METHODS FOR IMPROVING NHD STREAM NETWORKS WITH LIDAR - A FRAMEWORK FOR GLACIATED, GROUNDWATER DOMINATED LANDSCAPES IN THE HURON-MANISTEE NATIONAL FORESTS, MICHIGAN

The National Hydrography Dataset (NHD) is the authoritative source of hydrography data in the United States. Many federal, state, and local entities rely upon it to inform land management decisions, monitoring efforts, and research programs. Having accurate stream network data is crucial for various decisions made within national forests, including planning timber harvest mitigation efforts that protect riparian environments, and prioritizing road-stream crossing projects to improve aquatic organism passage. Recently acquired light detection and ranging (LiDAR) data for Huron-Manistee National Forests provide a more detailed look at the landscape and highlight dramatic inaccuracies in the current NHD flowlines for the area, thus providing an opportunity to update a vital national dataset. This work presents a method for using 2-foot resolution LiDAR Bare Earth digital elevation models (DEMs) to model stream networks for 6th level watersheds in the Huron-Manistee National Forests, followed by field-based data collection to verify stream sources, sinks, and paths using ArcGIS Collector and ArcPro. Key aspects of this process are highlighted including the use of additional remote sensing data to determine appropriate flow accumulation thresholds when modeling flowlines and to identify likely stream source locations. Additionally, challenges associated with generating modeled streamlines in a glaciated, topographically deranged, largely groundwater dependent landscape are discussed. As LiDAR and other high-resolution remotely sensed data becomes increasingly available nationwide, research aimed at refining methods to model, map, and refine NHD features will be increasingly important, especially at regional or local scales.