Keywords
hydrodynamic, plant, ecological, model, estuary, marsh, neural network, decision support
Start Date
1-7-2006 12:00 AM
Abstract
The Savannah Harbor is one of the busiest ports on the East Coast of the USA and is located just downstream of the Savannah National Wildlife Refuge (SNWR), which is one of the nation’s largest freshwater tidal marshes. The Lower Savannah River estuary has been studied for years by governmental agencies, water users, universities, and consultants having an interest in controlling water quality and predicting the potential impacts of a proposed harbor deepening. Consequently, many different databases have been created that describe the natural system’s complexity and behaviors. Variables having particular relevance include those describing bathymetry, meteorology, water level, and specific conductance. To evaluate the environmental impacts of the deepening, a three-dimensional hydrodynamic model (3DM) and a “marsh succession model” (MSM) were developed by different scientific teams. The 3DM predicts changes in riverine water levels and salinity in the system in response to potential harbor geometry changes. The MSM predicts plant distribution in the tidal marshes in response to changes in the water-level and salinity conditions in the marsh. To link the riverine predictions of the 3DM to the MSM, a “model to marsh” (M2M) model was developed using data mining techniques that included artificial neural networks (ANN). The ANNs simulated riverine and marsh water levels and salinity in the vicinity of the SNWR for the full range of 11½ years of data from riverine and marsh gaging networks. The 3DM, MSM, and M2M were integrated in a decision support system to make it easy for various regulatory and scientific stakeholders to use.
Integrating 3D Hydrodynamic Transport and Ecological Plant Models of the Savannah River Estuary Using Artificial Neural Network Models
The Savannah Harbor is one of the busiest ports on the East Coast of the USA and is located just downstream of the Savannah National Wildlife Refuge (SNWR), which is one of the nation’s largest freshwater tidal marshes. The Lower Savannah River estuary has been studied for years by governmental agencies, water users, universities, and consultants having an interest in controlling water quality and predicting the potential impacts of a proposed harbor deepening. Consequently, many different databases have been created that describe the natural system’s complexity and behaviors. Variables having particular relevance include those describing bathymetry, meteorology, water level, and specific conductance. To evaluate the environmental impacts of the deepening, a three-dimensional hydrodynamic model (3DM) and a “marsh succession model” (MSM) were developed by different scientific teams. The 3DM predicts changes in riverine water levels and salinity in the system in response to potential harbor geometry changes. The MSM predicts plant distribution in the tidal marshes in response to changes in the water-level and salinity conditions in the marsh. To link the riverine predictions of the 3DM to the MSM, a “model to marsh” (M2M) model was developed using data mining techniques that included artificial neural networks (ANN). The ANNs simulated riverine and marsh water levels and salinity in the vicinity of the SNWR for the full range of 11½ years of data from riverine and marsh gaging networks. The 3DM, MSM, and M2M were integrated in a decision support system to make it easy for various regulatory and scientific stakeholders to use.
