Joint 69th Annual Southeastern / 55th Annual Northeastern Section Meeting - 2020

Paper No. 12-25
Presentation Time: 8:00 AM-12:00 PM

GROUND AND DRONE BASED PHOTOGRAMMETRIC MAPPING AND MODELLING OF SHORELINES, OUTCROPS, AND RELATED ARTIFACTS


BLACKWELL, Arlene E., CERRA, Jack V., WHITCOMB, Danielle A. and HYATT, James A., Environmental Earth Science, Eastern Connecticut State University, 83 Windham Street, Willimantic, CT 06226

Digital photogrammetry (DP) is a versatile and increasingly popular technique used by the geologic community to map, model, and measure eroding coastlines, outcrop, and other targets. Here we describe our efforts to learn and use ground and airborne DP to examine coastal sites on Block Island, RI and an outcrop of glacially eroded Jurassic conglomerate in Glastonbury, CT. Field work involved collecting high quality overlapping images and surveying ground control points (GCP) to construct models of eroding shorelines at Clay Head (CH) and the West Beach (WB) Transfer Station on Block Island, and an outcrop at Dufford’s Quarry (DQ). Ground-based images were captured using a tripod-mounted Nikon D610 full-frame camera while a Mavic Air DJI Drone was used to manually fly and image outcrop from elevations of 5, 15, and 30 meters above ground. The D610 was also used in the lab to image/model eroded samples of cobbles (CH) and metal debris (WB) for 3D printing. GCP’s for approximately 10 photogrammetric targets placed at different elevations were located using RTK-GPS (CH and WB) and a total station (DQ). Approximately 79 (CH), 327 (WB), and 27-81 (DQ) images were analyzed using Agisoft Photoscan to model the field sites, while images of the CH cobbles (n=138) and WB metal (n=153) were captured from all angles using a turntable in a light tent. Image processing and DP modelling, as explained in our poster, is complex making use of Lightroom, Photoscan, Blender, and Cloud Compare software. Ongoing analyses seek to quantify coastal change (i.e. erosion and deposition) by comparing 2019 models with previously constructed 2017 and 2018 models. WB experienced significant (1-2 m) bluff retreat from 2017-18 and a complete change in 2019 due to remediation and construction of a revetment. Changes at CH are subtler with erosion evident in rills and by undercutting cemented cobbles, while deposition created small talus cones. DQ outcrop model elevations differ most at locations furthest from GCP’s. Large clasts can be recognized and measured best using low altitude (5 m height) models. Glacial striations and grooves in the conglomerate are visible in model DEM’s but are only recognized in extracted cross-sections from the lowest flight-line model.