In Brazil, the verticalization of urban spaces due to higher living standards and the construction of tall buildings in urban areas, in addition to the progressive growth of land use by mining, civil engineering, and agriculture activities, have generated impacts on the physical environment, as in the case of karst terrains. In this context, the significant areas of karst terrains in several Brazilian cities (Cajamar-SP, Lapão-BA, Teresina-PI, Curvelo-MG, Bocaiúva do Sul-PR, Sete Lagoas-MG, Almirante Tamandaré-PR e Colombo-PR), means that there are potential geological risks for phenomena that generate subsidence and collapse of cavities under these cities. According to Vestena (2002), the subsidence which has occurred in the cities of Cajamar-SP, Sete Lagoas-MG, Almirante Tamandaré-PR, and Colombo-PR are associated with karst terrains, where the lowering of the water table contributed to the process of soil collapse / subsidence.
In light of this, the geological risk associated with karst areas is worrying since in this type of geological environment, soil subsidence and cavity collapse can become a social and economic threat, which is intensified by the urbanization of these cities (Vestana 2002; Waltham et al. 2005; Gutiérrez et al. 2015).
Karst terrains are formed by the surface and subsurface dissolution of carbonate and evaporite rocks, which are commonly characterized by challenging geotechnical conditions (Rodriguez et al. 2014). However, some authors call attention to less soluble rocks that can develop karst settings, such as sandstones and quartzites (Martini 2000; Frumkin 2013). It’s worth noting that in karst settings, problems related to subsidence and collapse can be caused by natural phenomenon or by anthropic activity. Subsidence sinkholes are the main risk in karst terrains (Waltham et al. 2005; Gutiérrez 2010; Rodrigues et al. 2014).
Considering the heterogeneous character and susceptibility of karst terrains to subsidence and collapse, geophysical studies and investigations using multiple approaches are extremely important for providing information on the subsurface, enabling the identification and characterization of features associated with sinkholes (Milanovic 2000; Thierry et al. 2005; Ezersky et al. 2006; Ezersky 2008; Argentieri et al. 2015). According to Lelièvre et al. (2009), due to the ambiguity of the results obtained by a single geophysical method, the integrated application of various geophysical methods in the study of an area is necessary in order to provide more reliable information and results (Dourado et al. 2001; Kruse et al. 2006; Lago et al. 2006; Lago et al. 2008; Nouioua et al. 2012; Sevil et al. 2017; Pazzi et al. 2018; Hussain et al. 2020).
The effectiveness of a geophysical survey is conditioned by the existence of contrasts between the measured physical property values. Therefore, considering changes in the physical properties of materials caused by the processes of dissolution, erosion, and subsidence involved in the development of sinkholes, geophysical methods are an excellent tool for indirect investigation (Hoover 2003). In general, the use of geophysical surveys in the characterization of karst terrains consists of the detection and mapping of the extension of sinkholes as well as information about the depth of the water table, direction of the underground flow, and depth of the karst rocks (Chalikakis et al. 2011).
Among the geophysical methods, the Electrical Resistivity (ER) method is widely used in several fields of study such as mining (Arifin et al. 2019), engineering geology (Rucker et al. 2011), hydrogeology (Revil et al. 2012), environmental studies (De Lima et al. 1995; Chambers et al. 2006), agricultural studies (Michot et al. 2003) and cavity studies (Smith 1986; Martinez-Lopez et al. 2013; Hussain et al. 2020). However, despite the countless geophysical investigations carried out on karst terrains worldwide, (mainly for mapping cavities) the Ground Penetrating Radar (GPR) method has proven to be the most efficient geophysical method for identifying geometric karst features in urban environments. As such, over the past couple of decades the use of the GPR method has increased and many improvements have been successfully implemented (McMechan et al. 1998; Zisman et al. 2005; Kruse et al. 2006; Rodriguez et al. 2014; Sevil et al. 2017; Hussain et al. 2020). In light of the results of these authors, the GPR method has shown effectiveness in mapping urban areas affected by subsidence and collapse in karst environments and has contributed to a better assessment of the risks associated with this geological environment.
The physical principle and data acquisition of GPR methodology are similar to the seismic reflection and the sonar techniques, except that the GPR is based on the reflection of electromagnetic waves (Casas et al. 2000). According to Annan (2002), this method stands out for shallow investigations, due to its high resolution and the acquisition of a large volume of data in a short period of time. The depth of investigation is a limitation of the GPR method and can be influenced by the following factors: geometric scattering, attenuation by the terrain, and partition of energy at the interfaces, which are all related to the loss of energy during the propagation of the electromagnetic wave (Bradford 2007). The depth of investigation and the resolution of the GPR vary according to the frequency of the antenna. The higher the frequency, the higher the vertical resolution and the lower the depth of investigation, and vice versa.
Kruse et al. (2006) in the publication “Sinkhole Structure Imaging in Covered Karst Terrain”, show that GPR and resistivity techniques have been widely used to map the locations of sinkholes in covered karst terrain. The authors acquired GPR and resistivity data in the west-central region of Florida, USA. According to the authors, the GPR method provided detailed information on the geometry of sinkholes developed within covered karst terrain. Rodriguez et al. (2014) applied the GPR method in two studies of sinkholes developed in the mantled evaporite karst of Zaragoza city in Spain, with the purpose of evaluating the potential of GPR for the characterization of sinkholes in covered karst. The authors concluded that GPR was an effective technique for the identification and characterization of shallow collapse and sagging subsidence structures in covered karst areas. GPR allowed for the reliable mapping of the limits of the sinkholes (characterizing their internal geometries), their inferred subsidence mechanisms, and estimates of their morphometric parameters. Continuing geophysical studies in the city of Zaragoza, Spain, Sevil et al. (2017) analyzed and compared the data acquired by the GPR and Electrical Resistivity Tomography (ERT) methods. In addition, the authors suggested the use of shielded antennas as the best option for surveys with the GPR method in urban areas.
Within this context, the study at hand consists of the imaging of subsidence areas by the GPR method in the city of Teresina-PI. According to geotechnical studies, 9% of the city of Teresina-PI has a high chance of a collapse and 45% of the city has a medium chance of collapse (Aquino 2020). The collapse and subsidence of soil and rock are the result of the evolution of karst reliefs, the degree of dissolution of the rock, and the evolution of cavities in the subsoil. The city developed mainly on sedimentary rocks included in the Pedra de Fogo Formation (Permian). The rocks of this formation are susceptible to dissolution processes, as seen in the processes of subsidence and / or collapse that have occurred widely in the central region of the city.
The Geological Survey of Brazil (CPRM), as part of the Emergency Action of Sectorization of Areas with Risks of Subsidence in Karst Environments, carried out a geophysical study with GPR for the evaluation and characterization of the sinkhole events that occurred in the city of Teresina-PI. Taking into consideration the lack of published articles of this nature in the technical scientific literature, this article is innovative in advancing the knowledge of this particular case study in Brazil. The article presents a model which includes a set of geological and pedological factors that caused the sinkholes and collapses in the city of Teresina-PI, as well as integrating results obtained through the GPR method with data from regional aeromagnetometric studies.
The capital city of Teresina, in the State of Piauí, is located on the banks of the Parnaíba River and covers an area of 1,756 km2. The central area of the city, located between the Parnaíba and Poti rivers, has a highly fragile subsoil and subsequently a history of sinkholes, which have been triggered by the drilling of tubular wells, fluctuations in the water table, and leaks in the sewage and hydraulic networks (Pimentel 2008; Fig. 1). Two sinking events deserve to be highlighted: the first occurred in 1999 at Simplício Mendes Street and was related to the drilling of a cavity in the subsoil that caused the collapse of both homes and the street itself (Fig. 2a). The second occurred on July 31, 2008, at Francisco Mendes street (Fig. 2b).
Teresina-PI was created in 1852 as a planned city with a geometric shape in the form of a chessboard, and was the first planned city in Brazil (Viana 2007). From 1970 onwards a process of verticalization began, which accelerated strongly in the late 1980s with the construction of many high-rise buildings (Façanha 1998, 2003). However, it is not possible to say whether the geotechnical weaknesses characteristic of the soils and rocks of Teresina were well known to the municipal technical entities and particularly to the construction companies operating in the municipality.
On Francisco Mendes street, the soil (silty sand) presented mechanical and microstructural behavior typical of collapsible soil as shown in the geotechnical study made by Aquino (2020). From the author’s perspective, there was no single cause for the sinking of this street, rather there were a combination of factors which contributed to the triggering of the accident, both of a pedological and geological nature. The collapse events in the city were related to natural factors aggravated by anthropic action including: leakage from a pipe, infiltration of rainwater, leakage in the hydraulic network, fluctuations in the water table, and tectonic events (leading to fractures).
Fractures are associated with the formation of dolines or small cavities in the subsoil, leading to the possibility of free circulation of water and gases. The limestone intercalation with other rocks, typical of the Parnaíba sedimentary basin (Castro et al. 2016), increases the dissolution power of the water on the rocks.
The city of Teresina-PI is developed mainly on sedimentary rocks included in the Pedra de Fogo Formation, which consists of alternating horizons of medium to fine sandstones with cross-stratification and discontinuous silexite and carbonate horizons. At the base of the Pedra de Fogo formation, there are soft sandstones (fine kaolinitic) that are very friable. These lithologies are overlapping the sediments of the Piauí Formation.
The sediments of the Pedra de Fogo Formation are susceptible to dissolution processes that occur at carbonate levels and which may be a condition of the collapse processes that are occurring widely in the central region of the city. The collapses in Teresina-PI are related to natural factors aggravated by anthropic action (rainwater infiltration, leakage in the hydraulic network, water table fluctuations and geological faults), and considering these factors, a Geological Model was proposed in this work for the city (Fig. 3).