THE EFFECT OF DATA AVAILABILITY AND MODEL COMPLEXITY ON BUXAHATCHEE CREEK TMDL

This project evaluated the effect of data quality and quantity on the total maximum daily load (TMDL) for a moderately complex stream system. A moderate data collection effort on Buxahatchee Creek was performed to gather the information to support the study. Increased levels of data availability resulted in an increasingly stringent TMDL. One investigation calculated a final TMDL of 2 mg/L with a range of 1–10.5 mg/L under different levels of data availability. A second study using somewhat different modeling assumptions calculated a TMDL of 1.8 with a range of 1–52 mg/L. Using actual stream velocity data rather than an estimation resulted in the largest improvement in model accuracy and the largest reduction in TMDL uncertainty. This information also had the effect of greatly improving model prediction error when paired with high quality rate and/or load data. Rate information and multiple sets of field data were also important throughout the creek in reducing model error.Model complexity did not have a large effect on the overall TMDL due to the fact that the simpler model selected was most appropriate for a stream impaired mainly due to carbonaceous biochemical oxygen demand (CBOD). The overall study was enhanced by the information provided by the more complex algae model.The final objective of this study was to prepare summer and winter TMDLs for the Alabama Department of Environmental Management (ADEM). There is an implicit margin of safety incorporated into these TMDLs. The summer TMDL at the Calera wastewater treatment plant (WWTP) is 2.0 mg/L CBODu at a discharge of 0.75 MGD (12.5 lb/day). At 1.5 MGD, the TMDL concentration is 1.4 mg/L CBODu (17.5 lb/da). The ammonia level at both of these discharges is 1.0 mg/L. The winter TMDL at the Calera WWTP is 9.2 mg/L at 0.75 MGD (57.5 lb/da) and 7.5 mg/L CBODu at 1.5 MGD (93.8 lb/da). The overall uncertainty in these TMDLs is considered to be low as the most sensitive parameters in the model are field-verified. In addition, the model was able to simulate a second data set well. Winter TMDL uncertainty is slightly higher. This uncertainty is dealt with in a conservative manner protective of stream quality. Nitrogen was determined to be the limiting nutrient in the stream. Any efforts to reduce algal blooms should focus on nitrogen reduction.