Active geomorphic features of rivers like sandbars provide habitat for endangered and threatened riparian plant and animal species. However, human development has altered flow and sediment regimes, thus impairing formation of sandbars and islands. Large scale mapping of the fluvial geomorphology in river ecosystems like the Connecticut River is are necessary to understand the dynamics of these features and preserve habitat. Orthophotographs from 2012 from United States Department of Agriculture's Farm Service Agency (FSA), National Agriculture Imagery Program (NAIP) were used to develop a model in ArcGIS Pro to identify fluvial geomorphic features in the Connecticut River and 12 of its major tributaries. This multi-stage image classification model identifies and ranks pixels of proximity and similar color to identify and map sandbars and islands. Locations of sandbars distribution were mapped and analyzed for each river. In the majority of rivers, sandbar area per reach decreases downstream. For the mainstem, sandbar area decreased towards the mouth but with three increases of sandbars due to meandering and major tributary confluences of the White and Deerfield rivers. Dams tend to decrease sandbars downstream but the effect of dams is context specific. Sandbars are stored upstream of the impoundment on the Black River as expected, sandbars appear downstream of a dam on the mainstem if a tributary confluence is present. Conservation of high sandbar area reaches and naturally eroding stream banks are necessary for preservation of endangered species. This spatial model for sandbar mapping can be applied in other river ecosystems across the region.
| Identifer | oai:union.ndltd.org:UMASS/oai:scholarworks.umass.edu:masters_theses_2-1643 |
| Date | 21 March 2018 |
| Creators | Backiel, Bogumila |
| Publisher | ScholarWorks@UMass Amherst |
| Source Sets | University of Massachusetts, Amherst |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | Masters Theses |
Page generated in 0.095 seconds