one-way nesting simulation (using ww3_shel) shows much better results than two-way nesting (using ww3_multi)

[yuehuan]

Hello all, I have conducted global-regional northwestern Pacific nesting simulations using both one-way nesting (ww3_shel) and two-way nesting (ww3_multi). The only difference between these simulations is the nesting method employed. Upon comparing the simulated results with altimeter data, I found that the one-way nesting simulation (the first figure below) performs much better than the two-way nesting simulation (the second figure below). Does anyone have insights into the potential physical reasons for this? Furthermore, I have encountered some confusion regarding nesting:
(1) In the one-way nesting simulation, using more frequent (hourly) boundary conditions did not yield better results compared to less frequent (daily) boundary conditions; in fact, the former performed worse.
(2) Is there a parameter within the ww3_multi.inp file that controls the data transfer between lower and higher rank grids for two-way nesting?
(3) The manual for two-way nesting mentions three types of data transfer between grids How can I change the type of data transfer specified in the inp file to "Transfer of data from lower to higher rank grids in the WWIII wave model"?
Thank you very much for any feedback or comments.