You are here:
Development and Applications of Rapid Streamflow Duration Assessment Methods
Citation:
Topping, B., Tracie-Lynn Nadeau, K. Fritz, AND R. FertikEdgerton. Development and Applications of Rapid Streamflow Duration Assessment Methods. 13th National Water Quality Monitoring Conference, Virginia Beach, VA, April 24 - 28, 2023.
Impact/Purpose:
Flow duration classification is used to implement several federal, state and local stream management programs. Because the flow duration of streams via existing maps, remote sensing, and gauging is constrained, field-based tools are often needed by practitioners. The presentation describes regional studies and resulting published beta methods to rapidly classify stream reaches as ephemeral, intermittent, and perennial in the Arid West, Western Mountains, Great Plains, Northeastern, and Southeastern U.S.
Description:
Resource managers, regulators, and regulated entities need rapid, reach-scale methods to classify streamflow duration to implement and comply with many federal, state and local programs. The U.S. Environmental Protection Agency is working cooperatively with the U.S. Army Corps of Engineers and other partners to develop regional streamflow duration assessment methods (SDAMs) for nationwide coverage that use hydrological, geomorphological, and/or biological indicators, observable in a single visit to classify reaches as perennial, intermittent or ephemeral (https://www.epa.gov/streamflow-duration-assessment). Since hydrologic monitoring of non-perennial streams is limited, we deployed and maintained our own network of data loggers for up to two years to confirm streamflow duration class in study reaches throughout the country. This project exemplifies how the use of random forest models and machine learning applied to regional datasets can lead to the development of accurate and efficient methods across large regions. We present published beta SDAMs covering the Arid West, Western Mountains, Great Plains and Northeastern and Southeastern U.S. The beta methods resulted from multi-year studies conducted at reaches across the country and included the evaluation of 141 candidate indicators expected to control or respond to streamflow duration at reaches of known streamflow duration class. Through recursive feature elimination, we identified the subset of indicators with the highest diagnostic accuracy of streamflow duration class within each region. Each beta SDAM developed using this approach has a unique set of indicators and will be re-evaluated in light of public comments and additional data during development of final methods in 2023. Additional analysis of the nationwide data will include data resulting in the final SDAM for the Pacific Northwest. The presentation will describe the process used to develop beta and final SDAMs including data analysis steps, regional beta SDAMs developed to date, future steps to complete nationwide coverage of final methods, and opportunities for more refined regionalization. Additionally, the presentation will discuss how the collection, management, analysis, and publication of these data sets has been organized and carried out to support application across multiple management decisions and support further study by others.
