Artificial oscillations in MT3DMS

[SSaalfield]

Hi all -

I am working on an MT3DMS model of a strongly advection-controlled plume (~4 miles long, 600 ft wide), in GMS Version 8.1.4. As expected for such an advection-dominated system, the TVD/ULTIMATE solution scheme for advection seems to work best.

However, I am getting what appears to be artificial oscillation, resulting in zones of negative concentrations along the edges of the plume (I attempted to attach an illustration). I have tried altering the advection solution options, but have not found a way to address these oscillations (without encountering the substantial numerical dispersion associated with the other solvers). There doesn’t seem to be much ability to tweak the TVD/ULTIMATE solver in GMS.

Have any of you dealt with a similar problem? Are there parameters in GMS that can be changed to improve the quality of such a solution?

Any advice would be greatly appreciated.

thanks,

Samantha

Plume.pdf