Hydrodynamic and Water Quality Modeling of Lake and Estuary Systems Considering Inflow and Outflow Effects

Abstract

Algal bloom, fish kill, hypoxia, and toxicity are some of the water quality issues faced by many waterbodies, from lakes to estuaries. Internal or external nutrients cause excess growth of algae and then lead to low or no dissolved oxygen (DO) in the waterbody, which impairs the life of living organisms, fish, etc. This has a direct influence on public health, the biological integrity of natural resources, and the economy. This study focuses on understanding the nutrient dynamics and using appropriate water quality model to mimic the nutrient conditions on a range of waterbodies, small lakes to estuaries. An existing lake water quality model MINLAKE2012 was modified to include nitrogen, phosphorus and DO sub-models along with sediment flux calculation to simulate the water quality parameters for long term. Six lakes in Minnesota were simulated using this model and the average standard error for DO simulation of these lakes decreased by 24.2% from the original MINLAKE2012 model, which indicates better model performance. This model is suggested as a reliable tool to managers. A comparative study of 1-D MINLAKE and 3-D EFDC model has been conducted for three lakes in Minnesota. It was found that the EFDC model is suitable where the spatial variance is significant within the waterbody. However, one drawback of EFDC is the absence of a snow model which is important for cold regions. The 3-D EFDC model, being reliable for capturing spatial variance and being more detailed, was used to simulate water quality in Cotton Bayou-Terry Cove system in Mobile, Alabama, which has multiple inflows and very complex bathymetry. This estuary has been suffering from very poor water quality, and fish kill issues which have caused economic and cultural loss. Following the nested model approach of EFDC, the model was calibrated using the continuous observed data for water level, water temperature and DO collected at three locations in Cotton Bayou-Terry Cove system. The standard error for DO simulation ranged from 1.09 mg/L to 1.50 mg/L at the three locations. Different restoration techniques were tested to select the most efficient one by using different boundary conditions in EFDC which includes aerator and dredging. A total of sixteen aerator sets were required to keep the DO above the target DO, 3 mg/L. When both aerator and dredging were used, a dredged channel in Cotton Bayou and nine aerator sets were able to keep the DO above the target DO at all time. Finally, the best restoration techniques were tested against future climate and sea level rise scenarios to predict how the estuary would behave with the restoration techniques implementation in the future. This study gives an overall idea of water quality modeling which will be useful for lake and estuary restoration and help the managers to make informed decisions.