experiments_rep.txt

Convective organisation/mode of convection (0.0 ... 0.5 .. 0.8 .. 1.5)
It seems like modification of the momentum advection term mostly influences our results via the mode of convection that appears in a simulation. Low values for "control_vadv" (vertical advection of u and v) lead to poorly organised convection instead of a supercell modus, which leads to a cut-off of convective plumes, due to the downdraft motion. This is seen in the MSE cross-sections, but needs to be compared to the effect of downdrafts and subsequent cold pools in a x-y plot.
When vertical advection of horizontal is amplified, smaller cell sizes lead to well organised convection, but smaller cell sizes. 
In the condensational heat - divergence space, we see a tight relation between the two quantities for high resolution reference (ensemble) runs and runs with modified lve, which is nearly linear as expected. There is a little uncertainty. Modifying the advection of horizontal momentum in the vertical leads to a little more variability, which may be related to cell size and degree of organisation and will be related to mixing/entrainment processes.

Modification of advection of moist static energy (0.8 and 1.2)
Unconserved equations with modification of the moist static energy advection appear to give rise to sources and sinks of heat in the stratosphere after the simulation has started (possibly due to subtle gravity waves triggered by the convection and flow disturbance below? Or an effect of the boundary conditions at the top?). Subsequently, a turbulence source propagates downward from the stratosphere to the surface during the two hour simulation. 
For values of 0.5 and 1.5 teh simulations "explode" after ten or twenty minutes.
A solution could be to couple vertical advections of water vapor and potential temperature to conserve moist static energy. It might solve problems. 


_________________________________________________

About/after squall line case
Is it the area that is precipitation from the anvil and exposed to dry midlevels at a large "surface" of the cloud which ventilates condensed air to evaporation and subsequently allows for net no divergence? Mixing is at least less concentrated at the tropopause interface then with the supercell and more or less constrained to this region. Additionally, much more MSE is ventilated by the squall line and over a much wider region. There is also more source air lifted in the squall line case with a higher turnover.

We should think per (kilo)metre squall line compared to the supercell in this context
Additionally the interplay between dynamics and instability which distinguishes the cell types and particularly the supercell...

Maybe also think along the pdV = W reasoning if partly consistent?