Modelling stratification in shallow estuarine lakes

I am modelling the saltwater stratification in some estuarine lakes using MIKE3FM. The depth of the lakes vary from 3 to 15 m. The observed halocline can occur within 200mm where the salinity increases by 13PSU. I tried to use different turbulence models and vertical resolutions. The modeled and observed surface and bottom salinities matched pretty well , but I  could never get that sharp vertical gradient salinity ( see the attached example). I was wondering if  anyone has successfully modelled such sharp stratification using MIKE3 in this kind of shallow lake environment. Many thanks.
Yafei

Hi Yafei,

do you have an explanation for this phenomenon in natural?
Do you know something of the depth dependent flow directions or flow velocities?
Do you know something about the depth dependent temperatures?

Or do you assume, that there is no flow in your estuarine lake? Is there a barrage or anything else?

Best regards, Christian

Hi Yafei

Have you tried to use a mixed sgima-z vertical mesh? Place start of the z mesh a little bit abvoe the halocline. A pure sigma coordinae model is difficult to calibrate in this case.

Michael

Hi Christian,
Thank you for your reply. Strong stratification occurs in the lakes when there is flood coming from the tributary rivers. The top 5m surface water of the lakes becomes mostly fresh after the flush  while the bottom remain saline. At the mean time, salt is also pumped into the lakes beneath the fresh water through the entrance where it is connected to the ocean. After  a few months the lakes would be well-mixed again. The largest temperature difference between the top and bottom water is less than 2 degrees Celsius. Unfortunately, we did not have the velocity profile in the lakes during the floods but we have the water levels  at different locations calibrated to observed ones. So I assume that the model should well reproduce the hydrodynamics of the lakes. As far as I know, there is no barrage in the lakes. I tried to put a artificial one at the level of halocline and it just stops salt going upstream. We suspect that it might be caused by numerical diffusion. We were  recommended by DHI that the classic version of MIKE3 might do a better job, so I am going to give that a go. Hopefully it can solve our problem. i will keep you posted.
Kind Regards
Yafei



[quote author=Christian Tomsu link=topic=1333.msg3332#msg3332 date=1358931650]
Hi Yafei,
do you have an explanation for this phenomenon in natural?
Do you know something of the depth dependent flow directions or flow velocities?
Do you know something about the depth dependent temperatures?

Or do you assume, that there is no flow in your estuarine lake? Is there a barrage or anything else?

Best regards, Christian

[/quote]

Hi Michael,
Thank you for the recommendation. I did have combined sigma&z vertical mesh with z layer above the halocline, but the model does not seem to work well.
Regards
Yafei
[quote author=Michael Potthoff link=topic=1333.msg3345#msg3345 date=1359536009]
Hi Yafei

Have you tried to use a mixed sgima-z vertical mesh? Place start of the z mesh a little bit abvoe the halocline. A pure sigma coordinae model is difficult to calibrate in this case.

Michael
[/quote]

Hi, I'm aware this is an old topic however it's not been resolved and I have a similar problem (thermocline not maintained over model simulation). Is there anyone who can add anything to the above?