Table Of Contents

This Page

02.07.2014: Behavior of Z P system with chaotic advection and zero diffusion

This case (run_002/equilibrium_6_diffk_0) is similar to case #1 in this note except that there is now a chaotic advection of the field (but still zero diffusion). As in the V-P system case, the edges of the bloom get excited as if there was diffusion. As I wrote in the V-P case, the diffusion is still present even when we increase the temporal and spatial resolution and, at this point, I still do not understand it.

../../../../../../_images/equilibrium_6_diffk_0_snapshots.png

Figure 2: Snapshots in case run_002/equilibrium_6_diffk_0. The velocity field is shown in the lowest panels.

Next things to do

  1. Check, using the time series of the spatial averaged quantities, that the diffusion is really not sensitive to IORD (maybe, it is a bit sensitive and we need to use a lot more iterations to reduce the diffusion)
  2. Using again the time series of the spatial averaged quantities, check that the impact of numerical diffusion is also significant when the advection flow is stronger (maybe, the numerical diffusion is not too important when the advection is strong; I doubt it though).

Notes after meeting with Kelvin (02.07.2014)

Run with enough spatial resolution to minimize the numerical diffusion. Note that this might depend on the strength and type of flow. Compare to the case with no advection (and no explicit diffusion) using spatially averaged quantities.


Snapshots are plotted with plot_snapshot_current_run.py in RESEARCH/MODELISATION/marine_viruses/current_version/runs/run_002/analysis/ on ipu1.