|
Title: Thermistor-chain data assimilation to improve hydrodynamic modelling skill in stratified lakes and reservoirs. Authored by: Yeates, P.S., Imberger, J and Dallimore, C. Abstract:
Simulation results from any 3D hydrodynamic model show that the predicted distribution of the gradient Richardson number, which is commonly used to parameterise diapycnal mixing as is also the case in ELCOM, decreased rapidly due to numerical diffusion, thus increasing the vertical mixing dynamics of the simulation and so introducing a positive feedback that quickly drives the predictions offcourse. A numerical filter technique is proposed that assimilates thermistor chain data and adaptively corrects the computed density field, negating the numerical diffusion, to minimize the difference between the thermistor chain measurements and the computed temperatures. The adaptive gradient sharpening filter ensures that the buoyancy frequency at the position of the thermistor chain remains within bounds of the measured value and so the computation better preserves the internal wave spectrum of the internal wave motions. .
Reference: Yeates, P.S., Imberger, J and Dallimore, C., 2008, Thermistor-chain data assimilation to improve hydrodynamic modelling skill in stratified lakes and reservoirs., 134(8):1123-1135 Keywords: Kinneret |
Publications options:
Browse by year: |
