that there was no discussion beforehand to determine exactly which mods and forcings to use. Doug Smith has kindly given us a basis file from one of the Hadley Centre's all-forcings runs, it is job xaaon on the gorgon UMUI. It differs on many points to the existing GCEP all-forcings run.

I think we all agree that we do not want any flux corrections in our transient run (and indeed neither of the all-forcings runs mentioned above have them). There is still a question on whether to use the indirect aerosol effect. The Hadley Centre run does not switch on the model option, but rather includes the effect through ancillary files. However, when the model does not have the indirect effect enabled, it uses a two-mode distribution for cloud condensation nuclei (CCN). One value for the ocean and one value for the land. When the indirect effect is enabled the model has a much more realistic and interactive distribution of CCN. It could be argued that by switching on the indirect effect in the model we would obtain a larger and more realistic spread in our ensembles.

I would propose the following for an all-forcings run:

• Use all the same mods the Hadley Centre have which are relevant for vn4.5
• Enable the indirect effect and use mods related to it
• Use the same ancillary files and forcings as the Hadley Centre

-- LeonHermanson - 21 Feb 2007

UPDATE: The final form of the models used in the ensemble was that of the ALL ensemble in Stott et al (Science 290, p2133, 2000). The indirect effect is disabled, though included in a non-interactive manner through an ancillary file. The mod SO4_indirect is used for this. There are no flux corrections used in the ocean-atmosphere coupling. The ancillary files used for ozone, aerosols, etc. are the same as at the Met Office.