Jean-Michel and Patrick, thank you for info re heat fluxes. For the existing cubed-sphere integrations we have no explicit horizontal diffusivity, only that which is implicit in the "tempAdvScheme=33", so DIFy_TH would be zero, no? Dimitris