Use of Very High Resolution Remote Sensing to Derive Geomorphic Features Indicative of Stream Stability

    Streams are dynamic landscape features that vary across temporal and spatial scales. A stream attempts to maintain a stable state through the adjustment of slope or gradient, sediment load, width to depth ratio, and discharge. Climatic events and anthropogenic disturbance may cause instability through the alteration of one or more of these variables, causing the stream to adjust until it returns to a stable state. These changes can often be detected in the straight or meandering pattern exhibited by a stream. While it is possible to study stream pattern fluctuations in the field, the area covered by streams within multiple watersheds requires a view from above that is the planform pattern of the stream, where it is possible to assess the current condition and conditions that have interacted or persisted over years or decades.

    In Vermont, the Agency of Natural Resources River Management Program (VT ANR/RMP) is tasked with assessment and monitoring of streams to maintain long-term ecological and economic goals (VT ANR, 2004). VT ANR/RMP utilizes a three phase protocol for the collection of stream geomorphic data of which some processes are automated through a variety of software packages. However, VT ANR/RMP estimates one to three months are necessary to conduct the initial phase for a watershed and one to two days per stream mile are necessary for the second phase in a given stream (VTANR, 2004). The assessment of all Vermont streams could be a daunting task when one considers that there are over 9000 miles of streams in Vermont and, as streams evolve from any number of impacts, there may be a need to revisit certain streams over time.

    With the advent of high resolution imagery and digital imaging technology, remote sensing provides new opportunities to study stream geomorphology and monitor stream stability. The ability to detect geomorphic features with remote sensing could provide oversight agencies such as VT ANR/RMP with cost effective tools to identify unstable stream reaches. The goal of this study is to assess the utility of very high resolution remotely sensed data (0.16m digital orthophotography, QuickBird satellite data, and LIDAR) combined with advanced object-oriented classification techniques to map and monitor geomorphic features (i.e. sinuosity, channel and valley gradient, sediment bars, riparian vegetation, impervious surfaces) indicative of stream condition.

Check out our poster!