DETERMINING STREAM COMPLEXITY USING LIDAR DATA: FOCUSING ON THE PRESENCE OF LARGE WOOD AND DAMS IN NEW ENGLAND STREAMS

Quantifying channel complexity is a long standing area of research in geomorphology. Complexity is important to have full understanding of retention of organic matter, sediment transport, flow routing, and habitat quality. Large wood (LW) is a central factor in increasing the complexity of streams by altering flow patterns, causing overbank flows and increasing anabranching. The primary goal of this study is to predict the presence of wood from LiDAR data in streams by comparing field survey data to variability of longitudinal profiles. The secondary purpose is to understand how wood alters complexity at both the planform and profile scale, and if we can measure these changes using remotely collected data. We are attempting to understand the direction of change in complexity created by stream blockages in the form of anthropogenic dams, beaver dams and LW. Field studies were conducted in three New England watersheds with drainage areas ranging between 3.57 to 13.34 km2 from May to August 2015. Each respective basin contains large quantities of LW, beaver dams, and two out of the three reaches contain anthropogenic dams. Surveyed LW was restricted to pieces greater than 1 m in length and 10 cm in diameter, with the average for all three reaches being 5 m and 28 cm, respectively. Using a global positioning system, the coordinates were determined and we determined the type of stream blockage from field observations. These GPS coordinates were then imported into ArcGIS, and analyzed simultaneously with LiDAR data of each watershed. LiDAR elevation values were measured along the channel centerline. The thalweg was smoothed using a moving-average filter and the remaining minima connected to ensure a monotonically decreasing downstream profile. Then the residuals between the raw and smoothed profiles were calculated. We hypothesize that the residuals from a longitudinal profile, developed in ArcGIS using 1 meter LiDAR data, will correspond to locations of dams and LW jams. Preliminary data analysis suggests that in steeper reaches wood jams are related to areas of greater bed variability whereas in the low gradient parts of the system the wood is undetected using LiDAR analysis. We hope that future analysis will enable us to identify pool frequency in relation to wood locations using LiDAR data.