CUR - The Council on Undergraduate Research

Photogrammetric Reconstruction of Roof Snow Load

Madeline Beck, Roxanne Holmes and Eric Sproles, Department of Earth Science, Montana State University, Culbertson Hall, 100, Bozeman, MT 59717

In March 2019, two flat-roofed recreational gyms of the Marga Hosaeus Fitness Center at Montana State University collapsed. While the accumulated snowpack was above average at that time, these collapses were not expected based on the current understanding of total load capacity. Therefore, our innovative project is designed to provide facilities management with accurate snow load and depth measurements on a selection of flat-roofed buildings around campus to complement their existing systems of measuring roof collapse risk.  This project uses Unmanned Aerial Vehicles (UAVs) to measure snow depth using photogrammetric reconstruction. UAV technology is still new and has not yet been applied to monitor rooftop snow. To gather the data, UAV flights are conducted over the buildings of interest and a series of geo-located photos are taken in a predetermined flight path. Photogrammetric processing of the photos constructs a 3-D model of the buildings and their roofs without snow.  Later 3-D models with snow present provide maps of snow depth on the rooftops at sub-meter resolution. The snow depth maps are combined with snow density measurements from a rooftop snow scale and snow depth sensor to provide a map of snow load for the buildings selected. This map is shared with facilities management through a web-based interface to identify if, when, and where snow mitigation is necessary on existing buildings.

Additional Abstract Information