Based on the findings of Colson (2006), a plan for accurately mapping headwater streams was developed. The basic plan elements are:
- Use LiDAR data acquired from the NC Floodplain Mapping Program to create new topographic maps at a resolution of 20 feet.
- Conduct field identification and mapping of 1st and 2nd order intermittent and perennial headwater streams and their origins in selected watersheds.
- Build a geodatabase of study watersheds to include field data, DEMs and DEM-derived data and other GIS data, e.g., slope, contributing drainage area, curvature, soils, precipitation and Ecoregion.
- Conduct spatial analysis using GIS data and field data.
- Use field data and other GIS data to create predictive models of streams and origins.
Accurate mapping of headwater streams must address stream representation as part of the larger concept of a stream network.
The basic questions asked when referencing streams on current maps are does the stream truly exist on the ground and, is there a stream on the ground that is not depicted on the map?
Subsequently, if a stream is depicted on the map, is the length of the stream correct, and is that stream length correctly labeled as ephemeral, intermittent, or perennial?
Each question refers to a specific element of a mapped stream within the stream network.
Model performance with respect to these questions can be evaluated by performing a stream network accuracyassessment that specifically addresses different components of the stream network.
The term “stream network accuracy” encompasses topography, geometry and hydrology, all of which are components of stream networks that influence geomorphic, hydrologic and biologic processes.
Stream network accuracy considers multiple scales and is most important for DWQ, USACOE and EPA applications. Therefore, the objectives of this study are designed to capture several aspects of stream network accuracy of the predictive model.
The objectives, in order of priority, are to develop a spatially-based model that most accurately predicts:
- Presence or absence of a stream in the correct valley,
- Stream length, and
- Flow Duration – i.e., ephemeral, intermittent or perennial.
Development of predictive models requires the evaluation of landscape characteristics and processes to determine influencing factors on headwater streams, the processes these factors represent and how these factors and processes vary spatially. Understanding the strength and contribution of these factors is critical in developing a meaningful model and for evaluating its success in predicting the presence and extent of headwater streams.