FourierFlows / GeophysicalFlows.jl

Geophysical fluid dynamics pseudospectral solvers with Julia and FourierFlows.jl.
https://fourierflows.github.io/GeophysicalFlowsDocumentation/stable/
MIT License
155 stars 32 forks source link

Systematic error in budgets? #109

Open navidcy opened 4 years ago

navidcy commented 4 years ago

I ran the 2D Navier-Stokes example that computes the energy and enstrophy budgets with higher amplitude excitation and for longer so it became nonlinear. Both residuals for energy and enstrophy budgets seems to have a systematic negative bias, although still the residuals are ~3 orders magnitude smaller that the each of the terms in the budgets.

Here's a plot of the budgets when I let the energy/enstrophy example run for up to μt=2.

budgets

I'm pretty sure that the terms are correct. Perhaps testing that the budgets close when in the nonlinear regime with deterministic forcing is a good idea. There are less ambiguities there...

@BrodiePearson @glwagner

navidcy commented 4 years ago

Btw, as a side-note, we should add the expression of each term in the budgets in the doc strings so that people know what the functions are supposed to be computing. :)

navidcy commented 4 years ago

or if you a bit run longer... budgets

navidcy commented 4 years ago

@cesar-rocha, I remember you computing energy/enstrophy budgets. Did you notice anything like the above? As a remark, these are simulations of 2D Navier-Stokes with linear drag and hyperviscosity. There is neither high-wavenumber filtering or dealiasing.