slowdown when Hrms at boundary is very small

[Jim Gunson]

Hi,

I am investigating the decay of infragravity waves in XBeach.
I am running XBeach, instat=41, with half-hourly forcing at
the seaward boundary with Jonswap parameters:
1.20 5.0 260.00 3.300 10.000 1800.000 0.300
after 12 hours of model time, I drop the Hrms to a very low value
(1mm);
0.001 5.0 260.00 3.300 10.000 1800.000 0.300

Something I notice is that the CPU time per half-hour model time
quadruples when the boundary wave height drops. Can anyone
explain why this happens ?
(The XBlog.txt file is 7MB, can post it if you like)

Regards,

Jim Gunson