Evaluation Of Digital Surface Models Determined From UAS-Based Cameras Over Great Lakes Ice Features

Keywords: Sea Ice, Great Lakes, Unmanned Aerial Systems, Geographic Information System, Digital Surface Models
Abstract Type: Guided Poster Abstract

Abstract

Understanding how to collect ice data via Unmanned Aerial Systems (UAS) based cameras are integral to the future study of ice topography which helps determine ice thickness, seal nesting areas, and estimate hazardous ice for navigation. Supporting the work of a previous campaign, this project will examine the ability of 3 cameras mounted on a UAS to determine ice surface elevation measurements over the Straits of Mackinac using photogrammetry. The imagery was calibrated and normalized to create 2D orthomosaics and Digital Surface Models (DSMs) for each camera (DJI Matrice 600, Multispectral TetraCam MCAW, and GoPro Hero 9) through Pix4D software. A preliminary evaluation of the DSMs suggested that the DJI and GoPro produced the most vivid ice visualizations. Upon further analysis, the TetraCam produced the most realistic figures of ice feature elevation. Using GIS to normalize the datasets by converting each camera’s values to non-unit measurements allowed for comparison across the data, determining overall areas of higher and lower ice feature elevation. Therefore, it is recommended that all cameras be used for the next step of the campaign in Alaska to obtain both higher quality, clear DSMs, and provide relative accuracy of the data. Further steps taken to improve data collection, DSM calculations, and processing will resolve issues and enhance the results. Successful implementation of UAS devices to collect cryospheric data will reduce human risk and financial concerns in data collection, improve climate planning, and support future ice studies in global and rural areas.