GPR investigation of karst guided by comparison with outcrop and unmanned aerial vehicle imagery Antonio L. Fernandes Jr. a , Walter E. Medeiros a,b,c, ⁎, Francisco H.R. Bezerra a , Josibel G. Oliveira Jr. b , Caroline L. Cazarin d a Programa de Pós-graduação em Geodinâmica e Geofísica, UFRN, Natal, RN, Brazil b Departamento de Geofísica, Centro de Ciências Exatas e da Terra, Universidade Federal do Rio Grande do Norte, UFRN, Natal, RN, CEP 59072-970, Brazil c INCT-GP/CNPq, Instituto Nacional em Ciência e Tecnologia em Geofísica do Petróleo, Brazil d CENPES/PETROBRAS, Av. Horácio Macedo, 950, Cidade Universitária, Rio de Janeiro, RJ CEP 21941-915, Brazil abstract article info Article history: Received 18 June 2014 Accepted 25 November 2014 Available online 29 November 2014 Keywords: Karst GPR UAV imagery Potiguar basin Fracture Dissolution The increasing importance of carbonate rocks as aquifers, oil reservoirs, and for urban problems is demanding detailed characterization of karst systems, a demand that can be partially satisfied with GPR imaging. However, the goal of imaging and interpreting karstified carbonate rocks is notoriously difficult due to the complex nature of the geometry of the dissolution and the GPR intrinsic limitations. One way forward is the direct comparison of GPR images with similar outcropping rocks. A joint study involving a 200 MHz GPR survey, unmanned aerial vehicle imagery (UAV), and outcrop characterization is presented aiming to improve the interpretation of sedimentary structures, fractures and karst structures in GPR images. The study area is a 500 m wide and 1000 m long carbonate outcrop of the Jandaíra Formation in Potiguar basin, Brazil, where sedimentary, fracture, and karst features can be directly investigated in both vertical and horizontal plan views. The key elements to interpret GPR images of karstified carbonate rocks are: (1) primary sedimentary structures appear in radargrams as unaltered imaged strata but care must be taken to interpret complex primary sedimentary features, such as those associated with bioturbation; (2) subvertical fractures might appear as consistent discontinuities in the imaged strata, forming complex structures such as negative flowers along strike–slip faults; (3) dissolution may create voids along subhorizontal layers, which appear in radargrams as relatively long amplitude shadow zones; and (4) dissolution may also create voids along subvertical fractures, appearing in radargrams as amplitude shadow zones with relatively large vertical dimensions, which are bounded by fractures. © 2014 Elsevier B.V. All rights reserved. 1. Introduction Ground Penetrating Radar (GPR) has been increasingly used in the last twenty years as a non-invasive method for characterizing near sur- face clastic sediments, which allows in general high quality imaging of both stratigraphic and structural features (e.g., Bristow and Jol, 2003a; Neal, 2004). However, its use for imaging carbonate rocks (e.g. Martinez et al., 1998; Al-fares et al., 2002) appears to be comparatively sparse (Bristow and Jol, 2003b) at least until to the beginning of the last decade. According to Pueyo-Anchuela et al. (2009), GPR has been used in karst environment to map hazard zones, where the main task has been the detection of cavities and paleo-collapses that might be a potential risk to human constructions (e.g., Benson and Yuhr, 1989; Collins et al., 1990, 1994; Doolittle and Collins, 1998; McMechan et al., 1998; Chamberlain et al., 2000; Gómez-Ortiz and Martín-Crespo, 2012). In this context, detailed imaging of karst structures has been fre- quently overlooked, although the contribution of GPR goes beyond the mere detection of cavities and may unravel the evolution of karstic areas, specially when the GPR interpretation is integrated with geolog- ical and/or geomorphological data (Pueyo-Anchuela et al., 2009). In the last few years, the increasing importance of carbonate rocks has motivated an increase in detailed GPR studies in carbonate aquifers (e.g. Mukherjee et al., 2010; Chalikakis et al., 2011; Carrière et al., 2013), oil reservoirs analogs (Jorry and Bievre, 2011; Forte et al., 2012; Jahnert et al., 2012), and for urban problems (e.g. Pueyo-Anchuela et al., 2010). The last two cases demanded an advanced understanding of karst evo- lution (e.g., Pueyo-Anchuela et al., 2014). However, GPR investigations of karst systems still present problems (Grasmueck et al., 2013). For karstified carbonate rocks, besides characterizing the primary structures, considerable effort must be done to identify joints, fractures or faults (e.g. Jeannin et al., 2006), and dissolution structures such as caves from the GPR images. Moreover, it is necessary to establish spatial and genetic Journal of Applied Geophysics 112 (2015) 268–278 ⁎ Corresponding author at: Departamento de Geofísica, Centro de Ciências Exatas e da Terra, Universidade Federal do Rio Grande do Norte, UFRN, Natal, RN, CEP 59072-970, Brazil. Tel.: +55 84 3215 3794. E-mail addresses: antljunior@hotmail.com (A.L. Fernandes), walter@geofisica.ufrn.br (W.E. Medeiros), bezerrafh@geologia.ufrn.br (F.H.R. Bezerra), josibel@geofisica.ufrn.br (J.G. Oliveira), cazarin@petrobras.com.br (C.L. Cazarin). http://dx.doi.org/10.1016/j.jappgeo.2014.11.017 0926-9851/© 2014 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Journal of Applied Geophysics journal homepage: www.elsevier.com/locate/jappgeo