Comparison of Empirical and Analytical Physical Assessment Approaches for Stream Restoration: A Case Study on Abrams Creek, Great Smoky Mountains National Park, Tennessee

A physical assessment approach referred to as natural channel design (NCD) is commonly used today by stream restoration practitioners, which requires an empirical-based comparison between study and reference reaches. Use of available analytical tools, or models, into pre-design physical assessments is not widely applied, an approach that does not require a reference condition. As a case study, a comparison of empirical and analytical approaches was conducted on Abrams Creek, located in the Great Smoky Mountains National Park (GRSM), Tennessee. Historically, the Abrams Creek valley was used for subsistence agriculture beginning in the 19^th century, and cattle grazing more recently between the 1930's and 1960's. Several tributaries and a few mainstem sections were channelized for agricultural purposes. GRSM resource managers requested that a channel stability assessment be conducted to evaluate whether restoration was needed because a consulting group was promoting a project on Abrams Creek in order to obtain mitigation credits. This study provided an opportunity to compare geomorphic data input/outputs used in NCD and analytical approaches. The NCD approach utilized stream classification and various geomorphic channel attributes at "bankfull" for a departure-type analysis. The analytical approach utilized HEC-RAS and CONCEPTS models. Supporting these approaches, assessment of channel condition also included: aerial photo interpretation, rapid geomorphic assessments for stability indices, and longitudinal profile knickpoint analysis. Some of the findings included: 1) bankfull flow was greatly overestimated by the empirical approach, whereas HEC-RAS with a hydrological analysis provided a more reasonable estimate, 2) a departure analysis found that C4 study and reference reaches were similar for most geomorphic attributes, except for channel slope and bank material, which made restoration needs difficult to discern, and 3) CONCEPTS provided useful information on bed aggradation/degradation and bank failure rates, whereas the NCD approach could not. hi addition, similarities and differences of data input needs, and where professional judgment influences assessment outcomes were identified to better understand uncertainties associated with outcomes for both assessment approaches.