results_plan.txt

1. present 3 meshes - mention overlapping, non overlapping
2. w poitwise spatial error for a couple of different time slices, mention error at the boundaries: spatial boundary peopbably peaks due to the edge function peaks.
3. MSEM truly mimetic: preserves conservation of linear momentum: present L2, Linf w error vs N for 3 K. Refer again to the balance eqn and point out this relation is preserved exactly indeed. Discuss results: 1. Linf bigger than L2 due to bad approximation at the boundary (recommenation for the future? flatten them at the boundary?), 2. Both error metrics increase at the boundary due to increasing condition number (and edge boundary peaks?). present kappa vs N graph. Also, present residual vs N (TODO). 3. Further increase in K does not change anything: The problem is fully resolved in the T dimension and further increase in K does not influence the solution. Now, only ME in spatial can help to refine res's.
4. (TODO) TOtal energy in time. For high  order two graphs. K = 1. K = 5. Show the boundaries between elements and discuss the peaks, drops. Again, probably increasing K beyond certain threshold will have no influence.
5. a (slope) vs c graph. Discuss the slope dicreases asymptotically until the mesh starts overlapping itself (c around .3); after that: divergence in errror for increasing N (Limitation!). May check how that looks for other domain transformtaions (recommendation).
6. Mention sth about the order of integration?