Scottish Shelf Model. Part 6: Wider Domain and Sub-Domains Integration

Recommendations for future work

There are some outstanding issues which could be discussed further but these are outside the scope of the present work:

6. How do wild salmon influence dispersion of sea lice?
7. What about the wall effect mentioned in several papers about particle dispersion e.g. Amundrud and Murray (2007; 2009)? Some rules may be applied when the particle comes close to land. There is no rule applied in the FVCOM offline particle tracking code.
8. The effect of winds will not be as strong as likely under certain wind events, as we are only including monthly mean wind stress in model climatology.

There are potential improvements to the particle tracking code which could be made relatively easily

1. Implement the option for lat/lon model output (Dmitry Aleynik has already produced code with this fix, which could be adopted).
2. Jim Clark (Plymouth Marine Laboratory) has provided a modified code in September 2015 - adding a local search for velocity elements to yield a substantial speed up in run times.
3. Another suggestion is to speed up the I/O by reading a whole daily file at once rather than hour by hour (this will have memory implications however).
4. Remove the hard-coded year number of offlag.f90.
5. Allow particle tracking to go beyond the year end and cross boundaries between April/May and October/November (latter is due to only 2 nested models being run for whole year so combined grid changes size).