Introduction

Volcanology has often been successful in reconstructing the active history of volcanoes, based on the surface exposures of their eruptive deposits. These have thus become a fundamental tool in understanding the risks associated with volcanic activity (e.g., Carey et al. 1995; Arce et al. 2005; Siebe et al. 2017). However, to understand the magmatic processes occurring below volcanoes, and shaping that activity, petrology and geochemistry have been traditionally used as a “subsurface probe” (e.g., Andrews et al. 2008; Townsend and Huber 2020; Sosa-Ceballos et al. 2014). In addition, though, geophysical methods have contributed to understanding the plumbing architecture below volcanic edifices, i.e., the part of the system that stores and brings magma to the surface (e.g. Yang et al. 2021; Petronis et al. 2021; Sosa-Ceballos et al. 2021). Combining the results provided by all three disciplines (volcanology, petrology, and geophysics) is thus essential for a complete and well-constrained understanding of the subsurface characteristics of the magmatic conduits feeding a volcanic system and how the storage and plumbing system relates to surficial deposits.

Despite recent progress combining petrology and geophysical results (e.g. Martin-Del Pozzo et al. 2002; Power et al. 2021; Wassermann et al. 2022), more volcanic systems must be studied with crossover petrology-volcanology-geophysics interpretations to derive essential information that can be applied to wide-ranging cases. As a contribution to such a direction, we present the results of our geophysical investigation on the Sierra Nevada volcanic system in central Mexico, which is composed of Tlaloc, Telapón, Iztaccíhuatl, and Popocatépetl volcanoes (Fig. 1); the last two being the most active, voluminous, and tallest. In terms of frequency of activity, Popocatépetl is the second most active volcano in Mexico, only behind Colima volcano, but is foremost in terms of risk due to its presence in Mexico's most populated areas, and being some 42 km from Mexico City.

Fig. 1
figure 1

The upper part of the schematic tectonic map of central Mexico shows the Trans-Mexican Volcanic Belt. The yellow box locates the studied area. The red triangles represent important volcanoes and calderas: Ce = Ceboruco; M = Mascota; Te = Tequila; P = La Primavera; Co = Colima; Pa = Parícutin; Ta = Tancítaro; Jo = Jorullo; Nt = Nevado de Toluca; Tl = Tlaloc; Te = Teyotl; Iz = Iztaccíhuatl; Po = Popocatépetl; Ac = Acoculco; Ma = Malinche; Hu = Humeros; Pe = Perote; Pi = Pico de Orizaba. At the bottom, there is a photograph that shows the central and southern part of the Sierra Nevada volcanic range where Popocatépetl and Iztaccíhuatl volcanoes are located. Abbreviations of summit volcanoes and dome of Iztaccíhuatl are H = head; B = breast; K = knees; and F = feet

Full size image

The Sierra Nevada volcanic range has always been a region of geologic interest (e.g., Steele 1971; Robin and Boudal 1987a, b; Nixon, 1989), but the current activity of Popocatépetl, which began in 1994 (e.g., Armienta et al. 1998; López-Loera 2002; Delgado Granados et al. 2008), has spurred the attention of researchers into several aspects of its volcanic activity. The seismicity associated with its volcanic activity has been studied in some detail, primarily because of its well-known precursory significance (National Academies of Sciences 2017; Roman and Cashman (2018); Sparks 2003). For Popocatépetl, the review of seismic data by Kuznetsov and Koulakov (2014) yielded information regarding the structures beneath the volcano, revealing the extent and depth of anomalous seismic velocity anomalies, which can be associated with current activity. In addition to the seismicity approach, geophysical methods, especially self-potential and aeromagnetic methods, are also applied to infer internal structures of volcanic regions (Takeshi et al. 2004; Sawires and Aboud 2019; López-Loera et al. 2020). In the case of the Popocatépetl-Iztaccíhuatl area, aeromagnetic data have been acquired through the years by the SGM (acronym in Spanish of the Mexican Geological Service) and made available to the public (https://www.gob.mx.sgm). Regarding the study of magmatic activity across an entire volcanic complex, little attention has been directed to studying the internal structure of Iztaccíhuatl and Popocatépetl volcanoes.

We here use these data to characterize the magnetic anomalies present in the area, and, through inversion of the data, we identify a number of subsurface bodies causing the anomalies. As a result, we infer the presence of anomalous bodies under the volcanic edifices, as well as surficial domes, and lava flows. Some of the anomalies are present in neighboring areas where only isolated volcanic activity is apparent today. Thus, we present plausible models of bodies that are likely magmatic intrusions associated with the region's past and ongoing volcanic activity. Our study thus contributes to understanding subsurface structrures beneath Popocatépetl and Iztaccíhuatl volcanoes through the analyses of the SGM´s aeromagnetic data.

Geologic and tectonic framework

Popocatépetl and Iztaccíhuatl are the two largest volcanoes of the Sierra Nevada volcanic range. The Sierra Nevada is a 45-km N-S trending long volcanic range in central Mexico that divides the basins within which the cities of Mexico and Puebla have grown (Fig. 2). The range is comprised from north to south by four main volcanoes: Tlaloc, Telapón, Iztaccíhuatl, and Popocatépetl. The rocks of the Sierra Nevada are calc-alkaline andesites-dacites generated by the subduction of the Cocos-Rivera plate beneath the North American plate (e.g., Nixon 1989; Schaaf et al. 1995; Sosa-Ceballos et al. 2015). The Moho and the subducting slab under the Sierra Nevada are about 40- and 70-km deep, respectively (Espíndola et al., 2017). The activity of the volcanoes of the Sierra Nevada varies from north to south, with volcanoes becoming progressively younger towards the south. Tlaloc is the oldest edifice, where activity initiated at 1.8 Ma that peaked at 1.07–0.84 and 0.13–0.02 Ma (Cadoux et al., 2011; Macias et al. 2012, Rueda et al. 2013). Instead Telapon’s activity began, started at 1.4 Ma, with peaks at 0.8 and 0.32 Ma (Cadoux et al., 2011), and Iztaccíhuatl, in turn, began activity at 1.1 Ma with eruptive maxima at 0.6 and 0.45 Ma (Nixon 1989; Macias et al. 2012). Finally, Popocatépetl’s activity started at 0.330 ka (Sosa-Ceballos et al. 2015), with persistent activity during the late Pleistocene (Siebe et al., 1995) and is currently in an ongoing eruptive phase that began in 1996 (Zobin and Martínez, 2010; Gómez-Vazquez et al. 2016; Schiavo et al. 2020).

Fig. 2
figure 2

Simplified geologic map of the Popocatépetl and Iztaccíhuatl volcanoes after (Nixon 1989; García-Tenorio, 2003; 2008; Sosa-Ceballos et al. 2015; Siebe et al. 2017)

Full size image

The aeromagnetic data used here focuses on the area comprising Iztaccíhuatl and Popocatépetl volcanoes. Popocatépetl is the second highest peak in Mexico, lying 65 km SE from Mexico City, the most densely populated area in North America. Its volcanic activity occurred in two eruptive periods, termed El Ventorrillo and Popocatépetl. These periods we characterized by at least three constructive-destructive events (Robin 1984; Robin and Boudal 1987a, b; Siebe et al. 1995, 2017; Sosa-Ceballos et al. 2015). Associated with this, the debris avalanche deposits found to the south of the Sierra Nevada have been attributed to the above activity of Popocatépetl volcano (Robin and Boudal 1987a, b; Siebe et al. 1995; Fig. 2). The age of the first collapse, which is associated with a paleostructure named Nexpayantla (Mooser 1958), is unknown. The second edifice was partially destroyed by a sector collapse at about 22.8 ± 0.9 ka, as inferred from a date obtained from the associated White Pumice Plinian deposit (Siebe et al., 2017). The third destructive event occurred during the Tutti Frutti Plinian Eruption (TFPE), which destroyed the NW portion of the cone (Sosa-Ceballos et al. 2012). This eruption formed the Milky Tephra (MT) and Gray Tephra (GT), which occurred approximately at 14 ka and resulted in the formation of a summit caldera (Sosa-Ceballos et al. 2012). Although the volcanic edifice has been constructed mainly with summit lavas and pyroclastic deposits since the last partial destruction, two flank eruptions have occurred to emplace the Nealtican lava flow (Sosa-Ceballos et al. 2015) and the Ecatzingo-Temocaxusco lavas (León-Miramón, 2013). Both series of lavas occur at the base of the cone and are aligned along the NE-SW Atexca fault (Martin del Pozzo et al. 2017; Arámbula-Mendoza et al. 2010; Ramírez-Uribe et al. 2021) that cuts the main cone (Sosa-Ceballos et al. 2015).

Iztaccíhuatl is a volcanic complex (IVC) formed by superimposed domes (Figs. 1 and 2) (Mooser 1975; Nixon 1982; Mooser et al. 1996; García-Tenorio 2002, 2008). The ages and compositions of the IVC rocks can be subdivided into two groups: an Old Volcanic Series (> 0.6 Ma) and a Young Volcanic Series (< 0.6 Ma) (Nixon and Pearce 1987; Nixon 1988a, b, 1989). In contrast to its effusive activity, the explosive activity of Iztaccíhuatl has been poorly studied, and only recent findings have called attention to some isolated layers, including a debris-avalanche deposit produced around 0.45 Ma BP by the collapse of the southern part of Iztaccíhuatl (García-Tenorio, 2008; Macias et al. 2012). Even since pre-Hispanic times, inhabitants of the region have envisioned the form of a lying woman in the relief of Iztaccíhuatl volcano, thus nicknaming it “the sleeping woman.” The different parts of the imagined female body correspond to eruptive centers of the complex (Figs. 1b and 2).

Materials and methods

Geology

The geologic map of the study area resulted from the integration of data from the geographic spatial database from the National Institute of Statistics, Geography, and Informatics (INEGI 2007) and the Mexico Reception Station of the Spot Satellite Constellation (ERMES). We used digital elevation models (DEMS) with a 10 by 10 m resolution. DEMs were used to prepare shaded-relief models. Panchromatic (1:20,000) orthophotos and the SPOT panchromatic satellite image (with 2.5-m and 10-m pixel resolutions, respectively) were used to build a preliminary map that includes lithological and structural contacts. All geographical information was processed using ArcGIS9.1 and Erdas 9.3 software.

Geophysics

Aeromagnetic data acquisition and basic processing

The aeromagnetic survey was part of a volcano research program supported by CONACYT-SGM to investigate the internal structures of Mexican volcanoes (including Popocatépetl and Iztaccíhuatl), their association with faults, fractures, and alteration zones, and their probable relationship to any ongoing active eruptive phase.

The survey of the area was carried out in collaboration with the SGM (Urrutia-Fucugauchi et al., 2002; López-Loera 2002), using an Islander aircraft BN-B27, equipped with a cesium magnetometer with a sensitivity of 0.001 nT, the Picodas P-data acquisition system 101, and a Sony Pcam-500 video system for positioning. The hourly and daily drift correction was made with a base station using a GSM-19 GEM magnetometer and the altitude control of the flight with a Sperry altimeter radar. The aeromagnetic data were obtained through N-S flight lines separated 1000 m. Control lines were made every 10,000 m in an E-W direction. The flights were made barometrically at an altitude of 5000 m and in the neighboring areas at approximately 500 m, terrain clearance. The data were standardized to an altitude of 5500 m by upward continuation. Other aeromagnetic position/elevation and topographic corrections were performed at the SGM The raw data were also corrected for aircraft motion (magnetic compensation) and daytime drift.

Residual magnetic field

The geomagnetic data encompasses an area of approximately 1125 km2 (40.6 by 27.7 km). The data is provided with aeromagnetic position/elevation and topographic corrections already included. Further processing was carried out using the commercial software Geosoft from Oasis Montaj Data Processing and Analysis System. The Residual Magnetic Field (RMF), shown in Fig. 3a–b was obtained by subtraction of the International Geomagnetic Reference Field (IGRF 1995) from the total magnetic field (TMF) to isolate the field associated with the Earth’s crust. Because the mean values for the magnitude in the aeromagnetic survey data (06/1995) were 42,172 nT, with an inclination of 47°06′ and declination of 6°69′, the RMF anomalies were displaced from the causative sources. We reduced the anomaly to the magnetic pole and, after this transformation, obtained the first vertical derivative, upward continuations, and the analytic signal to facilitate the interpretation. These transformations are discussed below.

Fig. 3
figure 3

a Digital elevation model and residual magnetic field (RMF) of the Popocatépetl and Iztaccíhuatl volcanoes. b Aeromagnetic domains (I to V) described in the contour map of the RMF. Symbols: SWn = South West north; SWs = South West south; N = Nexpayantla; PC = Paso de Cortés anomalies = nw = Northwest; ne = Northeast; sw = Southwest; se = Southeast. Abbreviations of Iztaccíhuatl landforms as in Fig. 1

Full size image

Reduction to the magnetic pole

The interpretation of the aeromagnetic anomalies was aided by transforming the data as if it would have been obtained at the magnetic north pole, where the inclination is 90° and the declination is 0°. This procedure called “reduction to the pole” (RTP), following Baranov and Naudy (1964) is now a standard procedure to treat intermediate latitude data (Figs. 4a, b). After transformation, the magnetic anomalies are simplified, and are due to causative bodies magnetized by a vertical field, similar to those produced by density anomalies in the gravity field. We carried out this transformation of the RMF using the mean values of inclination and declination given above. Once we obtained the RTP, we applied the transformations of Henderson and Zietz (1949) to obtain the first derivative (Fig. 5a), the upward continuation (Henderson 1970), and the analytical signal (Nabighian, 1972; Fig. 5b). The first derivative was applied to recognize the magnetic gradient zones, which correlate with contacts between zones and faults or underground fractures. The upward continuation attenuates the contribution of frequencies associated with magnetic noise from the Earth’s surface. Finally, the analytic signal is helpful in intensifing the edges of the magnetic sources, thus disclosing their dimensions. Based on the generated contour maps of the RMF and RTP, areas with similar magnetic susceptibilities are contoured, yielding the different aeromagnetic domains (AMD) shown in the resultant maps (López-Loera 2002). Each AMD is associated with a different geological unit.

Fig. 4
figure 4

a Contour map of the reduced to the pole magnetic field (RTP), and b aeromagnetic domains (AMD I to VI) defined in the RTP. Symbols: SWn = South West north; SWs = South West south; N = Nexpayantla; PC = Paso de Cortés. Abbreviations of Iztaccíhuatl landforms as in Fig. 1

Full size image
Fig. 5
figure 5

a First derivative of the RTP and upward continued for 500 m. b Analytic signal upward continued 1 km. c Rose diagram of alignments. Symbols: SWn = South West north; SWs = South West south; N = Nexpayantla; PC = Paso de Cortés; T = Teyotl.; abbreviations of Iztaccíhuatl (Iz) landforms as in Fig. 1

Full size image

Magnetic models

AMD-I is the aeromagnetic domain of greatest interest since it encompasses the most significant volcanic structures (Urrutia-Fucugauchi et al. 2002; López-Loera 2002). We obtained 3D susceptibility models with the VOXI technique (Ellis et al. 2012), using possible values of magnetic susceptibility from 0.09 to 0.004 IS (Fig. 6).

Fig. 6
figure 6

VOXI models of the Popocatépetl and Iztaccíhuatl volcanoes located in the aeromagnetic domain (AMD I): k = magnetic susceptibility; IS = international system. a k = 0.09 IS (red color); b k = 0.09 (red) and 0.05 IS (yellow); c k = 0.09 (red), 0.05 (yellow), and 0.01 IS (light blue); d k = 0.09 (red), 0.05 (yellow), 0.01 (light blue), and 0.005 IS (green). The green vector (Y) indicates the position of the North. P = Popocatépetl volcano; PC = Paso de Cortés; T = Teyotl. Abbreviations of Iztaccíhuatl (Iz) landforms as in Fig. 1

Full size image

Results

Geology

We carried out image analysis and fieldwork to define the eruptive products of the Popocatépetl-Iztaccíhuatl volcanoes and the Paso de Cortés-Xalitzintla road (Figs. 2 and 3a–b). We correlated these results with the aeromagnetic results by overlaying geological and aeromagnetic maps. Our map reveals previously mapped lavas and domes and some eruptive products related to the magnetic anomalies below Paso de Cortés (such as the Buenavista lava flow and the Cerro Gordo dome).

Residual magnetic field (RMF)

In the RMF contour map (Fig. 3), Popocatépetl volcano is associated to an average dipolar anomaly with a polar distance of the order of 2150 m with a magnetic anomaly amplitude varying between 352.5 and − 186.3 nT, giving a total intensity amplitude of 538.8 nT. The shape of the magnetic high is ellipsoidal, with dimensions of 4.1 and 3.3 km in the N-S and E-W directions, respectively. The magnetic high extends towards the SW, forming a band ending in two other monopole anomalies. The first anomaly has an elongated shape which is 4.4-km long in the E-W direction and 2.3-km-long N-S, respectively. The second anomaly is located towards the SW portion of the anomalous zone and shows two elongations towards the N-S and the NE-SW with overall dimensions of 3.4 km by 1.8 km. The magnetically anomalous area around Popocatépetl extends for 18.4 km in an NE-SW direction, with a width of 14.5 km. The magnetic low directly associated with Popocatépetl has an elongated shape in the NE-SW direction with dimensions of 3.8 km, and 2.4 km in the NW–SE direction, with a magnetization intensity of -163 nT.

Reduction to the magnetic pole (RTP)

The appearance of the anomalies after RTP transformation can be seen in Fig. 4. Popocatépetl volcano is related to a monopole anomaly extending in a NE-SW direction from the crater. At a distance of 10.8 km from the summit, the anomaly shows a peak seemingly related to a volcanic structure covered by lava from an ancient volcano (SWn). To the south of SWn, another anomaly is present. This is probably the SWs end of the anomalous area of the SWn now covered by volcanic debris deposits. To the NW of the primary anomaly, another anomaly correlates with the geologic structure of Nexpayantla (N, Figs. 3 and 4).

Analysis of lineament trends

Lineament analysis has been traditionally applied to identify lithological structures such as faults, fractures, and dykes through remote sensing (e.g. Hall 1986; López-Loera 2002; Ramachandra et al. 2019); to identify deep-seated lineaments, geomagnetic data have been typically employed (e.g., Feumoe et al. 2012; Lee and Morris 2013; Abdellatif and Elkhateeb 2019). This task is often carried out with the first vertical derivative transformation, which enhances those features. Figure 5a shows the results of the transformation of the RTP data upward continued for 500 m. In the upward continuation algorithm, the data is transformed into the values expected at a plane 500 m higher than the original ones to eliminate some noise in the contours. We used this map to estimate the orientation of the lineaments in the area. As shown in the rose diagram of Fig. 5c, most of them trend 55 ± 5° NE-SW, with a secondary set trending almost E-W (85 ± 5° NW–SE). The nature of these lineaments cannot be inferred from aeromagnetic data alone, but generally, they are due to contacts between intruded and host rocks, faulting (Henkel and Guzmán, 1977; Hall 1986), or other tectonic features.

Aeromagnetic domains

The patterns yielded by the RMF and RTP allow us to recognize five (Fig. 3b) and six (Fig. 4b) aeromagnetic domains (AMD, López-Loera 2002), respectively. The main difference between the two results is that in the RMF, the AMD-I is not well defined to the south, whereas in the RTP, the corresponding domain is closed by another domain. In the RMF chart, the AMD-I comprises three identifiable sectors lying to the south, center, and north of the area covered by this domain. The southern sector contains the anomalies under Popocatépetl volcano, the Nexpayantla structure, and the SW anomalies, “SWn” and “SWs” (Fig. 3, Table 1). The central area contains the anomalies associated with Paso de Cortés and Iztaccíhuatl volcano to the north. The characteristics of the sections of this AMD are listed in Tables 1, 2, and 3.

Table 1 Characteristics or the Southern portion of AMD I
Full size table
Table 2 Characteristics of the center portion of the AMD I
Full size table
Table 3 Characteristics of the Northern portion of AMD I
Full size table

In the RTP pattern (Fig. 4), the anomalies associated with the edifice of Popocatépetl, the Nexpayantla, and the S.W. structures are better defined, and the domain is closed to the south. These anomalies are associated chiefly with volcanic debris deposits. From the RTP anomalies, we derived the Analytic Signal (Nabighian, 1972). As stated before, such transformation allows imaging of geologic boundaries such as contacts and faults. We upward-continued the results to filter out the high frequencies and obtain a better definition of the anomalies. Figure 5a and b displays the result of these transformations as applied to the RTP. In this chart, the anomalies under Popocatépetl and Nexpayantla are isolated by a lineament trending roughly NE-SW.

The central portion of AMD-I is located under the topographic low of Paso de Cortés (PC) that lies between Popocatépetl and Iztaccíhuatl. In the map of Fig. 3, this sector resembles a horseshoe open to the north and formed by a series of minor anomalies. The anomalies have wavelengths (λ) in a range of 2770 m to 3900 m, with magnetization intensity values of − 126 nT to − 8.6 nT. In general, the anomalies in the RMF maps (Fig. 3) comprise an area of 38.65 km2 (7.58 km EW by 5.1 km NS). The four magnetic anomalies that make up the PC area are very distinct. One of them, located to the NW, is ellipsoidal with a long axis E-W, an approximate area of 9.1 km2 (3.5 km E-W by 2.6 km N-S), and altitudes between 3480 and 3640 m.a.s.l. (hereafter, all elevations are given in meters above sea level). This anomaly shows the highest magnetization intensity values ranging from − 8.6 to − 46 nT. The magnetic anomaly located at the SW of the PC (Fig. 3) is elongated in an E-W direction. It covers an area of 6.1 km2 (3.45 km E-W × 1.77 km N-S) and has magnetization intensity values of − 21.2 nT to − 42 nT. The topography of the area has altitudes between 3270 and 3740 m.

There is another small ellipsoidal magnetic anomaly to the SE of Paso de Cortés. Its large axis is oriented NE-SW, and it occupies an area of 2.4 km2 (2.0 km NE-SW by 1.2 km NW–SE) with magnetization intensity values from − 26.5 to − 40.9 nT and topographic altitudes from 3656 to 3825 m. Another magnetic anomaly located to the NE of the PC is oriented NW–SE, spans an area of 1.4 km2 (1.55 NW–SE × 0.9 km NE-SW), shows magnetization intensity values from − 31 to − 38.2 nT, and altitudes from 3631 to 3797 m. In Fig. 4, the PC zone appears as part of the sizeable aeromagnetic anomaly under Popocatépetl, Nexpayantla, and the SW structures, given that its contour line encircles all of these structures. In the map of the analytical signal, the PC structure is characterized by positive anomalies and magnetic lows. The positive anomalies are observed only towards the NW and SW, whereas the lows are seen in the central and SE portions (Fig. 5b).

To the north, the AMD I consists of an anomaly associated primarily with Iztaccíhuatl volcano. As seen in the RMF, this area is a cluster of overlapping anomalies with a shape resembling the number seven with its long axis trending E-W (Fig. 3b). The anomaly shows two magnetic highs that are associated with most of the domes and craters of Iztaccíhuatl. To the south, a steep descent into a magnetic low coincides with the feet of the edifice (Fig. 3). In Fig. 4, the area of Iztaccíhuatl shows a magnetically anomalous zone elongated in the NE-SW direction. This zone contains two anomalies, one to the south, associated with Iztaccíhuatl, and another to the north, related with the Teyotl volcano. It is approximately 9.5-km long in the NE-SW direction and has an average width of 2.8 km. The anomaly associated with Iztaccíhuatl (3.2 by 2.8 km) reaches intensities from 75 to 291 nT. The anomaly under Teyotl is 3.5-km long and 2.1-km wide, with magnetization intensities from 98 to 261 nT.

Figures 4a and 5b, the zone of Iztaccíhuatl, shows an elongated anomaly with a long axis in the NW–SE direction and dimensions of 7.1 km by 2.2 km, with magnetic highs to the N and S separated by a zone of lows trending NE-SW. Two domes are associated with the northern anomaly, and two other points to the southern anomaly. A third dome approximately in the middle of the complex is located at the boundary between the two anomalies, and it is associated with a lineament of magnetic lows. The altitudes related to the domes and craters of Iztaccíhuatl vary from 4419 to 5212 m. To the NE of the Iztaccíhuatl anomaly, another 3 by 1.5 km anomaly is oriented NE-SW and is associated with Teyotl volcano, which has a maximum altitude of 4651 m. However, the edifice is not located at the peak of the anomaly but is situated to the SW. Such a departure is probably due to the magma following fractures or lines of weakness leading tangentially away from the main magma body in its ascent to the surface. The anomaly reflects the location of the source, that is, the position of its solidified magma chamber.

The AMD II is located towards the eastern portion of the Sierra Nevada, presenting a series of magnetic anomalies elongated in the NW–SE direction (Fig. 3). It shows three types of magnetic signals defined by different wavelengths (λ). The first type is found towards the N and has λ of the order of 3015 m, intensity values between − 59 and 74 nT, and a topographic average of 2600 m. It is separated from the central part of the domain by a fault zone whose presence is revealed by a low magnetic value of − 140 nT (Table 4).

Table 4 Characteristics of the AMD II
Full size table

The center consists of a λ of 1608 m, with values of magnetization intensity from − 72 to 44 nT, and a topographic average of 2670 m. It is limited to the N by a fault zone marked by a magnetic low showing intensity in the order of 101 nT.

The third type of anomaly is located towards the S with λ of the order of 2211 m, intensity values between – 97 and 48 nT, and a topographic average of 2614 m. It is separated from the central portion by a magnetic lineament with an NW–SE orientation.

The north portion of this AMD II is undefined towards the margin and shows a wavelength of the order of 1800 m, intensity values from − 97 to 48 nT, and a mean altitude of 2597 m.

The AMD III is located towards the western portion of the Sierra Nevada and is associated with magnetic lows (Fig. 3, López-Loera 2002). In geomagnetic surveys, high gradients are associated with faults, fractures, contact zones, or combinations of such features. These gradients are observed in the limits of AMD I and III. Domain III shows two patterns, one towards the south, characterized by frequencies with average λ of the order of 2300 m, and second towards the north with λ of the order of 3455 m (Table 5).

Table 5 Characteristics of the AMD III
Full size table

AMD IV lies towards the NW portion of the study area (Fig. 3). It is composed of two anomalies that are elongated in the NE-SW direction with λ of 3769 m. To the north, it has intensities from − 39 to 165 nT, and to the south, from − 88 to 155 nT (Table 6).

Table 6 Characteristics of the AMD IV
Full size table

AMD V is found in the extreme SW portion of the study area, but its western and southern limits are outside of the study area (Fig. 3); its dominant feature is an incomplete two-peaked positive anomaly (Table 7).

Table 7 Characteristics of the AMD V
Full size table

AMD VI is located towards the south and has an average magnetization intensity and wavelengths higher in comparison to other AMD (Table 8). Its southern and eastern limits are outside the study area (Fig. 4).

Table 8 Characteristics of the AMD VI
Full size table

Magnetic models

In the analyses discussed above, we pointed out that the Iztaccíhuatl and Popocatépetl volcanoes are located inside AMD I. Therefore, we chose the RMF of this domain to obtain models of the causative bodies in the area. First, we obtained a model for the whole of AMD I, which includes the anomalies associated with Popocatépetl, Paso de Cortés, and Iztaccíhuatl. We performed the modeling with the VOXI technique (Ellis et al. 2012) using possible magnetic susceptibility values from 0.09 to 0.004 IS.

For a magnetic susceptibility of k = 0.09 IS (Fig. 6a), we obtained three bodies for the region under Popocatépetl. The first is associated with this volcano’s anomaly. It consists of a body composed of a vertical plug reaching the area of the crater with a length of 4450 m and a maximum thickness of 590 m. It extends towards the bottom into a horizontally elongated body extending 2900 m in the NE-SW direction, 1750 m NW–SE, and with an average thickness of 1970 m. The second body, producing the SW anomaly, is a cylindrical body with the long axis in the E-W direction dipping towards the NE It is 2500-m long with a maximum width of 780 m. The shallower part of the body is located at a depth of 880 m deepening to 4600 m towards the NE (Fig. 6a). It has an approximate length of 3.08 km and a width that varies from 750 m close to the surface to 1480 m at depth. The third body is very shallow and is the source of a third anomaly farther SW; it consists of a minor body, elongated in the NE-SW direction with a length of 2.95 km and an average width of 660 m (Fig. 6). Finally, inversion of the anomaly on Iztaccíhuatl volcano (Fig. 6a) yields an elongated body with a long axis–oriented NE-SW. It has a length of 3.2 km and a thickness of 1.06 km. Its top and base are 2.69-km and 5.96-km deep, respectively.

With a magnetic susceptibility of k = 0.05 IS (Fig. 6b), the inversion of the anomalies to the south of AMD I generated the same causative bodies associated with the Popocatépetl and the SW anomalies. In the area of Iztaccíhuatl (Figs. 6, 7, 9, and 10), the model consists of a U-shaped body elongated in the NE-SW direction. One tip of the U is located below the structure corresponding to the northwesterly dome of Iztaccihuatl; the other end is located at the anomaly associated with Teyotl volcano (Fig. 9). The Paso de Cortés anomaly is related to a series of small, elongated, sub-vertical bodies.

Fig. 7
figure 7

Image displaying the theoretical (predict) and measure aeromagnetic data atop, the digital elevation model (DEM) below, and the VOXI models of the Popocatépetl (P), and Iztaccíhuatl (Iz) volcanoes underneath. PC = Paso de Cortés anomaly. k = 0.09 (red), 0.05 (yellow), 0.01 (light blue), 0.005 (green), and 0.004 IS (brown). The green vector (Y) indicates the position of North

Full size image

For k = 0.01 IS (Fig. 6c, d), the anomalies to the south and center of the domain are related to a single body presenting a series of apophyses associated with individual anomalies such as that at Nexpayantla. The body associated with the anomalies to the north of domain I, increases in size with depth and towards the south. For k = 0.005 IS (Fig. 6d), all the bodies are connected, forming a single entity with a series of apophyses projecting upward towards the surface. Finally, all bodies are inter-connected for k = 0.004 IS (Fig. 7).

For comparison, Fig. 7 shows the VOXI models obtained for the different susceptibility values. Here, bodies with the highest magnetic susceptibility (k = 0.09 IS, red color) are those that contribute the most to the generation of magnetic anomalies. The models also show bodies that enclose all of the main sources of magnetism (k = 0.004 IS brown), and bodies that encase the main sources individually (k = 0.005 and 0.01 IS, colors green and light blue, respectively).

Shape, size, depth, and volume of source bodies

Finally we estimated the physical characteristics of the bodies associated with the aeromagnetic anomalies through the VOXI models discussed above. In the study area, we associate the main anomalies identified with the volcanic structures of Popocatépetl, Paso de Cortés, Iztaccíhuatl, and Teyotl volcanoes. Beneath Popocatépetl, we find two source bodies (Fig. 6a); the first is cylindrical, with its top located at a depth of 0.73 km. It has an average diameter of 1.7 km and a length of 5.3 km. It is connected with the NE end of a semi-elliptical body with variable dimensions of 3.7 km N-S, 7.3 km E-W, and has a vertical dimension of 4.8 km. The total volume calculated for these two bodies is of the order of ~ 54.5 km3.

The main body in the Paso de Cortés area has a half-ellipsoidal shape dipping to the SW, with a depth between 140 and 200 m. It is 2.4-km long with a maximum width of ~ 2.3 km and an average of 1 km. Its volume is of the order of 4.7 km3.

Iztaccíhuatl and Teyotl volcanoes share a main body of irregular shape. It is elongated in a NE-SW direction and located at a depth of 4 km. It has its base at ~ 8.1 km, a length of 5.3 km in the EW direction, and 4.4 km in the vertical direction. Two branches stem from its top, associated with Iztaccíhuatl and Teyotl, respectively. The first branch has a conical shape; it is located at a depth of 1030 m, has a mean diameter of 1.1 km (EW), and a vertical length of 2.9 km. The second branch, associated with Teyotl volcano, has a cylindrical shape with its top at a depth of 1.8 km, a diameter of 1.2 km, and a length of 2 km. The volume of the entire Iztaccíhuatl-Teyotl body is ~ 30.5 km3.

Discussion

The aeromagnetic studies illustrate the potential of aerial methods to obtain subsurface data for volcanic areas. The aeromagnetic method reveals the main structures associated with associated volcanoes (e.g., faults and fractures or both), as well as the distribution of their products and buried feeding bodies, through contrasting their magnetic signals with the host rocks (Figs. 3, 4, 5). Aeromagnetic methods also assist in recognition of the extent of the different surficial volcanic products such as lava flows, pyroclastic deposits, and volcanic debris, which also possess contrasting magnetic susceptibilities. Such contrasts generate anomalies with different magnetic intensities and surface morphologies, identifiable in contour maps.

From the 3D VOXI models (Fig. 6) of the aeromagnetic anomalies, we estimated the shape, size, depth, and volumes associated with the bodies below Popocatépetl and Iztaccíhuatl volcanoes. Our results show that Popocatepetl has anomalies resulting from bodies with a much larger volume than those under Iztaccíhuatl (54.46 km3 vs. 30.54 km3). We interpret the differences in shape, size, and depth of these two bodies as being a consequence of a higher magma supply to the Popocatépetl plumbing system, which is prone to ascend through a more fractured system. Perhaps the more recent activity occurring at Popocatépetl during the last thousand years has contributed to these differences.

We assume that the anomalous bodies causing the anomalies are crystalized magmas that did not reach the surface. Therefore, they mark the ascent routes followed by these magmas, and their distribution is due to a variety of factors such as the magma supply, magma viscosity and temperature, toughness of the ambient rocks, local stresses, topography, and/or existing fractures and faults, to mention some of the most important controls (Acocella, 2021). Thus, our results provide a better understanding of the volcanic activity in the area and the underlying plumbing system.

The Popocatépetl plumbing system

Modeling of the anomalies shown in Figs. 3, 6, and 7 are consistent with the differences in the positive magnetic effects attributed to mafic rocks and the weakly magnetic effect of silicic rocks buried at the same depth. Such anomalies show the NE-SW trend to be plumbing system at Popocatépetl and intrusive bodies between Popocatépetl and Iztaccíhuatl.

The highest positive anomalies (k = 0.09 IS) are located directly beneath Popocatépetl’s summit and the SW edge of the Atexca fault (Martin del Pozzo et al. 2017; Arámbula-Mendoza et al. 2010; Ramírez-Uribe et al. 2021); a fault that runs across the summit (SWn and SWs; Figs. 2 and 3). The high susceptibility of these rocks and the depth at which these anomalies (SWn and SWs) are located (~ 2–6 km below the summit) suggest that they consist of mafic intrusive bodies. The anomaly has a “pipe-like” geometry that branches laterally, and, given its high susceptibility, it might stem from a swarm of mafic dikes and sills. These anomalies are consistent with the ascent of mafic magmas through the plumbing system during the last 23 ky of explosive activity at Popocatépetl (Sosa-Ceballos et al. 2014), and during its last reactivation in December 1994 (e.g., Macías et al. 2020). The SWn, SWs, and N anomalies differ considerably in depth and size with the anomalies below the summit. However, the variation between different susceptibilities persists where the high susceptibility anomaly below the summit can be related to the mafic intrusions that triggered the most explosive eruptions during the 1997–2001 crisis (Macías et al., 2020). The apparent size of the SWn, SWs, and N anomalies shows a limited subsurface distribution. We propose that these positive anomalies could be due to crystallized mafic magmas, which are leftovers of the magma pulse that formed the Ecatzingo and Temocaxusco lavas (León-Miramón, 2013). The vents that erupted these lavas (Figs. 2, 3, 4, 5) are located on the northern limit of the SW anomaly; therefore, the magma pulse probably migrated from the SE ascending through the SW-NE Atexca fault (Martin del Pozzo et al. 2017; Arámbula-Mendoza et al. 2010; Ramírez-Uribe et al. 2021) that crosses the Popocatepetl’s summit (Figs. 2, 3, 4, 5). As opposed to the SW sector of the fault, relatively weaker magnetic effects characterize the NE portion. This portion of the fault probably favored the injection of dikes and served as the vent that tapped the Nealtican lavas (Siebe et al., 1995; Panfil et al. 1999; Ramírez-Uribe et al. 2021); therefore, this anomaly may also represent magma trapped along dikes. Given the more mafic composition of the Nealtican lavas than the Ecatzingo-Temocaxusco lavas, the weaker anomaly in this sector might be related to the entrapment of magma at greater depth or lower viscosity mafic magma tapped through dikes.

Rocks with lower magnetic susceptibilities enclose the highly magnetic bodies causing the anomalies under Ecatzingo and the Popocatépetl summit; such anomalies (k = 0.05 IS) might represent intrusions of crystallized magma of more intermediate composition. Lower susceptibility rocks (k = 0.01 and 0.005 IS) are associated with these intrusions, extending into the Paso de Cortés zone, between Iztaccíhuatl and Popocatépetl. These anomalies may be felsic intrusive bodies in sedimentary rocks, where intermediate magmas were injected to create reservoirs below Popocatépetl. Abundant evidence for the existence of these rocks is present in the 14 ka Plinian eruption of Popocatépetl (Sosa-Ceballos et al. 2012). This event emplaced deposits that contained accessory lithics of granodiorite, marble, skarn, and sandstone reamed from the conduit or magma chamber walls. Beneath Paso de Cortés, the scenario is different as only low susceptibility rocks (with k = 0.01 and 0.005 IS) are found as felsic intrusives in sedimentary rocks. This suggests that magma was collected below the Popocatépetl-Paso de Cortés area, magma collected, eruptions did not occur, and full crystallization proceeded. This magma accumulation and crystallization occurred around 0.5 Ma, as dated in zircons from granodiorite xenoliths found in Popocatépetl deposits (Sosa-Ceballos et al. 2015). We propose that a series of frequent mafic injections partially melted the intrusive body to form a crystal mush (e.g., Cashman et al. 2017; Zou and Ma 2020). The intrusive body hosting the crystal mush was reheated, probably multiple times, as shown by its 40Ar/39Ar younger ages of 180 ka (Sosa-Ceballos et al. 2015). Evolved melts were produced in the mush, mixed with mafic injected magma and tapped between Iztaccíhuatl and Popocatépetl. We propose that these magmas also reached the Paso de Cortés zone (Figs. 7 and 8), resulting in the eruption of the Buenavista lava and the Cerro Gordo dome, both being products of the restricted plumbing system associated with Paso de Cortés.

Fig. 8
figure 8

Schematic model of the Iztaccíhuatl-Popocatépetl plumbing system; a magma accumulation formed intrusive bodies below the area that later built the Iztaccíhuatl and Popocatépetl volcanic edifices. The approximate geometry of the intrusives are constrained by the anomalies determined in this study; b a series of mafic injections partially melted the intrusive bodies and formed a crystal mush. Mixing between the mush-partial melts and more mafic injections generated volcanic activity through the two stratovolcanoes and along the Paso de Cortés area (Cerro Gordo and Buenavista flow); c present-day visualization of the plumbing system feeding magma to the Popocatépetl and Iztaccíhuatl reservoirs. Colors represent magnetic susceptibilities shown in Fig. 7. Popocatépetl show more voluminous zones with high magnetic susceptibility suggesting the local system is prone to mafic magma circulation

Full size image

The Buenavista lava is bracketed between the Tutti Frutti pumice (14 ka) and the Ochre pumice (5 ka; Siebe et al. 1996; Arana-Salinas et al. 2010). This young lava is exposed east of Paso de Cortés (Fig. 2) and was dated at ~ 7 ka (Siebe et al., 1995). These authors linked its source to the southernmost activity of the Iztaccíhuatl volcano; however, as shown in Fig. 5, this lava may have been fed from a reservoir disassociated with those of Iztaccíhuatl or Popocatépetl magma chambers. Hence, we suggest that volcanic structures such as Cerro Gordo and the Buenavista lava flow were generated by melts extracted from a crystal mush below the Paso de Cortés zone. The relatively young age of the Buenavista flow suggests that discrete volumes of magma can be generated below Paso de Cortés, but the local plumbing system favors the magma ascent through Popocatépetl (Fig. 8).

The Iztaccíhuatl plumbing system

The central and northern portions of Iztaccíhuatl are best characterized by the models for k = 0.09 IS, and k = 0.05 IS as shown in Fig. 6a–d. The bodies found here are less voluminous and not as deep as at Popocatépetl. This difference between the plumbing systems could be due to the more frequent injection of mafic magmas at Popocatépetl (Schaaf et al. 2005; Sosa-Ceballos et al. 2014). It is clear that the southern edge of Iztaccíhuatl (Los Pies) has a weaker magnetic effect when compared to the central-northern portion of the volcano (Fig. 4b, 7, 9c). The Los Pies sector collapsed 450 ka ago, forming a debris-avalanche deposit (García-Tenorio; 2012; Macias et al. 2012). This debris avalanche was superficial; thus, the lower susceptibility values for this magnetic anomaly could be related to rocks with similar compositions to those buried below the central-northern sector, although emplaced much deeper (Figs. 8 and 9).

Fig. 9
figure 9

VOXI models generated with a k = 0.1 (red), 0.05 (pink), 0.03 (purple), and 0.01 (brown) IS at the North zone of AMD I, associated with Iztaccíhuatl and Teyotl volcanoes; b k = 0.1 (red), 0.05 (pink), 0.03 (purple), 0.005 (light green), and 0.003 IS (dark green). c Map showing the contour map of the RMF of the northern portion of AMD I, composed mainly of the craters and/or domes of Iztaccíhuatl volcano (abbreviation as in Fig. 1). The Y axis (green) points to the north

Full size image

Low-susceptibility rocks (k = 0.01 and 0.005 IS; Fig. 6d) are also present below the whole of the Iztaccíhuatl structure and are not connected to the Paso de Cortés rocks like those of Popocatépetl. This suggests that the intrusive body below Popocatépetl-Paso de Cortés is not present below Iztaccíhuatl (Fig. 8). This could be a crystallized intrusive system, and/or low susceptibility body may consist of felsic igneous rocks intercalated with sedimentary rocks (Fig. 10).

Fig. 10
figure 10

Composite VOXI models of the northern part of AMD I for different values of k. a View from W to E displaying three bodies, b E to W view, c view from NE to SW and composition of four bodies. d View from NW to SE and a composition of three bodies. The Y axis (green) points to the north

Full size image

Conclusions

Crossover interpretations using petrology, volcanology, and geophysics are a powerful tool to reduce the uncertainties produced and open questions that remain when each discipline is applied separately. This investigation used an aeromagnetic survey and published petrology-volcanology data to interpret the plumbing system below an active volcanic area.

The structure of the magma systems below the southern portion of the Sierra Nevada is revealed by the magnetic susceptibility contrast between the Popocatépetl-Iztaccíhuatl volcanic rocks and the country rocks of the local basement. The magmatic bodies below Popocatépetl, Paso de Cortés, and Iztaccíhuatl are reveled through 3D inverse modeling of the aeromagnetic data. The magmatic activity that generated the intrusive bodies that we find and the comagmatic extruded products associated with these occurred during the Pleistocene, and its effects continue today. The volume of the subsurface bodies is significantly larger than the volume of the extruded materials that comprise the volcanic edifices and deposits, per the current understanding and observations that extruded comagmatic products are only a minor fraction of the magma chambers associated with large volcanoes (e.g., Marsh 2015; Mori et al. 1996; Baker and Malone 1991).

Interpretative maps and profiles obtained from the geophysical and geological data yielded information on several aspects of the studied area, such as the presence and distribution of the surface deposits and intrusive bodies below the volcanic edifices of the Iztaccíhuatl and Popocatépetl. EW and SE-oriented anomalies are recognized through the vertical derivatives of the RTP data; we interpret these as faults and fractures beneath the southern portion of the Sierra Nevada. These are associated with regional intrusions and favored magma migration through the shallow plumbing system.

Our aeromagnetic data of the Popocatépetl-Iztaccíhuatl area reveals the links between shallow plumbing systems and volcanic structures across this volcanic region, as well as the relationship with the local basement. Intrusions below Popocatépetl are much more voluminous than those below Iztaccíhuatl. As we see at Popocatépetl, such a situation can occur when magmas of higher mafic composition are partially erupted, but a significant amount of the magma becomes stalled, so as to crystalize at depth (i.e., below the edifice) and/or at very shallow depths (i.e. within the edifice). Our analyses also reveal that stalled magma bodies in the zone under the Paso de Cortés region could then feed eruptions of small volume. We thus interpret the anomalies below Paso de Cortés as an intrusive, mush-like body formed by magmas from the Popocatépetl plumbing system that drifted from the volcano’s principal network of conduits to feed the domes and lavas between the two centers. Our work highlights how the body of volcanic edifice growth can be endogenous and/or cryptic (cf. Francis et al. 1993; Allard 1997).