Remote Sensing & Geospatial Analysis Laboratory

May 24, 2018

May 31st – RSGAL Gavin Forster presenting @ SEFS Capstone Symposium

Loader
Loading...

EAD Logo
Taking too long?

Reload Reload document

|

Open Open in new tab

Download

If you missed Gavin’s presentations at the 21st Annual UW Undergraduate Research Symposium, you have one more chance at the SEFS Capstone Symposium.

Mapping Landslides on Mt. Baker-Snoqualmie National Forest with New LiDAR-based Contour Connection Method

Landslides are frequent hazards along the west side of the Cascade Mountains that result in major economic, environmental, and social impacts. The Mt. Baker-Snoqualmie National Forest (MBS) has a heightened awareness of landslide risks and potential consequences, particularly since the catastrophic Oso Landslide of 2014. Mapping of existing landslides, which tend to be the areas at highest risk of future instability, is a challenging, time-consuming, and expensive process. Current landslide mapping techniques tend to be inconsistent due to subjective interpretation by individual geologists and most result in insufficient resolution. In this study I use a new mapping tool, called the Contour Connection Method (CCM) that utilizes bare earth LiDAR, to detect landslide deposits on MBS land in an automated manner. The CCM approach requires less user input than other mapping algorithms, and focuses on general landslide geometry such as the slope of landslide scarps and deposits. Use of CCM also provides an opportunity to evaluate very large areas within matters of minutes or hours whereas the same areas, if evaluated based on field inventorying or manual interpretation of the high resolution DEMs, would have taken weeks or months. Publically available LiDAR data from the Department of Natural Resources was first formatted to be used with the CCM tool and then input into the program. Once the CCM landslide map was completed, I compared existing landslide inventory maps to the CCM map to assess the degree of agreement between the results and analyze the overall improvement in landslide mapping using this procedure. The implications of this study point to a less subjective, improved, consistent, and rapid framework for inventorying classified landslides and creating improved maps for designating activity avoidance areas for future projects, and for designating landslide areas crossed by existing infrastructure.