In the end, the student's work will be largely numerical, but comparisons with exact solutions, and checks that the solutions conform to provable qualitative behaviour are thoroughly worthwhile.
Part of this PhD work involves oscillations in lakes.
I'm assembling a couple of surveys. Part I is just on the constant coefficient Helmholtz equation.
Part II will treat varying bottom depths (equation (10.13.19) in the old classic Stoker "Water Waves", for example, in the context of surface waves).
I have over the years learnt lot about the Helmholtz equation - and Part I is evolving from adapting some of my earlier papers.