Re-Evaluating Established TMDLs: Applying True Adaptive Management to Improve Water Quality in High Value Receiving Waters

Background : Total maximum daily loads (TMDLs) are sometimes developed with limited datasets and simplistic modeling approaches. Nevertheless, these TMDLs can have large cost implications for regulated entities such as wastewater utilities and municipal separate storm sewer systems (MS4s). After TMDLs are officially approved, they can be difficult to change, and can result in costly MS4 requirements over multiple permit terms. If the original TMDL was poorly formulated, these investments might not result in true environmental benefits. This paper presents a case study of how two Pacific northwest localities (Pierce and Whatcom Counties, WA) engaged regulatory agencies and other stakeholders to reassess data-limited TMDLs and improve approaches to protecting water quality in high value waters.

* Stream Example— Clarks Creek: Clarks Creek is tributary to the Puyallup River to that drains about 13 mi2 of Pierce County and the City of Puyallup (Figure 1). High spring flows allow the stream to support one the region's healthiest populations of salmon and trout. The County has historically harvested aquatic plants in the stream to improve aesthetics and reduce flooding of adjacent properties (Figure 2). Clarks Creek was 303(d)-listed as impaired for low dissolved oxygen (DO) concentrations in 2012, and the Washington Department of Ecology (Ecology) finalized a TMDL in 2014. The TMDL called for large increases in DO in stormwater and groundwater, treatment of large stormwater volumes, increases in riparian shading, a 67-75% reduction in aquatic plants, and an 80% reduction in sediment oxygen demand (SOD). The original TMDL technical analysis (Ecology, 2014) was constrained by data and modeling limitations. In 2015, the County and Ecology agreed Ecology agreed that the TMDL study would be subject to a 5-year reassessment process. The County collected additional hydrologic and water quality, and performed special studies of aquatic plants and sediment characteristics. The County also developed a new modeling framework that included a hydraulic model (HEC-RAS), a watershed model (HSPF), and a receiving water quality model (QUAL2Kw) (Figure 3). A series of model scenarios were run to explore the response of DO in Clarks Creek to various management options, including riparian shading, stormwater treatment, stormwater diversion, and aquatic plant reduction (BC, 2022; Gawel, 2020). Some important new findings include: (1) Under dry weather conditions, DO in the creek is primarily determined by the DO of spring water (~8 mg/L) that is naturally lower than the water quality criterion (10 mg/L). (2) SOD is much lower than assumed in the original TMDL analysis, which leaves less potential room for DO improvement by SOD reduction. (3) The original TMDL's call for large amounts of traditional stormwater treatment would not likely benefit the creek, because DO in the creek is insensitive to the parameters that are normally removed in stormwater BMPs. (4) Conversely, practices that increase the DO of stormwater inflows (e.g., improved reaeration) could improve DO in the creek. (5) Average DO in the creek is actually higher if some aquatic plants remain in the creek. Based on the new data and models, the County is proposing major updates to the original TMDL analysis, including the development of a site-specific water quality criterion that reflects the high spring flows to the creek. The proposed implementation approach continues to emphasize riparian shading and SOD reduction. However, it put less emphasis on traditional stormwater treatment and aquatic plant removal, and more emphasis improved reaeration in stream channels and the stormwater collection system.

* Reservoir Example—Lake Whatcom: Lake Whatcom is an important ecological, recreational, and aesthetic resource and the main water supply for more than 100,000 people in Whatcom County (County) and the City of Bellingham (City) in Washington State (Figure 4). Motivated primarily by declining concentrations of dissolved oxygen (DO), the Ecology developed a TMDL for total phosphorus (TP) (Pickett and Hood, 2014). The TMDL modeling process was limited by relatively few hydrologic and TP data available at that time, and included only a two-year simulation period (2002-2003). In 2019, the City began a process for updating the models to reassess pollutant loading and lake response for the TMDL. The watershed and lake water quality modeling framework were significantly upgraded for the TMDL reassessment, including many more years of data (2002-2015). After calibration, the updated Lake Whatcom water quality model was applied to simulate the DO response of the lake to pollutant reductions (BC, 2023) (Figure 5). The revised modeling indicated a DO deficit of only about 20 percent of that predicted by the original TMDL modeling. The percent of land subject to expensive stormwater retrofits would decrease from 87% under the original TMDL to about 40%. Significance: These case studies demonstrate that some established TMDLs can be successfully reassessed and modified. A TMDL reassessment project requires significant investment in new data and modeling. Such projects must also be pursued in close cooperation with stakeholders including regulatory agencies, citizens, and environmental groups. However, the outcome can be lower implementation costs, improvement environmental protection, or both.