Quantifying active deformation on a dry maar’s bottom through a light unmanned aerial vehicle and Structure-from-Motion

Abstract

ABSTRACT In this work, we quantify the differential subsidence that occurred between 2015 and 2016 in a maar-lake that has been desiccating due to the large groundwater extraction rates of the aquifer that used to feed it. The relatively small size of the bottom of the crater (1.2 km in diameter) and the scale of its structures (e.g. mud-injection domes that range in height from a few decimetres to 12 m and individual fault scarp heights that vary from a few cm to 15 m), along with the high albedo of the maar’s sediments provide a difficult setting to develop a series of high-resolution Digital Elevation Models (DEMs) required to quantify its subsidence. Through the use of a small quadcopter that carried a consumer-grade compact camera along a portable Global Positioning System (GPS) – which was used to georeference the acquired images – we were finally able to develop two high resolution (4.7 cm) DEMs for years 2015 and 2016. These DEMs allowed us to identify the geologic structures (faults, domes, and fault scarps) developed on the dry bottom of the crater. Through a Difference of Digital Elevation Models (DoDs = (DEM2015)-(DEM2016)), we were able to prove that subsidence is not uniform in the crater, as subsidence is mainly occurring on its western side. We also found that the largest subsidence occurs on the crater’s southwestern region (with some areas subsiding more than 1 m) followed by the northwestern area – where the crater’s highest normal fault is found. From these results, we can conclude that: (1) movement of the sediments inside the crater can not be interpreted as a simple piston-like downward displacement occurring near the centre of the basin and (2) Unmanned Aerial Vehicles (UAVs) are an efficient tool to quantify terrain changes as periodic flights can be repeated over the same area.