The Iglesia basin (San Juan, Argentina), seismic interpretation, geometry basin, and implications for geothermal systems

. The Iglesia basin, in the San Juan Province, is the northern part of a Cenozoic piggy-back basin (the Iglesia-Calingasta-Uspallata basin), located above of the active flat-slab subduction zone of western Argentina. This basin is located between two mountain ranges, the Cordillera Frontal to the west and the Precordillera to the east, affected by thin and thick skinned tectonics, respectively. It is elongated in a north-south direction (70 km) and has a maximum width of 35 km. We have analyzed 17 seismic reflection lines corresponding to more than 500 km of interpreted sections. A strong reflection at 0.5-2 s is interpreted as indicative of shales and sandstones in the upper part of the Agua Negra Formation (Carboniferous); this represents a seismic basement for the Iglesia basin clearly separating stratified (shallower) from chaotic (deeper) reflectors. Several high angle faults have been recognized in the interpreted seismic lines affecting the Neogene fill and even the basement; some of them correlate with structures observed at the surface such us El Tigre Fault System. A 3D model shows the seismic top of basement smoothly deepening down to 3,650 m beneath the Pismanta area. The basement continues upward to the east with a steeper slope clearly defining an asym-metrical shape for the basin. Moreover, there are three depocenters of similar maximum depths around the geothermal hot spring of the Pismanta center. Some of the interpreted faults may be helping in the outflow process of meteoric water heated by a normal geothermal gradient. Considering earthquake data framework, our observations correlate with transpressive deformation likely associated with El Tigre Fault System, which affects Neogene and Pleistocene strata of the Iglesia Group.


Introduction
The Iglesia basin is located 300 km east of the pacific trench related to the interaction of the Nazca and South American plates, with a convergence rate of about 64 mm/yr (Vigny et al., 2009), and about 110 km above the subducted slab (Ammirati et al., 2016). The basin is the northern portion of the ~150 km long Iglesia-Calingasta-Uspallata basin between the Precordillera in the east, with an elevation of about 2,500 m a.s.l. and the Cordillera Frontal in the west, with an elevation of about 5,000 m a.s.l. (Figs. 1 and 2). The oceanic Nazca plate slab in this segment of the Andes has a nearly horizontal position, something only observed in 10% of the convergent margins on Earth. Not many of these flat-slab regions show mountainous continental chains with a wide intervening basin (like the Iglesia basin) in their backarc regions (Gutscher et al., 2000) (Fig. 1). Shallow subduction processes at the latitude of the study area (30° S) seem to have started at ca. 10-5 Ma according to magmatic, sedimentary and thermochronological studies (Jordan et al., 1993;Kay et al., 2005;Fosdick et al., 2015;Mackaman-Lofland et al., 2020).
Previous studies in this northern area of the Iglesia-Calingasta-Uspallata basin proposed that it includes a Neogene sequence up to 3.5 km thick, and it is interpreted as a piggyback basin that was developed during the eastward propagation of deformation within the Precordilleran thrusts Jordan et al., 1993). Former interpretations using seismic reflection data showed the Iglesia basin highly affected by a zone of deformation in its eastern part (Álvarez-Marrón et al., 2006), with steep reverse faults and folding of the sedimentary fill near the locality of Rodeo (~30.2° S; ~69.0° W). This area also contains the presence of the Pismanta mesothermal (45 °C) water spring center whose temperature is lower than that of the geothermal centers in the high Cordillera of about 75 °C (Sussini et al., 1948) (Fig. 1). This has gained importance in recent years due to the possibility of geothermal energy generation and socioeconomy-tourism development. Because of the flat slab subduction, there is no active volcanism in the Frontal Cordillera and Precordillera in the San Juan area (Kay et al., 1991). The association of geothermal areas in the Precordillera with a volcanic origin is thus not expected; however relatively higher temperature centers are present in the high Cordillera, such as Despoblados water springs linked to a magmatic heating source (Sussini et al., 1948); those areas, are located in remote low populated areas in the northwest of San Juan. In the Iglesia Valley there are also other thermal spring sources such as Centenario, Poblete at about 30 km distance from Pismanta, which are considered as part of the same geothermal system in this study ( Figs. 1 and 2).
This work presents a characterization of the Iglesia Valley basin, generating a Move model (MVE: Midland Valley and Petroleum Experts. https://www. mve.com/) with the main goal to delimitate its seismic basement as well as other geological structures around the hydrothermal system of Pismanta with available information. The shape of the basin (with its points of maximum depth) will be correlated with the location of the geothermal sources. We interpreted about 500 km form a set of seventeen seismic reflection lines acquired by YPF S.A. in the 90's. The work involved the 2D analysis of each seismic line in order to identify major faults, folds, the basement shape, the horizontal continuity of geological units as well as their correlation in contiguous 2D lines and in the surface.

Geological framework
The region of study comprises two different tectono-stratigraphic terranes, Chilenia in the west and Cuyania to the east (Fig. 1), being separated by a dissected belt of ophiolitic rocks. The Chilenia terrane is composed of high grade metamorphic rocks exposed to the west of the ophiolitic belt . According to Ramos (2004) the Cuyania terrane, in the Western Sierras Pampeanas and the Precordillera, exhibits Grenville rocks and Laurentian carbonate platform rocks of Cambrian-Ordovician age (Astini et al., 1996) (Fig. 1). Ramos (1984) suggests that the Chilenia and Cuyania terranes were both part of Laurentia and were later accreted to the western margin of Gondwana during Ordovician and Devonian time. The presence of about ~300 km igneous calcoalkaline rocks of Eopaleozoic ages has been related to the approach and collision of the Chilenia terrane, although its genesis and the geotectonic framework in which they have been originated are not fully understood (e.g., Pérez, 2015).
The Iglesia basin is located in the northwestern part of the San Juan province, between 30° S and 31° S and between 69° W and 69°30' W. From the Miocene, the compressive tectonics controls the structure of this region due to the flat slab subduction between the Nazca Plate and the South American Plate between 28° S and 32° S (Fig. 2) (Barazangi and Isacks, 1976;Cahill and Isacks, 1992;Anderson et al., 2007;Ammirati et al., 2015Ammirati et al., , 2016. To the west, the tectonic of the Frontal Cordillera is dominated by mechanical basement- FIG. 1. Study region (rectangle) showing morphostructural units and terranes (Ramos, 1995(Ramos, , 1999Astini et al., 1996), the 1894 (M>7) earthquake epicenter (yellow dot), and the cross section over the Pismanta geothermal system (A-B), which is shown in figure 2. Also shown are the depth contours, in km, of the Nazca Plate (NP) beneath the South American Plate (SAP) according to Anderson et al. (2007) and localities (black squares) mentioned in the text. The arrow in the inset represents the GPS relative convergence velocity between the NP and SAP according to Vigny et al. (2009). involved uplifts that exhume predominantly Upper Paleozoic-Triassic rhyolites and granitic intrusions, in addition Devonian-Carboniferous metamorphic and sedimentary rocks are also exposed (Heredia et al., 2002;Kleiman and Japas, 2009;Vallecillo et al., 2010;Rocher and Vallecillo, 2014). To the east, the Western Precordillera thin skinned foldthrust belt bounds the basin; this fold-thrust belt extends to the foreland accommodating the majority of the retroarc shortening (>100 km) and involves Paleozoic-lower Mesozoic sedimentary rocks and Cenozoic deposits Von Gosen, 1992;Cristallini and Ramos, 2000). Sedimentological studies carried out by Wetten (1975), Beer et al. (1990) and Ré et al. (2003) have described the Neogene to Pleistocene fill of the Iglesia basin, which is composed of sedimentary sequences of continental origin. The Iglesia Group involves the Pliocene Las Flores Formation and Miocene Lomas del Campanario Formation. (Fig. 3). According to Allmendinger et al. (1990), this sedimentation coexisted with the thrusts of the Precordillera whose main structures dip westward; thus, this basin constitutes a piggyback basin, which is being passively transported eastward on the activity of the Frontal Cordillera. The Carboniferous shale, sandstone and conglomerate of the Agua Negra Formation are observed on the west side of the valley, as part of the Frontal Cordillera, but not in the Precordillera on its eastern edge (Figs. 2 and 3); Yerba Loca Formation (composed of shales, sandstones and Ordovician basalts) is also observed on the west side of the valley (Fig. 3). We assume this may correspond to the basement of the Iglesia basin. The volcanic FIG. 2. 3D view (upper panel) and elevation profile A-B (lower panel) across the Pismanta geothermal system (see Fig. 1 for location).
The Iglesia Valley elevations are between 1,600 and 3,500 m a.s.l. The black arrow in the upper panel denotes the relative plate convergence velocity and direction according to Vigny et al. (2009). Blue, yellow and red arrows in the lower panel represent the schematic meteoric water track.
Permo-Triassic rocks of Choyoi Group also outcrop to the west of the valley (Frontal Cordillera) but not in its eastern margin (Precordillera). Wetten (1975) described the Pliocene Las Flores Formation as an upper member composed of conglomerates and a lower member made up of clays, limolites and fine sandstones. The Miocene Lomas del Campanario Formation is composed of two members, an agglomerate member consisting of pyroclastic material, andesites, dacites and andesitic tuffs and a conglomerate member with a high content of volcanoclastic rocks (Wetten, 1975). A Miocene to Pliocene pyroclastic sequence associated with lava flows and subvolcanic rocks is exposed along the Iglesia Valley Basin (Poma et al., 2017). In This sedimentary filling of the basin is consistent with predominantly Neogene deposits Ré et al., 2003) originally deposited in a former Oligocene foreland basin (Jordan et al., 2001). Structurally, the main feature in the western sector of the Precordillera is the El Tigre fault extending for more than 100 km between the Jáchal and San Juan rivers ( suggested for this regional fault (Bastías, 1986;Wetten, 1975;Siame et al., 1996Siame et al., , 1997Fazzito, 2011;Ruskin and Jordan, 2007). Pérez and Costa (2006), Perucca and Vargas (2014) and Rivas et al. (2019) have proposed that the non-instrumentally recorded 1894 ( Fig. 1) earthquake of magnitude M>7 (Richter scale) may have occurred in the northern continuation of the El Tigre fault system. Some exposed segments of this fault show an average strike of N10°±6° and an almost vertical dip (Costa et al., 2000); it affects six levels of alluvial fans with ages varying between 700,000 years for the oldest levels and 21,000 years for the most recent surfaces of the current level (Siame et al., 1997).

Hot springs
Several hot springs are located in the eastcentral part of the Iglesia Valley near the town of Pismanta ( Fig. 1) and other neighboring areas such as Centenario, Carrizalito, Rosales, Horcones, Arnobio, Poblete and La Salud (Fig. 4). We have FIG. 4. Locations of hot springs with their temperatures (Sussini et al., 1948)  used basic geochemical data of fluids in this sector by Sussini et al. (1948) to infer features from deep fluid reservoirs and their possible association with faults in this area. Measurements of the water source temperature at surface indicate relatively low enthalpy hypothermal springs (Muffler, 1979;Hochstein, 1988;Benderitter and Corny, 1990;Haenel et al., 1988), with temperatures <45 °C. At a regional scale, higher temperatures (~70 °C) hot springs are observed in the high Andes westward (Sussini et al., 1948;Barcelona et al., 2014); in contrast lower temperatures (~20 °C) hot springs characterize the eastern Precordillera to the east (Fig. 4). One possible explanation for the relatively higher temperature water sources outflowing at the surface near Pismanta, in the central eastern part of the Iglesia Valley, would be that the high elevation regions in the west (>5,000 m a.s.l.) allow the infiltration of rain and snow meteoric water into the system. Then, differences in the slope (Fig. 2) favor the infiltrated water motion from west to east while it heats up by a normal geothermal gradient. The water hosted in deep reservoirs may be reaching the fault systems in the east (western-central Precordillera). Thus, the fault system plays a role controlling the hydrothermal system. Another hypothesis suggests that recent isolated magmatism would contribute as a source in the heating of infiltrated meteoric water in tectonic environments associated with horizontal subduction; an example has been described in Despoblados area in the high Andes as a geothermal system (see location in Fig. 4) by Barcelona et al. (2014Barcelona et al. ( , 2019a. Compositionally, the Pismanta waters are classified as sodium sulfates (SO 4 2-Na + ) (Sussini et al., 1948). Ca-Na, Ca-K, Na-K and Silica geothermometers (Tonani, 1980;Fournier, 1983) were used to obtain reservoir temperatures; they yielded values of 155 °C, 69 °C, 49 °C and 91 °C, respectively; two of these geothermometer values (Ca-K and Silica) were also reported in the inventory by Pesce and Miranda (2003). Assuming these values of 69 °C and 91 °C, respectively, and a normal geothermal gradient of 28 °C/km (obtained from the Ansilta oil exploration well in the Calingasta Valley), it is possible to estimate reservoir depths of 945 m and 1,800 m. However, it is worth to note that Sussini et al. (1948) have even observed a greater variation of more than 105 °C over the four geothermometers. Thus, we suggest the presence of two reservoirs located at depths of less than 1 km and more than 2 km, respectively. Data on stable isotope composition consisting in 18 O/ 16 O and 2 H/ 1 H ratios from Pesce and Miranda (2003) clearly show a meteoric origin for the water in three samples from thermal springs near Pismanta (Fig. 4). The Carrizalito and Arnobio Poblete springs show a significant 18 O-depletion likely due to isotopic exchange between liquid water and CO 2 , a process that typically affects CO 2 -rich bubbling pools (Epstein and Mayeda, 1953;Pauwels et al., 1997) and HCO 3 hydrothermally enriched waters (Peiffer et al., 2014) (Fig. 4).

Methodology
A 2D view was used line by line to recognize each reflector and faults. Furthermore, a 3D view was used at the end in order to identify the lateral continuity of the reflectors previously recognized in each seismic line individually and the continuity of faults as well.
In this work we have analyzed information from more than 500 linear km of seismic reflection data from the industry, located in the Iglesia Valley (Fig. 3). This work was done on the MOVE TM platform with an academic license (https://www.mve.com/). These lines were acquired using an energy detonating cord as a seismic source, with a fold of 2,400% (line 40144 has a fold of 1,200%). The length of the lines varies from 25 km to 100 km for the longest north-south line (5334). The processing of the seismic lines followed the standard sequence of pre-stacking, stacking and post-stacking (Yilmaz, 2001). In the pre-stacking stage, we removed the random noise and we applied filters and frequency analysis because of compensation for loss of amplitude. We also applied static corrections by refraction (Cox, 1999). A velocity analysis was made using the semblance method (Taner and Koehler, 1969;Neidell and Taner, 1971;Hubral and Krey, 1980). In the stacking stage, velocity analysis was carried out by the constant velocity method (Yilmaz, 2001). In the post-stacking stage, we removed incoherent noise applying the FX deconvolution method (Yilmaz, 2001). Figure 5 shows a strong reflector at about 0.7 s and 2 s in west-east direction (yellow color; top panel) for the 2D view of seismic line 5325. The same figure 5 shows the same line in depth conversion in the bottom panel. It made possible to define the basin basement and some discontinuities associated with faults in the east sector. The geological structures were identified in time and then a depth conversion was made. We note consistency for a similar reflector recognized in all analyzed lines; thus, we interpreted this as the basement of the basin. In addition, several faults were recognized in each seismic line (Fig. 6). After identifying these geological structures in each line, an attempt was made to correlate them with each other and with the structures on the geological map resulting from previous studies (Ré et al., 2003;Álvarez-Marrón et al., 2006). We used Move software (MVE) and a tool allowing to superimpose surface information such as lithology, geological structures, digital elevation model (DEM) (Fig. 3), among other features.
In order to accomplish a time to depth conversion for the seismic lines a velocity function, representative for the study area, was required. Because of the non-availability of any well data in this region, a direct seismic velocity structure was not an option. Thus, we assumed an average value of v=2,500 m/s for the P wave velocity. This velocity may correspond to the upper (less consolidated) sediments of the basin (Pleistocene and Pliocene) in the following equation. Therefore, for each two way travel time (TWT) a depth can be calculated using Z(t)=v (e kt -1) (Marsden, 1992), where Z is the depth in m, k is the rate of change of velocity with depth (we used a value 0.5 Hz),

Results
Assuming that the basement of the basin corresponds to the top of the Carboniferous rocks of the Agua Negra Formation and the velocity function previously described in methodology, the maximum depth to the basement top is of 3,650 m below the surface. This is compatible with results from Beer (1989) and Beer et al. (1990), who calculated a maximum depth of the basin of 3,500 m based on interval velocities calculated from seismic reflection data. Figures 8, 9 and 10 show a depocenter (red colors) with three subzones labeled as 1, 2 and 3 for the study region. The location of the maximum depths of the basin correlates with the location of the hot springs at Pismanta area. These three maximum depth areas (of about 20 km radius) are not aligned in a north-south direction. The number 2 area in figure 9 is relatively displaced to the west with respect to areas numbered 1 and 3 probably due to the movement of the inferred faults, which displays a northwest-southeast orientation (labeled as "d" and "e" in Figs. 3 and 9). We have not observed a correlation of these inferred northwest-southeast faults with those shown on surface in the geologic map (Fig. 11). Figure 12 shows faults "d" and "e" along the N-S seismic reflection line 5335, here is inferred that they do not reach the surface. The shape of the basin is not symmetric as can be seen in figures 9 and 10. The basement depth curves are closer together on the eastern border of the basin than on its western side (Fig. 9), where the contours lines maintain a similar horizontal distance from each other. This correlates with the information obtained from seismic lines, which clearly show faults in the eastern sector of the study region, close to the Precordillera, in contrast to the absence of faults towards the western boundary. This is in agreement with the presence of the El Tigre Fault System and the geothermal Pismanta center in its northern part of the basin, above the deepest zones of the Iglesia basin ( Fig. 10) according to Gonzalez et al. (2020). These authors also provided the first evidence for an extensional control of the basin infill during latest Oligocene (?)-early Miocene. The analysis of lines 5322, 5323 and 5324 (northern sector of the basin) shows the presence of two faults with depths up to 6 km, which dip approximately 55° westward and another fault located further west with a dip of about 40° eastward and reaching a depth of up to 8 km (Figs. 6, 7 and 8). These structures seem to continue in a horizontal distance of about 40 km and in a north-south direction. We note a correlation with the same exposed structures denoted as "a", "b" and "c" in the geological map of figure 3 as reported by Cardó and Díaz (1999) and Cardó et al. (2005). and 6 for details); this information is consistent with a transpressive system involving faults (red lines) elongated in a north-south direction in the eastern part and northwest-trending in the northern part of the study area. The easternmost dark blue structure indicates the approximate location of El Tigre Fault System (Cardó et al., 2005).  Figure 10 shows a geological structure that was identified in the seismic lines and geological mapping, which would correspond to El Tigre Fault System (in dark blue color). These structures may be part of a transpressive fault system (Siame et al., 1996(Siame et al., , 1997Fazzito, 2011). Rivas et al. (2019) determined local seismic activity between 10 and 30 km depths and analyzed stress field over 33 focal mechanism solutions for the earthquakes (see inset in Fig. 11). Their results show a minimum vertical compressive stress σ 3 and an almost horizontal position for the maximum compressive stress σ 1 . Some studies have documented this type of motion along the El Tigre Fault System in the Western Precordillera (e.g., Bastías, 1986;Perucca and Vargas, 2014). The same faults identified in figures 6, 7, 9 and 10 are represented in the map view of figure 11 next to the El Tigre System main fault. This system may be connected to the deepest part of the basement near Pismanta favoring the fluids to reach higher depths and thus higher temperatures, as well as a conduit for transportation to the surface using these high angle faults in the upper 10 km. From its deepest part, the basement goes up steeper toward the east (~20°) while its depth smoothly diminishes toward the west (~15°). We note that several faults recognized in this work in the seismic lines 5322, 5323 and 5324 are affecting both the basin fill and the basement (Figs. 5 and 6). Figure 10 also shows the correlation of 5322, 5323 and 5324 seismic lines in a three-dimensional perspective. In this view, we have interpolated the presence of the faults that were identified in each 2D seismic line. We used black color in figure 10 for major faults (with depths up to 8 km) and red color for minor faults, showing a north-south continuity of these structures. At deeper levels, there is no other easily identifiable reflector in the basement. The deepest part of the basin is between line 5323 and line 5327. The Pismanta geothermal center is located in the same area. The next step consisted in an integration of all of the previously analyzed and marked seismic reflection lines to generate a 3D model; thus, the location of the basement was interpolated between each pair of seismic lines. We observe three zones of maximum depth in the 3D view, which are summarized in figures 9 and 10.

Discussion
Based on the analyzed seismic reflection data covering the Iglesia Valley, we obtained a 3D view indicating an asymmetric basin morphology with three dissected maximum depth zones. In addition,  Fig. 7). The geothermal Pismanta area (yellow star) and some localities (black squares) are also indicated for reference. The inset in the left bottom corner shows a schematic interpretation of the transpressive system related to El Tigre Fault System based on the geophysical information analyzed in this and another works (Siame et al., 1996;Pérez and Costa, 2006;Fazzito, 2011); indication of right lateral motion for structures has been suggested by neotectonic studies (e.g., Bastías et al., 1984). This agrees with stresses obtained from focal mechanism solutions of crustal seismicity in this region by Rivas et al. (2019), with maximum SW-NE compressive horizontal σ1, and minimum compressive vertical σ3.axes (right upper corner). we developed a schematic three-dimensional geological diagram located at approximately 30.2° S (Fig. 13). This 3D model integrates surface geological data, regional structural geology from previous studies describing El Tigre Fault system as a flower fault (Álvarez-Marrón et al., 2006) and our interpretation of one representative seismic reflection line (5325, Fig. 5). Reflectors showing the top of the basin basement delineate the basin geometry. The interpretation for basement rock units is based on data from neighboring outcrops in the Iglesia basin. Westward, in the Frontal Cordillera, Carboniferous metasedimentary rocks are cropping out, which are intruded by Permo-Triassic rocks (Furque et al., 2003). Following an association for the stratigraphy found southward, in the Calingasta basin (Barredo and Ramos, 2010), we could also suggest that Triassic sedimentary rocks might be found beneath the Cenozoic unit in the basin. However, much of the structure beneath the Iglesia basin still remains enigmatic due to seismic reflections lines are only capable of imaging up to 5 km depth and not drilling is available. This last unit would consist mostly of clastic Paleozoic rocks highly deformed, forming large-scale east-verging asymmetric folds, and very low grade of metamorphism (Álvarez-Marrón et al., 2006). It is worth to note that this unit crops out eastward being part of the western Precordillera. Depocenters between two mountain ranges is not a common observation worldwide. The Rocky Mountain Trench (USA) can be considered as a depocenter similar to the andean Iglesia basin, although the tectonic activity is different. Thus, topographic depressions may be controlled by extensional tectonics in the Rocky Mountain Trench (e.g., Van der Velden and Cook, 1996) or by compressional tectonic stresses in the Iglesia basin (Rivas et al., 2019). The results observed for the dissected maximum depth areas of the basin, described in the previous item combined with the modeling of the basin and the presence of faults associated with the El Tigre Fault System correlate well with a transpressive system (Siame et al., 1996(Siame et al., , 1997Fazzito, 2011) (Fig. 11). The geothermal analysis indicates that El Tigre Fault System plays a role providing conduits for the outflow of heated meteoric waters near Pismanta, above the deepest part of the Iglesia basin and assuming a normal geothermal gradient. FIG. 13. Schematic structural west-east section at 30.2° S across the Frontal Cordillera and Precordillera. Neotectonic structures at the surface in the Iglesia basin are in dashed lines (see Fig. 3 for references). The lithology and structure beneath the Iglesia basin and Frontal Cordillera are somewhat uncertain due to lack of information Díaz ,1999 andCardó et al., 2005).

Conclusions
Processing of 500 km of seismic lines data covering the Iglesia Valley and available geological information has allowed us to obtain a 3D model for the Iglesia basin, located between the Frontal Cordillera and the Western Precordillera of San Juan, Argentina. This helped us to correlate faults recognized at depth (in the seismic lines) with their geological expression at the surface. In three contiguous seismic lines (5322, 5323 and 5324) located between 30°5' S and 30°15' S there are three faults that affect both the Neogene filling of the basin and its Paleozoic basement. The continuity of these structures in a north-south direction is for about 40 km horizontal distance. These minor faults lie in the northern part of the El Tigre Fault and show changes in their dip.
We obtained an asymmetric shape elongated in north-south direction (70x35 km) for the Iglesia basin with three zones of maximum depth of about 3,650 m below the surface. In all seismic lines, geological structures were delimitated showing the presence of a major north-south structure with a high angle dip of about 85° to the west corresponding to the El Tigre Fault System; it exhibits the minor faults with opposed vergences and NW-SE trending in its northern part. Our interpretation corresponds to the presence of a transpressive system associated with the El Tigre Fault, which is consistent with the seismic analysis of the focal mechanism solutions for the earthquakes and the exposed geological information. The water of the hydrothermal system responsible for the hot springs in the Iglesia basin is probably supplied from the west, by meteoric waters infiltrating at the Frontal Cordillera. Thus, the hydrothermal fluids reach relative high temperatures at depth using the faults like fluids paths. The calculated depth of <2 km for the reservoir of Pismanta geothermal system roughly agrees with the depth of basement top. The 18 O/ 16 O and 2 H/ 1 H ratios indicate a meteoric origin, in good agreement with fluid migration through the faulting.
The basin exhibits three areas of maximum (3,650 m) depths, beneath Pismanta, which are not aligned in the north-south direction. These three deep zones are relatively horizontally displaced. Thus, we also infer the presence of faults of about 35 km length in a relatively east-west orientation that may be controlling the dissection of the maximum depth zones, although non geological evidence has been already documented.