“…The idea that simpler alternatives might exist is indeed justified by the fact that some recent APE studies successfully moved away from the use of sorting algorithms by resorting to approaches using probability density functions instead, as in the case of Saenz et al (), itself an extension of Tseng and Ferziger (), although it is unclear how such a method could be applied to a moist atmosphere. Also, it has long been known from the works of Andrews () and Holliday and Mcintyre (), which were recently generalized for multicomponent compressible stratified fluids by Tailleux (), that it is possible to construct a local theory of APE based on an arbitrary reference state defined by a reference pressure p 0 ( z , t ) and specific volume α 0 ( z , t ) in hydrostatic equilibrium. Based on Tailleux () and Novak and Tailleux (), this would lead one to define the APE density for a moist atmosphere as the work that a fluid parcel needs to perform to move from its reference pressure p r to its actual pressure p , namely, where θ l is liquid potential temperature and q T is total water content.…”