Message Boards

Coastal model Dissolved Oxygen (DO) boundary & computational issue

Danker Kolijn, modified 5 Years ago.

Coastal model Dissolved Oxygen (DO) boundary & computational issue

Youngling Posts: 10 Join Date: 11/4/14 Recent Posts
For my coastal scenario (3 open boundaries, large volume of water with north/south alongshore currents) I experimented with the SWRear settings where 7,9,10 (see various time-series) were recommended to compute a diurnal DO signal using temperature and cloud cover, etc. . However, when I imposed an average value of 6.7 mg/L (as a time series) at the boundary the model did not produce the proper output (dotted red line does not produce amplitude expected) when compared to measured (black line). It seems almost as if the flow of DO from the boundary damps the computational process (I confirmed this in a video of the simulation). When I implemented a -999 DO time-series at the boundaries I found that this is equivalent to introducing 0 mg/L of DO (red dots on plot), where the model seems to be initially inundated with low DO. However, it then starts to the generate the proper signal (in phase and somewhat reasonable amplitude) when compared to my measured data.

I feel like i'm missing something very simple. I don't want to impose a measured DO boundary condition. I want the model to compute DO on it's own given the temperature and climatic conditions I feed the model. Does anyone know why when I input a higher average DO at the boundary, the computational process of generating a diurnal DO concentration seems to be damped?
many thanks in advance
Christophe Thiange, modified 5 Years ago.

RE: Coastal model Dissolved Oxygen (DO) boundary & computational issue

Jedi Knight Posts: 125 Join Date: 11/15/12 Recent Posts
Hi Danker,

It looks indeed like your boundary conditions are dominating the solution here.
What is the average age of water in your domain?
Is the flow direction at the boundaries mainly constant or does it change regularly?
Are you using any timelags with your boundary conditions?

The fact that your runs with zero DO at the boundaries yield a signal with more amplitude is likely due to the fact that the reaeration rate also depends on the difference between actual and saturation DO concentrations.