Multi-approach monitoring of a high-risk sinkhole in the Ebro Valley, Spain using high-precision leveling, terrestrial laser scanner, photogrammetry, and BOTDA distributed optical fiber sensing

Abstract

<p>The analyzed active sinkhole is located on the outer bank of a meander of the Ebro River, affecting the Alcalá de Ebro village and a flood-control embankment dike. Between 1927 and 1957 the Ebro River channel experienced a major shift of 500 m until its present position abutting the village. The development of the sinkhole, around 100 m long, is related to dissolution of cavernous salt-bearing evaporites (i.e., halite and glauberite) underlying the unconsolidated alluvium. It experiences progressive sagging subsidence punctuated by the catastrophic occurrence of nested collapses in the village and the dike since 2007, leading to the demolition of a building and recurrent high-risk and uncertainty situations during floods. The sinkhole has been partially treated with a number of costly remediation measures, including shallow injection of polyurethane foam, compaction grouting and the installation of geogrids. Subsurface information on the spatial extent of the dissolution and subsidence phenomena has been obtained by boreholes, electrical resistivity tomography and ground-penetrating radar. Surface deformation at the site has been monitored utilizing multiple techniques, including: (1) several lines of high-precision leveling since 2015; (2) terrestrial laser scanner since 2014; (3) distributed strain and temperature sensing in optical fiber based on BOTDA since 2019; and (4) Structure from Motion Photogrammetry with drone images since 2020. The available surface displacement data provide an opportunity to: (1) compare the performance of the different techniques and identify their strengths, weaknesses and complementarity for sinkhole monitoring; (2) analyze the impact of floods and the associated water-table rises and drops on the activity of the sinkhole; and (3) assess the performance of remediation measures by comparing subsidence data before, during and after their application. Regarding the latter point, surface displacement data indicate that some measures significantly reduced subsidence activity (compaction grouting reaching the karstification zone), whereas other measures (shallow injection of polyurethane foam) aggravated the situation (subsidence acceleration and expansion). Moreover, the location of some treatments shows a significant offset with respect to the active subsidence area (i.e., inadequately sited without displacement data). The rapid sagging subsidence occurring at the present time on a paved area 11 m across is ascribed to a large cavity spanned by a geogrid, which eventually might produce a damaging catastrophic collapse. Precisely identifying the area affected by sinkhole subsidence and characterizing the spatial and temporal patterns of the surface displacement is essential for assessing the associated hazard and designing effective remediation measures. The most suitable monitoring techniques largely depend on the subsidence mechanisms (sagging vs. collapse), displacement regime (progressive vs. episodic), subsidence rates, and characteristics of the area.</p>