Calibration of MIKE21 flow model FM

Hi Members,
I am a new users of MIKE21 flow model. In my model, the simulated data for seawater temperature does not fit to the measured data. So, in order to calibrate the model, which coefficients (wind friction, eddy viscosity, bed resistance, horizontal dispersion) should be taken into account? Which is more important and should be considered first? I really appreciate if you let me know.
All the best,

Hi

It depends if you include a "real" heat balance calculation (i.e. with the influence of solar radiation etc.) or if you just handle temperature as the result of mixing different water bodies. A heat balance calculation may be nesseccary if you have longterm calculations and a large domain. However, calibrating it is quite complex. For smaller domains and short term calculations you can propably calculate termperature based on mixing alone. In this case you do as any other AD/transport model calibration. Depending on you model wind may be important but usually the dispersion /eddy settings determine the results. However, you need to have a proper flow first, i.e.

1) you need to calibrate your HD flow
2) Then calibrate the AD (maybe you can check salinity)
3) calibrate the temperature/heat balance

Thank you Michael Potthoff for responding.
Who has any other thought or suggestion?
Regards,