1st International Congress on Environmental Modelling and Software - Lugano, Switzerland - June 2002
Keywords
reservoirs, water quality, hydrodynamics, modelling, spatial and temporal variability, management tool
Start Date
1-7-2002 12:00 AM
Abstract
A one-dimensional (1D-DYRESM) and three-dimensional (3D-ELCOM) hydrodynamic model were coupled to a common ecological model (CAEDYM) and applied to 2 different, but inter-connected reservoirs. A 1D water quality (WQ) simulation (DYRESM-CAEDYM) of large (V=2 km3, A=82 km2, L=50 km) and deep (zmax=90 m) Lake Burragorang during a drought (1992-1995) compared well with field data. DYRESM-CAEDYM simulations of much smaller (A=5 km2), shallower (zavg=9 m) and low residence time (ca. 1 month) Prospect Reservoir over 8 years (1983-1991) were validated against a comprehensive WQ record, with no modifications to the Lake Burragorang application other than to daily forcing and bathymetry files. Lake Burragorang is subject to occasional flood events involving rapid temporal evolution of spatial variations that cannot be simulated by a 1D model. A winter flood in Jun. 1997, with comprehensive spatial monitoring, took ca. 1 week to travel to from tributary to dam wall (~ 50 km) as a nutrient-laden underflow. Grids (100-200m×100-200m×1-2m) needed for 3D model run times to follow the evolution of the flood event were too large to resolve the narrow and complex geometry of this reservoir, but by ‘straightening’ the domain, larger grid sizes with suitable run times yielded good validation results. A 3D simulation of Prospect Reservoir during the onset of seasonal stratification indicated poor WQ from the 2 inflows is ‘contained’ to below the metalimnion, where its accessibility to algae is restricted by the stratification. A simulation with a bubble plume destratification system indicated that these inflows would be inserted into the mixed layer and available to phytoplankton. This study demonstrates that physical events (e.g. floods, destratification) often dictate the dominant responses of the biogeochemistry that produce an observed pattern of WQ. This suggests accurate validation and prediction of physical processes is the basis of accurate forecasting of natural or anthropogenic influences on reservoir WQ.
Application of 1D and 3D Hydrodynamic Models Coupled to an Ecological Model to Two Water Supply Reservoirs
A one-dimensional (1D-DYRESM) and three-dimensional (3D-ELCOM) hydrodynamic model were coupled to a common ecological model (CAEDYM) and applied to 2 different, but inter-connected reservoirs. A 1D water quality (WQ) simulation (DYRESM-CAEDYM) of large (V=2 km3, A=82 km2, L=50 km) and deep (zmax=90 m) Lake Burragorang during a drought (1992-1995) compared well with field data. DYRESM-CAEDYM simulations of much smaller (A=5 km2), shallower (zavg=9 m) and low residence time (ca. 1 month) Prospect Reservoir over 8 years (1983-1991) were validated against a comprehensive WQ record, with no modifications to the Lake Burragorang application other than to daily forcing and bathymetry files. Lake Burragorang is subject to occasional flood events involving rapid temporal evolution of spatial variations that cannot be simulated by a 1D model. A winter flood in Jun. 1997, with comprehensive spatial monitoring, took ca. 1 week to travel to from tributary to dam wall (~ 50 km) as a nutrient-laden underflow. Grids (100-200m×100-200m×1-2m) needed for 3D model run times to follow the evolution of the flood event were too large to resolve the narrow and complex geometry of this reservoir, but by ‘straightening’ the domain, larger grid sizes with suitable run times yielded good validation results. A 3D simulation of Prospect Reservoir during the onset of seasonal stratification indicated poor WQ from the 2 inflows is ‘contained’ to below the metalimnion, where its accessibility to algae is restricted by the stratification. A simulation with a bubble plume destratification system indicated that these inflows would be inserted into the mixed layer and available to phytoplankton. This study demonstrates that physical events (e.g. floods, destratification) often dictate the dominant responses of the biogeochemistry that produce an observed pattern of WQ. This suggests accurate validation and prediction of physical processes is the basis of accurate forecasting of natural or anthropogenic influences on reservoir WQ.
