IMPROVING UPLAND DRAINAGE REPRESENTATION USING LIDAR

Understanding surface runoff patterns is fundamental to evaluating the transport of water and materials through a landscape. These patterns, both spatial and temporal, are strongly controlled by the density of upland flow paths, which are often underrepresented in published drainage network data. However, high resolution elevation datasets are increasingly becoming available and present an opportunity to better quantify the extent and patterns of these first order networks. This research explores the potential to use these data for the mapping of first order channels in the post-glacial landscape of coastal Maine.

LiDAR elevation data in both point cloud and two meter resolution digital elevation model raster form are now available for all of coastal Maine, extending inland to the head of tide of major rivers. Data of this resolution, at a correlated length scale of first order channels, make it possible to remotely map upland drainage networks using direct detection methods that analyze small topographic variations in digital elevation models to directly find channel head locations. This is in contrast to process-based methods such as accumulation area to initiation, which can at best make good estimations of channel head locations. Direct detection methods have the advantage of being able to pick up artificial or modified channel heads such as drainage ditches or pipe outlets that do not follow natural initiation processes, which is useful in urbanizing areas.

Here we discuss the development and calibration of code that employs one direct detection method, topographic openness, which has successfully been used in the Mid-Atlantic region of the United States. In testing, we use LiDAR data from the Webhannet River in Wells, ME, a low-lying coastal watershed; and Cromwell Brook in Bar Harbor, ME, a small, high-relief island watershed that begins on Cadillac Mountain in Acadia National Park. Future work will focus on the use of direct detection-derived drainage networks to investigate changes in upland surface drainage patterns with urbanization.