VISC2 valuse

[kobl1201]

Hi~

when i simulated model, which resolution is about 3km, i set visc2 ~ 1m^2/s.

i heard the rule of thumb approximation like this.

visc2 ~ (2*delta_x / pi)**2 / timescale

According to this formaula, when i set the 3km resolution and time scale is 100s, i should set the VISC2 ~ (2*3000/3.14)**2/100s.

it is very large value, i think.

Is it right?? or i have some wrong??

Thank you~

[kate]

The time scale is damping time, not the time step. You want to damp at about the growth rate of 2 dx noise.

[kobl1201]

Thank you Kate~

so, the time scale is demermined by myself???

[kate]

Yes, feel free to try different values.