Add mixing option

[bartvanwesten]

Before, after hydrodynamic mixing, the bed composition of the mixed layers was reset to the average of those layers.

On shorter term scale, this approach seems to work, but after longer simulation periods, the approach appears to result in an underestimation of fine particles emerging in the top layer. If armouring occurs for a long period, the average bed composition of all layers starts to become more coarse. After several years, during mixing no small particles were emerging to the top layer anymore.

To solve this, an optional mixing method could be implemented that does not reset the mixed bed to the current average of the layers, but to the original bed composition.

method_mixing = layer_average (default)

OR

method_mixing = reset_initial (new)

[Sierd]

@christavanijzendoorn and @bartvanwesten
I am not sure if we need that additional 'bedcomp_mixing_file'. Isn't the 'bedcomp_file' enough?

[christavanijzendoorn]

I agree with Sierd that there's no extra bedcomp_mixing_file needed. On line 109 of bed.py the bedcomp file is read and immediately converted into the mass parameter. This means the bedcomp file will still contain the initial bed composition and can be used in Bart's code to reset the bed composition. I think that'd save some lines of code and loading two bedcomp files.