Journal of South American Earth Sciences Volume 98, March 2020, 102412
https://doi.org/10.1016/j.jsames.2019.102412
1
Eruptive chronology of the Acoculco caldera complex – a resurgent
2
caldera in the eastern Trans-Mexican Volcanic Belt (México)
3
by
4
5
6
7
Denis Ramón Avellán1, José Luis Macías2, Paul W. Layer3, Giovanni Sosa-Ceballos2,
Martha Gabriela Gómez-Vasconcelos4, Guillermo Cisneros-Máximo2, Juan Manuel
Sánchez-Núñez5, Joan Marti6, Felipe García-Tenorio2, Héctor López-Loera7, Antonio Pola8,
and Jeff Benowitz3
8
1CONACYT
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
– Instituto de Geofísica, Universidad Nacional Autónoma de México, Antigua Carretera a
Pátzcuaro 8701, 58190 Morelia, Michoacán
2Instituto de Geofísica, Universidad Nacional Autónoma de México, Antigua Carretera a Pátzcuaro 8701,
58190 Morelia, Michoacán
3College of Natural Science, Mathematics and Geophysical Institute, University of Alaska Fairbanks,
Fairbanks, AK 99775
4CONACYT – Instituto de Investigaciones en Ciencias de la Tierra, Universidad Michoacana de San
Nicolás de Hidalgo, Francisco J. Mujica s/n, Felicitas del Río, 58040, Morelia, Michoacán
5Instituto Politécnico Nacional-CIIEMAD, Miguel Othón de Mendizábal s/n. Col. La Escalera, C.P.
07320 Del. Gustavo A. Madero, Ciudad de México, México
6Instituto de Ciencias de la Tierra Jaume Almera, CSIC, LLuis Sole Sabaris, s/n, 08028 Barcelona, Spain
7División de Geociencias Aplicadas, Instituto Potosino de Investigación Científica y Tecnológica A.C.;
Camino a la Presa San José 2055, Lomas 4a Sección, C-P. 78216, San Luis Potosí S.L.P.
8Escuela Nacional de Estudios Superiores, Universidad Nacional Autónoma de México, Antigua
Carretera a Pátzcuaro 8701, 58190 Morelia, Michoacán
32
Submitted to Journal of South American Earth Sciences
*Corresponding author e-mail: denisavellan@gmail.com
33
34
October 29, 2019
1
36
ABSTRACT
37
The Acoculco caldera complex (ACC) is located in the eastern part of the Trans-Mexican
38
Volcanic Belt in the northern part of the State of Puebla. The complex sits at the intersection of
39
two regional fault systems with NE-SW and NW-SE orientations. Acoculco ACC was built upon
40
atop Cretaceous limestones, the Zacatán basaltic plateau of unknown age, early Miocene domes
41
(~12.7-10.98 Ma), and Pliocene lava domes (~3.9-3 Ma). Detailed field mapping and stratigraphy
42
studies complemented assisted by 40Ar/39Ar and 14C dating allowed to divide the ACC volcanic
43
succession into identify 30 volcanic units of the complex. Based on all these data and previous
44
studies the Acoculco ACC eruptive chronology was grouped in four eruptive phases: syn-caldera,
45
early post-caldera, late post-caldera, and extra-caldera. Inception of the ACC volcanism began
46
around 2.7 Ma with the dispersion of an andesitic ignimbrite followed by the collapse of the
47
magma chamber roof as attested by the presence of a lithic breccia in isolated parts of the caldera
48
rim. The collapse produced a 18×16 km caldera depression which was partly filled by the
49
ignimbrite (total volume of ~127 km3) followed by the establishment of an intracaldera lake of
50
unknown total extension. Early post-caldera collapse activity (2.6-2.1 Ma) was restricted within
51
the caldera producing 27 km3 of lava flows and domes dominantly of basaltic trachyandesite to
52
basaltic composition. Late post-caldera collapse activity (2.0-<0.016 Ma) migrated dominantly to
53
the caldera rim and periphery emplacing 90 km3 of magma as rhyolitic domes, lavas, scoria
54
cones, and two younger ignimbrites. The 1.2 Ma Encimadas ignimbrite (26 km3) was generated
55
vented throughon the eastern margin of the caldera and dispersed to the northeast, and the 0.6-0.8
56
Ma Tecoloquillo ignimbrite and dome (11 km3) was producederupted from at the southwestern
57
margin of the ka calderascoria cones. The most recent eruption of this phase was vented close to
58
the southeastern caldera rim producing the Cuatzitzingo (<16,710 ± 50 years BP) scoria cone.
59
Extra-caldera activity (2.4-0.19 Ma) of the Apan–Tezontepec volcanic field produced scoria
2
60
cones and lava flows of basaltic trachyandesite to basaltic andesite composition that are
61
interbedded with the products of the caldera complex.
62
Aeromagnetic data further constraint the edge of the caldera rim and revealed is consistent with
63
the presence of at least four intrusive bodies at depths of >2 1 km hosted in the Cretaceous
64
limestones. These bodies might represent a series of horizontal mafic intrusions located at
65
different depths (183 ka) that besides heating and mixing with the Acoculco magma reservoirs,
66
provides the energy that maintain active the Acoculco geothermal system.
67
Keywords: Geology; geochronology; geothermal energy; Acoculco caldera; Puebla
68
INTRODUCTION
69
Silicic collapse calderas are volcanic depressions resulting from the subsidence of the
70
magma chamber roof caused by the rapid withdrawal of magma during the course of an explosive
71
eruption (Lipman, 2000; Martí et al., 2008; Geyer and Martí, 2014). The formation of a collapse
72
caldera is still an enigmatic geological phenomenon because of the structural complexity
73
involved in such type of volcanic eruptions. In addition, silicic caldera eruptions represent a high
74
risk due to the large amount of magma and eruption rates involved (Costa and Martí, 2016). The
75
resulting caldera depression may represent the site of important ore deposits and high enthalpy
76
geothermal reservoirs. This makes the study of caldera systems of great interest for modern
77
societies; as they represent a high risk but also a significant source of economic resources. In
78
many active volcanic areas, such as Iceland, New Zealand, Costa Rica, Japan, Indonesia, among
79
several others, an important part of their government energy requirements is covered by the
80
exploitation of geothermal resources associated with collapse caldera systems. The current global
81
economic crisis has increased the interest for exploring and exploiting this type of geothermal
82
resources, so a great number of research programs have been initiated all around the world to
3
83
identify and quantify this potential energy resources. In México, geothermal energy has been
84
used since 1959, when a program led by the Comisión Federal de Electricidad (CFE) started to
85
explore and exploit some of these potential energy resources, most of them located in calderas of
86
the Trans-Mexican Volcanic Belt (TMVB) (Hiriart et al., 2011). A successful exploitation of the
87
geothermal resources requires a detailed exploration of the hydrothermal reservoir, which needs
88
to be conducted using a precise combination of geology, geophysics and modeling. In the
89
particular case of geothermal fields installed in collapse calderas, it is very important to
90
determine the mechanism that formed the caldera and the post-collapse internal structure (Bibby
91
et al., 1995; Di Napoli et al., 2011; Molina et al., 2014; Afanasyev et al., 2015). to better
92
understand fluid paths and the location of reservoirs. It is crucial to determine the exact structural
93
limits of the caldera (e.g., Molina et al., 2014), the stratigraphy and age of the caldera-forming
94
deposits and the distribution of thermal anomalies. On this respect, the Acoculco caldera located
95
in the eastern TMVB (Fig. 1) represents an excellent case scenario to study the internal structure
96
of the edifice, its eruptive chronology, and surface manifestations of geothermal activity with
97
respect to fractures and faults. Acoculco has been considered for years a dry-hot rock reservoir by
98
CFE (Lorenzo-Pulido et al., 2010) and is the site of an on-going European-Mexican effort to
99
develop geothermal energy from non-conventional sources called GeMex Project (Calcagno et
100
al., 2018).
101
In this contribution, we analyzed previous studies combined with new fieldwork to
102
present a simplified volcanological map and a new composite stratigraphic column of Acoculco.
103
The results are assisted by 24 new
104
units that combined with previous information set a refined evolution model of the caldera. This
105
information is crucial to present the chronology of the caldera volcanic complex throught time. In
106
order to gain new insights on the caldera structure we produced airborne aeromagnetic models of
40Ar/39Ar
ages and one
4
14C
radiometric date of the volcanic
107
the area, analyzed the morphostructural distribution of the volcanic units, and their correlation
108
with subsurface units in geothermal wells. With Base on thisthese new information, we propose
109
that the late Pleistocene shallow intrusions are still the heat source of geothermal activity beneath
110
Acoculco.
111
GEOLOGICAL SETTING
112
The Acoculco Caldera Complex (ACC) is located within the TMVB, a calc-alkaline
113
volcanic arc produced by the subduction of the Cocos and Rivera plates beneath the North
114
American plate at the Middle American Trench (Pardo and Suárez, 1995) (Fig. 1). The Acoculco
115
area is located in the straddles the central and eastern sectors of the TMVB (Pasquaré et al.,
116
1991). Acoculco occurs ~140 km northeast of Popocatépetl volcano which defines in this region,
117
the active front of the TMVB (Siebe et al., 1995; Macías et al., 2012) (Fig. 1). Acoculco is
118
located over a 45-50 km-thick continental crust (Urrutia and Flores-Ruíz, 1996), and sits at 400
119
km from trench where the Cocos plate plunges into the mantle (Pérez-Campos et al., 2008). This
120
area is under an NW-SE oriented extensional regime, as deduced from the alignment of volcanic
121
vents, dike orientations, extension fractures and kinematics of faults at the Apan-Acoculco region
122
(García-Palomo et al., 2002; 2018). The Acoculco caldera is bounded to the west by the Apan-
123
Tezontepec volcanic field (ATVF) (~3–0.2 Ma) (García-Palomo et al., 2002; García-Tovar et al.,
124
2015), to the north and to the east by Cretaceous limestones of the Sierra Madre Occidental
125
(Avellán et al., 2018), and to the south by early Miocene volcanics of the TMVB (García-Palomo
126
et al., 2002).
127
Prior to the 80´s the Acoculco geology has not been extensively studied (e.g., Ledezma-
128
Guerrero, 1987; Castro-García and Cordoba, 1994), in fact, CFE produced the first regional map
129
of the area (De la Cruz-Martinez and Castillo-Hernández, 1986) followed by other detailed
130
studies as rock dating (López-Hernández and Castillo-Hernández, 1997; López-Hernández and
5
131
Martínez, 1996). More recentLater geophysical studies attempted to understand the internal
132
structure of the caldera (Alatriste-Vilchis et al., 2005; Campos-Enriquez et al., 2003) and the
133
hydrological features conditions (Huizar-Álvarez et al., 1997) of the Apan region. The first
134
volcanological study of the region was carried out by López-Hernández et al. (2009) who
135
concluded that Acoculco was a 18-km wide caldera active from ~1.7 to ~0.2 Ma. These authors
136
considered that Acoculco was nested within the largest 32-km wide Tulancingo Caldera, which
137
was active between ~3.0 and ~2.7 Ma. Recently, Avellán et al. (2018) presented the first detailed
138
geologic map and stratigraphy of the caldera assisted supported by nine
139
authors concluded that the caldera has a semi-circular shape (18–16 km) and that was active from
140
2.7 to 0.06 Ma, thus, not corresponding with the previous timing of the caldera formation (López-
141
Hernández et al., 2009). This geology of Acoculco was used to present a simplified geologic map
142
of the caldera to describe its geochemical evolution (Sosa-Ceballos et al., 2018) and a
143
preliminary 3D model of the Acoculco subsurface structure (Calcagno et al., 2018).
144
METHODS
145
40Ar/39Ar
dates. These
Twenty-four whole-rock samples, phenocryst-free, were crushed, sieved and washed in
40Ar/39Ar
146
deionized water for
isotopic dating (Table 1). For most samples phenocrysts-free
147
ground mass chips (whole rock) were separated for dating. For sample Ac11, both a plagioclase
148
mineral separate and a whole rock sample were separated and dated, and for Ac69, two different
149
rock types were analyzed. The samplesThe samples were irradiated in position 5C at the
150
McMaster University Nuclear Reactor in Hamilton, Canada, for 0.75 MWh. The standard mineral
151
TCR-2 with an age of 28.619 Ma (Renne et al., 2010) was used to calculate the irradiation
152
parameter, “J”. Samples Ac39 and Ac123, were irradiated with the standard mineral MMHb-1
153
with an age of 523.5. The standards were fused and, with the exception of AC-90. The standards
154
were fused and samples step-heated using a laser dating system consisting of 6W argon-ion laser
6
155
at the Geophysical Institute, University of Alaska Fairbanks, following the technique described in
156
Layer (2000) and Layer et al. (2009). For Sample Ac90, an obsidian, 7 small shards were fused,
157
with 6 of the 7 yielding enough gas to calculate a fusion age. The samples were analyzed in a
158
VG3600 mass spectrometer and the measured Ar isotopes were corrected for system blank, mass
159
discrimination and Ca, K, and Cl interference reactions, according to procedures outlined in
160
McDougall and Harrison (1999) and using the standards reported inof Renne et al. (2010).
161
System blanks were 2x10-16 mol 40Ar and 2x10-18 mol 36Ar, which are 5 to 50 times smaller than
162
fraction volumes. Mass discrimination was monitored by running calibrated air shots.
163
The structural analysis of the caldera system comprised a review of previous and new
164
structural data, after the evaluation of digital elevation models, aerial photographs, topographic
165
maps, fieldwork and a morphological evaluation of the landforms. The fault geometry was
166
characterized with field measurements and geomorphic analyses of the faults using ArcGIS on a
167
15 m-resolution digital elevation model from INEGI (Instituto Nacional de Estadística y
168
Geografía). Grid references on maps are in the WGS 1984 UTM Zone 14N projection. Rozeta 2.0
169
software was used for plotting the fault datas on stereographic projections (lower hemisphere)
170
and to perform trend-frequency analysis with rose diagrams. Stereonet 9 software was used for
171
plotting the fault planes and poles, from where the compression and tension areas were inferred
172
into the synthetic right dihedral diagram from the two pairs of conjugate fault systems, according
173
to their slip models (e.g., De Vicente et al., 1992).
174
To estimate the volume of the volcanic units we used the geological map, the Digital
175
Elevation Model (DEM), Spot-6 satellite images (1.5-m panchromatic and 6-m multispectral),
176
and the shaded relief DEM (15-m resolution) in the ArcMap 9.3 software. The difference
177
between the actual topography and the geomorphologic element of each unit was used to obtain a
7
178
z value, and to create a 3D surface geology on the shaded relief DEM and the volume with the
179
Surface Difference tool in ArcMap 9.3.
180
VOLCANIC STRATIGRAPHY
181
In this study we analyzed the regional (García-Palomo et al., 2002; López-Hérnandez et
182
al., 2009) and local geologic maps of Acoculco the ACC and its surroundings (Avellán et al.,
183
2018; Sosa-Ceballos et al., 2018; Calcagno et al., 2018), and their geochronological data include
184
in them. We compiledto present a new simplified geologic map (Fig. 2) supported by 24 new
185
40Ar/39Ar
186
Acoculco units, two pre-caldera, 30 ACC, and nine extra-caldera units. The ACC units were
187
grouped subdivided in syn-caldera, early post-caldera, and late post-caldera, corresponding to
188
different phases of the caldera formation defined by their distribution, stratigraphic position, age,
189
mineral and chemical composition (Fig. 3). The chemical variations of the ACC rocks have been
190
previously documented and discussed and will not be treated in this contribution (Sosa-Ceballos
191
et al., 2018). Next, we succinctly described all units that span in age from ~2.7 Ma to <0.016 Ma
192
with their location given with respect to the caldera rim (Figs. 2 and 3).
193
Basement
194
Undifferentiated Cretaceous Limestones (Ksl)
and one
14C
dates (Tables 1 and 2). The maps shows includes three basement pre-
195
The oldest rocks in the region are Cretaceous limestones (Ksl) of the Sierra Madre
196
Oriental that are exposed to the northeastern, eastern, and southeastern parts of the mapped area
197
(Fig. 2). Good outcrops appear in the Tenexapa and Ajajalpan canyons, close to the towns of
198
Chignahuapan and Zacatlán. At the Chignahuapan hot springs these rocks occur as light-gray
199
parallel stratified limestones with chert concretions between stratified beds sometimes affected by
200
vertical joints fractures filled with hydrothermal minerals (e.g., calcite). Although these rocks are
201
not exposed inside the ACC, they were cut in the two geothermal CFE wells at depths of ~1,200
8
202
m in EAC-1 (López-Hernández et al., 2009), and 350-450 m in EAC-2 (Viggiano-Guerra et al.,
203
2011), likely offset by a fault (Fig. 2).
204
According to these authors, Ksl was intruded by a light-gray phaneritic granite found at
205
the bottom of EAC-1 drill hole at depths from 1800 to 2000 m. The intrusion produceding a
206
methamorphic aureole of skarns (Viggiano-Guerra et al., 2011). A sample analyzed of this well
207
resulted to be an apliteic dike is made of alkali feldspar, plagioclase, quartz, amphibole, chlorite,
208
and Fe-Ti oxides. An isochron 40Ar/39Ar age of this sample yielded an age of 183 ± 36 ka (Table
209
1, isochron age; Fig. 4A), which may correspond to eithera younger intrusions beneath Acoculco
210
(e.g. dike swarms)or likely within a reset age of an older a regional plutonic body (López-
211
Hernández et al., 2009; Calcagno et al., 2018).,. whose older age could have been reset by these
212
younger intrusions.
213
Zacatlán basaltic plateau (Za)
214
Za is aA dark-gray, up to 200 m thick basaltic lava plateau dubbed Zacatlán (Za)that,
215
discordantly overliesying the Cretaceous limestones near the Chignahuapan and Zacatlán towns
216
(Fig. 2). The lava flow is aphanitic with columnar jointing and spheroidal weathering. The age of
217
this unit is unknown, however, its stratigraphic position indicates that Za is older younger than
218
the Cretaceous limestonesm.iddle Miocene.
219
Peñuela dacitic dome complex (~13 to 10 Ma) (Mv) (Pdd)
220
These rocks are exposed around southern the ACC and belong to the first beginnings of
221
regional volcanism of the eastern part of the TMVB (Fig. 2). The oldest pre-caldera rocks are the
222
Peñuela dacitic (Pdd unit) and Quexnol andesitic dome complexes exposed to the SW and SE
223
parts of the ACCAcoculco, respectively (Figs. 2 and 3). The Peñuela dome was dated with the K-
224
Ar method at 12.7 ± 0.6 Ma (García-Palomo et al., 2002). Another sample of this rock (Pdd) was
225
dated here with the 40Ar/39Ar method at 10.98 ± 0.07 Ma (Table 1, plateau age; Fig. 4B).
9
226
Pre-caldera units (~3.9 – 3.0 Ma) (Pc)
227
The Puente (Pald) and Terrerillos (Tdld) lava domes are exposed to the N-NW and SW of
228
the geologic map, respectively (Figs. 2 and 3). These rocks are light-gray andesitic to dacitic lava
229
domes with greenish-gray enclaves. Pald yielded in this work an isochron integrated age of
230
3,.6208,883 ± 220.71 kaMa (Table 1; Fig. 4C). García-Tovar et al. (2015) reported a K-Ar age of
231
3.0 ± 0.4 Ma for Tdld, which is coherent consistent with its stratigraphic position.
232
Syn-caldera unit (Acoculco andesitic ignimbrite, ~2.7 Ma) (Sc)
233
The Acoculco andesitic ignimbrite unit (Aai) corresponds to a yellow to white massive,
234
andesitic ignimbriteunit. It consists of rounded pumice and angular to sub-angular accidental lava
235
fragments (gray, pink and banded) fragments supported by a matrix of coarse to fine ash (Fig.
236
5A). Tube-banded pumice (greenish-gray to white) with alkali feldspar and amphibole
237
phenocrysts is also present. The Aai crops out in small gullies and in some places of high
238
topographic relief where it is greatly covered by younger deposits (e.g., Hbl, Srl, Atal, Fig. 2). To
239
In the southwestern and northern parts of the caldera, Aai underlies ≤40 m thick lacustrine
240
deposits (Fig. 5B). The contact between Aai and the underlying Pdd and Pald units was not
241
observed, however, accidental fragments of these units occur in the ignimbrite. At site 24, occurs
242
a massive lithic breccia (mlBr) made of heterolithologic lavas and scarce pumice (Figs. 2 and 5C-
243
D). This breccia (5-7 m thick) contains angular accidental lavas fragments set in a coarse to fine
244
ash matrix. Some lLava fragments are aphanitic (light-gray and ochre), and other porphyritic
245
(dark-gray, greenish, and ochre) composed ofwith clinopyroxene and plagioclase phenocrysts.
246
The mlBr matrix contains reddish aphanitic lithics and loose isolated crystals of plagioclase,
247
pyroxene and amphibole, disseminated pumice and silty minerals. Laterally, this mlBr grades to a
248
massive fine ash layer enriched in pumice and crystals of plagioclase and amphibole. In three
249
sites (24, 65, and 114) mlBr is interbedded between flow units of Aai (Figs. 2 and 5D).
10
250
The Aai unit was transected at the EAC-1 exploratory well at depths between 210 and 560
251
m, where it covers the Pald unit (Pre-caldera units) and is overlain by the Pedernal rhyolitic lava
252
unit (Pdl, which is a Llate post-caldera unit) (Fig. 2). Well EAC-1 shows nearly 800 m of
253
volcanic materials resting on atop a skar which wasa interpreted as the metamorphosed
254
calcareous basement (skarn) (López-Hernández et al., 2009; Viggiano-Guerra et al., 2011). These
255
authors interpreted the volcanic column from top to bottom as: Acoculco ignimbrite (0-130 m),
256
Cruz Colorada dacite (130-210 m), Alcholoya ignimbrite (210-580 m), and las Minas rhyodacite
257
(580-790 m). However, based on the new geologic map and revised stratigraphy of Acoculco, we
258
consider that the successions would correspond to our Pdl, Aai and Pald units (Figs. 2 and 3).
259
The Pedernal rhyolitc lava (Pdl) unit crops out around the CFE drill holes as a highly altered
260
vesicular rock with phenocrysts of feldspars, plagioclase, quartz and mafic unrecognizable
261
minerals together withand recrystallized lithic clasts and fully corroded, and highly vesicular lava
262
fragments. We consider that these highly altered lavas were mistaken with pumice blocks and
263
erroneously described as the Acoculco ignimbrite. In addition, we propose that their Alcholoya
264
ignimbrite (~2.7 Ma; López-Hernández et al., 2009) described between 210 and 580 m depth
265
would corresponds to our Acoculco ignimbrite Aai (Figs. 2, 3 and 5A). Unfortunately, Aai was
266
not recognized in the succession of well EAC-2 (Fig. 2). This well only shows 340 m of volcanic
267
infill (Pdl and Pald units) resting over the skarn including 200 m of the porphyritic unit
268
(Viggiano-Guerra et al., 2011) or our Pdl unit. As mentioned before, EAC-1 and EAC-2 wells are
269
500 m apart but their stratigraphy suggests that they are separated by a fault that is not
270
recognizable at surface. By averaging these thicknesses in the EAC-1 exploratory well, gullies
271
and outcrops described at surface (~470 m) we estimate a minimum Aai volume of 127 km3 (Fig.
272
2 and 3).
11
273
The AaiA pumice fragment separated from Aai was dated with the
40Ar/39Ar
method in
274
plagioclase at 2,732 ±185 ka (Avellán et al., 2018). In this work, we obtained two new whole
275
rock 40Ar/39Ar dates of Aai pumice samples that yielded younger ages of 2,041 ± 38, and 2,185 ±
276
65 ka (Table 1, plateau age; Figs. 4D-E). However, Aai underlies several early-post caldera lava
277
flows as Aguila (Atal, 2.44 Ma), and Manzanito (Mtal, 2.2 Ma), and the Sayula dome (Srl, ~2.55
278
Ma) (see figures 2 and 3). Based on these stratigraphic relationships, we concluded that the age of
279
Aai must be older than 2.55 Ma for which the 40Ar/39Ar in plagioclase (2,732 ±185 ka) is the best
280
age approximation of the formation of the caldera collapse.
281
A sequence of ≥ 40 m thick lacustrine sediments (ls) is exposed inside the caldera rim in
282
the south, southwest, and northern parts. It consists of a tilted alternation of white clayed laminae,
283
and dark-gray cm-thick, volcaniclastic beds. These beds are made of rounded lava fragments set
284
in a fine-grained matrix barren of fossils. At sites 69 and 119, ls overlies the Acoculco ignimbrite
285
(Aai) (Figs. 2 and 3).
286
287
Early post-caldera units (~2.6-~2.1 Ma) (Epc)
288
Four of these units are exposed inside the caldera depression as basaltic to basaltic
289
trachyandesite lava flows (Hbl, Atal, Vtal and Mtal) that partially cover Aai (Figs. 2 and 3). They
290
are highly eroded and have asymmetric morphologies similar to flatirons with their apex towards
291
the center of the caldera developing a sub-radial exorheic drainage. These units typically appear
292
as light-gray to dark-gray, blocky lava flows with porphyritic to aphanitic textures. These lavas
293
frequently present greenish, yellowish and reddish hydrothermal alteration zones, and host
294
xenoliths of sub-rounded limestones, sandstones and fine grain granite. Two units were dated
295
with the 40Ar/39Ar method at 2,323 ± 48 ka (Vtal) and 2,199 ± 24 ka (Mtal) (Table 1, plateau age;
296
Figs. 4F-GD). Avellán et al. (2018) obtained a
40Ar-39Ar
12
whole rock age of 2,441 ± 234 ka for
297
Atal. The absolute age of Hbl is unknown; nevertheless, it underlies the Srl dome dated at ~2.55
298
Ma (see below). Other two early-postcaldera units (Srl and Atad) are exposed on the external
299
northwestern and southeastern rim of the caldera, respectively (Fig. 2). The Srl unit (2,553 ± 110
300
ka; Table 1; Fig. 4H) consists of gray to black and brown, banded obsidian lava flows with
301
holohyaline texture and partially devitrified to light-gray spherulites and lithophysae (Fig. 3). The
302
Atad unit (2,179 ± 26 ka; Table 1; Fig. 4I) is a blocky greenish to light-gray, porphyritic lava
303
flow (Table 1). For all the early post-caldera units we estimated a volume of 27 km3 (Fig. 3).
304
Late post-caldera units (~2.0 - ~0.016 Ma) (Lpc)
305
These units are represented by 11 domes (Alrd, Lrd, Amrc, Prld, Crcd, Trcd, Crd, Ard,
306
Crl, Arcd and Mrcd), 5 lava flows (Coal, Tal, Cual, Pdl and Plrld), 2 ignimbrites (Eri, Tr1l), and
307
4 scoria cones (Plc, Tlc, Clc1 and Clc2). The Alrd, Lrd, Amrc, Crcd, Trcd, Crd, Ard, Crl Arcd
308
and Mrcd dome and lava flow units occur on the caldera border and periphery (Fig. 2). These
309
structures have predominant rhyolitic compositions with coulée and asymmetrical morphology
310
delimited by very steep levées. The rocks of these units typically appear as light-gray to pinkish-
311
gray, banded to massive obsidian lavas with mottled structure given by abundant spherulites and
312
lithophysae (Fig. 3). López-Hernández et al. (2009) and García-Tovar et al. (2015) reported K-Ar
313
ages in hornblende and whole-rock of the Lrd (1,700 ± 400 ka), Crd (1,300 ± 600 ka), and Crl
314
(1,274 ± 27 ka). In this work, seven new units were dated with the
315
yielding ages of 1,870 ± 36 ka (Alrd) (Fig. 4J4E), 1,438 ± 24 ka (Amrc) (Fig. 4KG), 1,394 ± 8 ka
316
(Crcd) (Fig. 4L), 1,360 ± 15 ka (Trcd) (Fig. 6A), 1,283 ± 88 ka (Ard) (Fig. 6B4H), 998 ± 36 to
317
1,145 ± 14 ka (Arcd) (Fig. 6C-D4J) and 1,066 ± 42 ka (Mrcd) (Fig. 6E4K), that are in agreement
318
with their stratigraphic position. On the other hand, on the western edge the undifferentiated
319
Maguey unit formed of a sequence of pyroclastic surges and fall with a 40Ar-39Ar age of 1,084 ±
320
22 ka (Table 1; Fig. 6F).
13
40Ar/39Ar
method (Table 1)
321
The Coal, Tal, Cual, Pdl and Prld units are situated inside the caldera partially overlying
322
the syn-caldera and early post-caldera units (Fig. 2). They appear as a stack of bedded lava flows
323
with andesitic (Coal, Tal and Cual) and rhyolitic (Pdl and Prld) compositions. The andesitic lava
324
flows (Coal, Tal and Cual units) are porphyritic, black to dark-gray in color, with reddish to
325
yellowish intense hydrothermal alteration (Fig. 3). The Coal unit was dated by Avellán et al.
326
(2018) with the
327
yielded 1,600 ± 35 ka (Table 1; Fig. 6G). In this study we dated the Tal unit with the same
328
method at 1,708 ± 54 ka (Table 1; Fig. 6H 4F). López-Hernández et al. (2009) obtained a
329
40Ar/39Ar
40Ar-39Ar
method at 2,027 ± 40 ka. A dike 40Ar-39Ar age that cuts this unit
age of 1,600 ± 200 ka for the Cual unit.
330
The rhyolitic lava flows (Pdl and Plrd units) are pinkish-gray to pinkish-white porphyritic
331
rocks. These units are white and highly hydrothermally altered lavas that crop out in the vicinity
332
of Pedernal and Acoculco towns (Fig. 2). The lava flows are corroded by hydrothermal fluids
333
along highly vesicular breccia structures, where phenocrysts and matrix have been replaced by
334
mineral alteration minerals. López-Hernández et al. (2009) reported two K-Ar ages for these
335
lavas of 1,600 ± 100 ka (Pdl), and 1,400 ± 200 ka (Prld).
336
The Encimadas unit (Eri) is a rhyolitic ignimbrite widely exposed on the east-northeast
337
external parts of the ACC (Fig. 2). It has a moderately dissected peneplain that mantles the
338
Zacatlán basaltic plateau and partially covers some of the post-caldera units. Eri is a welded
339
ignimbrite with several flow units that appear as massive, light-gray to white, beds. Each bed
340
consists of matrix-supported fine ash particles with abundant feldspar and quartz phenocrysts. It
341
has an approximated volume of 26 km3 (Fig. 3). López-Hernández et al. (2009) reported a
342
40Ar/39Ar
343
obtained in this study (Table 1, plateau age; Fig. 6I4I).
age in sanidine of Eri at 1,300 ± 200 ka. A similar 40Ar/39Ar age of 1,278 ± 14 ka was
14
344
The Tecoloquillo ignimbrite (Tr1) is a rhyolitic ignimbrite widely exposed to the south-
345
southwest parts of the caldera (Fig. 2). South of the caldera, Tr1 partially covers some units
346
belonging to the pre-caldera, syn-caldera and post-caldera. The Tr1 unit consists of two main
347
beds; the lowermost part is massive, monolithologic, brittle and matrix-supported with highly
348
friable pumice fragments embedded in a medium to fine ash matrix. Both pumice and matrix
349
contain bipyramidal quartz and alkali feldspar phenocrysts. The upper part is massive with pink-
350
gray, corroded lava blocks, supported by a crumbly medium ash matrix. The upper part of Tr2 is
351
a rhyolitic dome made of angular light-pink to gray lava blocks. The rock is moderately vesicular,
352
fibrous and porphyritic with quartz, alkali feldspar and amphibole phenocrysts. One pumice
353
sample collected from the basal part of the Tr1 unit yielded a 40Ar/39Ar age in plagioclase of 611
354
± 72 ka (Table 1; Figs. 6J). Other pumice sample that yielded older age of 762 ± 9 ka (Table 1;
355
Fig. 6K). López-Hernández et al. (2009) reported a 40Ar/39Ar age in sanidine of 0.8 ± 0.1 Ma for
356
this unit.
357
The last four late-post-caldera units are cinder cones (Paila, Tuliman, Cuatzitzinguito and
358
Cuatzitzingo) and associated lava flows of basaltic andesite composition. These scoria cones
359
occur above or close to the topographic caldera rim (Fig. 2). The Paila (Plc) and Tuliman (Tlc)
360
cinder cones are exposed on the southeastern and northwestern parts of the caldera rim,
361
respectively. The Paila unit directly overlies the Atad and Eri post-caldera units and Tuliman unit
362
lies discordantly on top of Srl post-caldera unit. The Cuatzitzinguito (Clc1) and Cuatzitzingo
363
(Clc2) Other scoria cones are lies on the southern flank of the Paila unit (Cuatzitzinguito and
364
Cuatzitzingo units) (Figs. 2 and 3). All these scoria cones are composed of massive poorly-sorted
365
fallout beds with dense blocks and bread-crust scoria and spatter bombs, as well as, black to dark-
366
gray blocky lava flows associated with effusive activity. The Tulimán scoria cone was dated with
367
the 40Ar/39Ar at 63 ± 9 ka (Avellán et al., 2018). A 40Ar-39Ar age of 71 ± 17 ka was obtained here
15
368
for the Paila scoria cone (Table 1; Fig. 6L). A paleosol underneath a fallout tephra of the the
369
Cuatzitzinguito scoria cone deposit yielded a
370
obtain an absolute age for the Cuatzitzingo scoria cone, however, its stratigraphic position
371
indicates that it should be youngerst than 16 ka.
372
14C
age of 16,710 ± 50 BP (Table 2). We did not
We estimated a total volume for all the late-post caldera units including Eri and Trl of 90
373
km3 (Fig. 3).
374
Extra-caldera units (~2.4-0.19 Ma) (ATVF)
375
NineteenTwentyEighteen scoria cones and threefour small-shield volcanoes of the ATVF
376
occur around the caldera complex (Figs. 2 and 3). Four of these scoria cones, known as
377
Amanalco (Asc; ) (2,408 ± 58 ka; Avellan et al., 2018), Huixtepec (##Hsc), Tecolote (Tsc) and
378
Apapasco (Asc) are located ca. 7 km to the southeast of the caldera border. Five scoria cones
379
called (Buenavista (##Bsc), Comal (##Csc), Calandria (##Csc), Toronjil (##Tsc), and Tezontle
380
(Tsc##)) and the Coatzetzengo small- shield-volcano are located at ca. 4 km to the northwest.
381
Three scoria cones named Moxhuite (Msc; ) (239 ± 34 ka; Avellán et al., 2018), Matlahuacala
382
(Mlcsc), and Cazares (Csc) lie discordantly on top of the Encimadas ignimbrite at ca. 6 km to the
383
east. Two small- shield volcanoes, Camelia (Callc; ) (2,033 ± 84 ka; Avellán et al., 2018) and
384
Tetelas (Tlcal,) (1,060 ± 84 ka; Avellán et al., 2018) are located at 10 km to the south of the
385
caldera border. Four Three scoria cones are aligned in a NW-SEW direction, these are Cuate
386
(##), Tecajete (##Tsc) (1,235 ± 62 ka), Blanco (Bsc; ) (1,274 ±62 ka; Avellán et al., 2018) and
387
Hermosa (Hsc##). Another, fourthree scoria cones, Coliuca (Clc; ) (188 ± 6 ka K-Ar age; García-
388
Tovar et al., 2015), Coloradoyote (##Csc), El Conejo (Csc) and Tezoyo (Tsc##), and the Coyote
389
small-shield volcano are situated at ca. 7 km to the southwest of the caldera border.
390
CHARACTERIZATION OF FAULT SYSTEMS
16
391
The ACC is in a highly tectonized area region affected by NW-SE, NE-SW and E-W
392
structures (Fig. 6A7A). The NW-SE and NE-SW structures belong to two regional fault systems.:
393
the NW-striking Taxco-San Miguel de Allende and the NE-striking Tenochtitlan-Apan. The NW-
394
striking (NNW to NW) Taxco-San Miguel de Allende fFault sSystem contains the oldest regional
395
structures. Some of tThese faults werefollow the trend of inherited from the older structures, such
396
as fold axes and Cretaceous folds and thrust faults from the Laramide orogeny in the Sierra
397
Madre Oriental (Suter, 198491; Rocha et al., 2006; Lermo et al., 2009), locatexposed just north
398
of the study area. The NW-SE system began its activity ~30 Ma with the NE-oriented Cenozoic
399
extension (Henry and Aranda-Gómez, 1992) was active in this region from middle to late
400
Miocene, creating NW-striking normal faults and NE-striking strike-slip faults synchronous with
401
regional volcanism that represent the southernmost part of the Basin and Range province in
402
México (Aranda-Gómez and McDowell, 1998; Aguirre-Díaz et al., 2005)(Andreani et al., 2008),.
403
In the ACC, the NW-SE fault system is represented by several major normal and oblique (right-
404
lateral component) faults and numerous minor fault strands with lengths between 2 and 5 km,
405
with an average azimuthstrike of 130˚ dipping mainlyboth to the NE and SW at average angles of
406
50˚, creating horst and graben-like structures (García-Palomo et al., 2002; ) (Fig. 67A). The
407
Manzanito structure, also known as the Tulancingo-Tlaxco fault system (López-Hernández et al.,
408
2009), is the most representative fault pertaining to this system in the ACC (Avellán et al., 2018).
409
It has an en échelon array of normal faultsstructure, which extends for ca. 30 km that measures
410
ca. 30 km (Fig. 6B) and displacesing the 1.7 Ma Lobera rhyolithic dome by at least 145 m, and
411
the 1.07 Ma Minilla rhyolitic dome by 120 m (this study; Fig. 6B). Also, atThe most recent
412
activity of this system in the ACC occurred around 1.6 Ma, when the ~160˚-striking and ~2.5 km
413
long Colorada basaltic dikes were emplaced at the southeastern part of the caldera (Fig. 2; Coal;
414
Avellán et al., 2018). These dikes, together with parallel faults indicate a NE-oriented Pleistocene
17
415
tectonic extension that was synchronous with volcanism. But there is no evidence of recent
416
(Holocene) activity in the local NW-striking faults.
417
The NE-striking (NNE to NE) Apan-Tlaloc Fault System (García-Palomo et al., 2018),
418
Tenochtitlan-Apan Fault System (García-Palomo et al., 2002) or Tenochtitlan Shear Zone (De
419
Cserna et al., 1988), has been active since the Miocene and is still active (García-Palomo et al.,
420
2002; 2018; Fig. 76). This is the most important fault system in the region, consisting of normal
421
faults with a left-lateral component with an average azimuthstrike of 040˚ dipping both to the
422
NW and SE with an average dip angle of 75˚ that present two generations of striae (~30 and 80°).
423
This system has created regional horst and graben-like structures, obeying a NW-oriented
424
extension regime that affects eastern México (this study and García-Palomo et al., 2018; ) (Fig.
425
67A). In general, this faultit system shows a good geomorphic expression represented by several
426
major faults (e.g. Apan-Tlaloc Fault and Chignahuapan Fault) and numerous minor fault strands
427
and fractures with 1-4 km individual lengths. Field exposures show prominent scarps; the ~1.34
428
Ma Mesa ChicaCanoas rhyolitic dome is displaced by the Atotonilco faultscarp (northern caldera
429
rim) by 150 m, and the ~2.2 Ma Ajolotla trachyandesitic dome is displaced by the Chignahuapan
430
fault by 200 m (this study). The most recent volcanic structures of the ACC (Plc with 71 ka, and
431
Tlc with 63 ka units) are cut by these faults (this study; Figs. 2 and 67B).
432
The NE- and NW-striking normal fault systems intersect each other (García-Palomo et al.,
433
2002; Lermo et al., 2009) creating an orthogonal arrangement of grabens, half-grabens and
434
horsts. The NE-SW Rosario-Acoculco Horst (García-Palomo et al., 2002) is delimited to the west
435
by the ~235°-trendingstriking and NW-dipping Apan-Tlaloc Fault (Mooser and Ramírez, 1987;
436
Huizar-Álvarez et al., 1997) and to the east by the ~055°-trendingstriking and SE-dipping
437
Chignahuapan Fault (Avellán et al., 2018defined in this study). The ACCcoculco Graben is a
438
volcano-tectonic grabendepression inside the Rosario-Acoculco Horst, delimited by parallel
18
439
faults but with opposite dipping. As for the E-W structures, they are represented by at least 10 E-
440
W striking (~N085°) and SE-dipping normal fault strands. They can only be observed inside the
441
ACC, mainly affecting post-caldera volcanism in the central part of the ACC (this study; Figs. 2
442
and 76).
443
AEROMAGNETIC DATA
444
We produced an aeromagnetic map by reprocessing the airborne data of the Mexican
445
Geological Survey obtained in 2000 (Fig. 78A). This map shows different aeromagnetic
446
anomalies inside or outside of the ACC. The outer domains may be caused by regional and
447
topographic anomalies. Instead, inner domains are possibly associated to four positive local
448
anomalies (subdomains) and conforming a semi-circular shape of the Acoculco caldera. The first
449
subdomain is located in the central part of the ACC; it shows an NE-oriented elongated shape
450
(9.7 km long and 4.8 km wide) with magnetic intensities between -81.8 and 57.8 nT. The second
451
anomaly is located at the southeastern boundary of the ACC, ENE-oriented, 6.6 km long and 4.1
452
km wide with magnetic intensities between -135.1 and 25.7 nT. The third anomaly is located in
453
the south-central portion of the ACC, NE-oriented, 8.9 km long and 3.5 km wide with magnetic
454
intensities between -64.5 and 36.5 nT. The fourth anomaly appears in the western part of the
455
ACC, NW-oriented, 5.4-km long and 3-km wide with magnetic intensities between 84.5, and -2
456
nT. The map shows a contrast between the caldera (higher magnetic values) and the surrounding
457
areas (lower magnetic values). Different aeromagnetic domains or anomalies described in this
458
study are located either inside or outside of the ACC.
459
DISCUSSION
460
Tectonic implications
19
461
The regional tectonic setting has controlled the formation and evolution of the ACC. The
462
ACC formed on top of the Acoculco-El Rosario-Acoculco horst, which is bounded by the Apan-
463
Tlaloc and Tlaxco-Chignahuapan grabens (Fig. 67). Furthermore, some pre-existing tectonic
464
faults were used as pathways for dike intrusions and volcanism (aligned scoria cones parallel to
465
the NE-trending fault system), but also as part of the ring fault system that controlled the caldera
466
collapse and exerted the main control on the location of post-caldera vents (Fig. 2). The
467
southwestern caldera rim coincides with the NW-SE Manzanito fault, while the northern caldera
468
rim is marked by the Atotonilco scarp (Avellán et al., 2018; ) (Figs. 2 and 67). The sub-linear to
469
sub-circular collapsed structures form a 18 x 16 km rhombohedral-shape, steeply dipping towards
470
the central part of the ACC (Figs. 2 and 67). In fact, the left-lateral movement in the NE-SW
471
faults during the middle Miocene (García-Palomo et al., 2000; 2018) and/or the right-lateral
472
movement of the NW-SE faults could have been responsible for the fracturing and creation of the
473
space necessary to accommodate the magma reservoir beneath the caldera (pull-apart basins in
474
transtensional regimes; e.g., Bursik, 2009; Saxby et al., 2016). Activity of the NE-SW and NW-
475
SE regional systems continued after the caldera formation and modified the trace of the caldera
476
border, causing its displacement at several points (e.g., northeastern tip of the topographic caldera
477
rim, outcrops 27 and 28; Fig. 67B). Tectonic movements affected the interior of the caldera until
478
very recent time, causing displacements of intra-caldera blocks and disturbing the position of
479
intra-caldera volcanic vents and products. For example, E-W faulting in the central part of the
480
caldera (outcrops 24, 32 and 54; Fig. 67B)
481
Caldera reactivation during the emission of the 26 km3 Encimadas (~1.28 Ma) and 11 km3
482
of the Tecoloquillo ignimbrites (~0.6-0.76 Ma) was possibly originated near or along the E and
483
SW borders of the caldera, respectively (Figs. 2 and 3). Therefore, it would not be surprising if
484
the caldera used these faults systems to nucleate the rest of the ring fault that controlled the
20
485
collapse event, as it has occurred in other well documented calderas (see Aguirre-Díiaz et al.,
486
2008; Martí et al., 2013; Molina et al., 2014). Despite the strong, selective erosion that has
487
affected parts of the area, the northern topographic caldera rim has not retreated significantly
488
from its original position. This may imply that caldera subsidence continued for a long time
489
(±500 ka, between Encimadas and Tecoloquillo eruptive events), dissecting younger rocks
490
emplaced close or through the ring faults. On the contrary, in the eastern side of the ACC, neither
491
the morphological nor the structural border are visible at surface, however, it is depicted in the
492
aeromagnetic map (Fig. 78). At the surface, the eastern border is not visible because it has been
493
coverburied by younger deposits and it is also likely an uneven collapse of the caldera. The E-W
494
intra-calderastructures fault system areis limited to the caldera interior and therefore isthey could
495
represent local deformation related to the resurgence of the caldera.
496
The NE-SW and NW-SE fault systems coexist in the ACC (Fig. 67B). This can be
497
explained by two different constructive phases. The first one took place in the Oligocene-
498
Miocene, ruled by ~NE-oriented extension, forming ~NW-SE dip-slip faults and ~NE-SW
499
sinistral strike-slip faults (left-lateral movement) ((Henry and Aranda-Gómez, 1992). The second
500
phase took place in the Pliocene-Pleistocene, and it is controlledruled by ~NW-oriented tectonic
501
extension, forming ~NW-SE dextral strike-slip faults (oblique and strike-slip faults with right-
502
lateral component) and ~NE-SW dip-slip faults (oblique and normal faults with a minor left-
503
lateral component; ) (García-Palomo et al., 2018). This is consistent with the left-stepping en
504
échelon geometry of the NW-SE Manzanito fault originated by right-lateral slip. Actually, this
505
latter phase of NW-oriented extension is still active (García-Palomo et al., 2018). Therefore, both
506
fault systems occur in the same region and are still active under the same stress regime. This is
507
possible because the NW-SE structures are acting as transfer faults of the NE-SW normal faults.
508
In order to demonstrate that both fault systems are moving under the same strain field, a synthetic
21
509
right dihedra diagram was created for the two pairs of conjugate fault sets, according to the mean
510
measured planes (Fig. 67B). This slip model implies an orthorhombic symmetry of faults, where
511
superposed areas divide the figure into the different compressional and tensional areas in the
512
ACC.
513
Aeromagnetic interpretation and geothermal implications
514
Previous studies used the aeromagnetic and gravimetric data of the Acoculco-Atotonilco
515
region obtained by Petroleos Mexicanos (PEMEX) in 1980 (García-Estrada, 2000; López-
516
Hernández et al., 2009). These authors interpreted a depression bounded by faultsstructure
517
coincident with gravimeetric data and low magnetic values with much higher values at the
518
interior defining the border of the Acoculco caldera. This coincides with our interpretation of the
519
airborne data of the Mexican Geological Survey obtained in 2000 displayed in the aeromagnetic
520
map of figure 78A. This map clearly shows an approximated match between the edge of the 18 x
521
16 km Acoculco caldera as reconstructed from the distribution of the Acoculco ignimbrite, the
522
co-ignimbritic breccia, and the Atotonilco scarp, and the Manzanito fault (Avellán et al, 2018;
523
this work) (Fig. 67). The aeromagnetic anomalies suggest a rough elliptical half-rounded limit of
524
the caldera that may be bounded to the east by the approximated venting location of the
525
Encimadas ignimbrite and to the southeast by the basaltic andesite la Paila scoria cone (Fig.
526
78A). The positive aeromagnetic anomaly located at the NE portion of the ACC is may be
527
produced by intrusive bodies underneath the Cretaceous limestones.the Zacatlán basaltic plateau
528
that Tthe Zacatlán basaltic plateau with high remanence and susceptibility. This basaltic plateau
529
is overlaid by the Encimadas ignimbrite and, both covering the Cretaceous limestones (Fig. 2).
530
The positive anomalies located to the SW and NW of the ACC may be are associated with
531
intrusive bodies linked to the ATVFthat gave place to monogenetic volcanoes of the ATVF
532
thandesitic units of Apan, the andesite-dacite units of Peñuela, and the late Miocene-late
22
533
Pleistocene basaltic andesitic units of ATVF (Figs. 2 and 78A). The distribution of these positive
534
anomalies are similar to the topographic features. The interior of the this tectonic depression the
535
caldera is not homogeneous but includes higher and lower magnetic values that might reveal the
536
presence of several small-scaleer horsts and grabens combined with the emplacement presence of
537
shallow intrusive bodies (i.e., laccoliths, sills) of basaltic and basaltic and intermediate andesitic
538
(i.e. sills) composition intruded inside the Cretaceous limestoneshosted inside the Cretaceous
539
limestones at depths (>1 km; (Fig. 78B).. These positive anomalies inside the caldera could have
540
fed the ca. 2.6 Ma Huistongo, 2.2 Ma Manzanito, 0.71 Ma Paila and <16 ka Cuatzitzingo
541
eruptions (Figs. 2, 3 and 7). at depth. rocks during the evolution of the caldera inside the
542
depression. The anomalies described in figure 78A have rough NE-SW orientations separating a
543
local horst on top of which the CFE exploratory wells were drilled, cutting skarns at ~790 m deep
544
(EAC-1), and argillitic limestones at 350 m deep (EAC-2) (Viggiano-Guerra et al., 2011). These
545
positive magnetic anomalies might represent intrusive bodies injected during the post-caldera
546
phase (Huistongo lavas 2.6 Ma), which may have an important role to keep active the geothermal
547
system active. Sosa-Ceballos et al. (2018) concluded that after the caldera collapse some 2.7 Ma
548
ago, the stress field on the ACC magma chamber and its immediate surroundings was modified.
549
Such change hindered magma migration through the collapsed reservoir and promoted lateral
550
tension zones where dikes and sills converge and eventually form new magma chambers. Due to
551
the heavily active tectonics deformed upper-crust in the Acoculco area, a new plumbing system
552
developed after the caldera collapse and gradually favored the ascent of deep seated,
553
independentphysically isolated peralkaline and calcalkaline mafic magmas. In addition, Sosa-
554
Ceballos et al. (2018) found that magma mixing-heating is the main magmatic process that
555
modified the ACC rock suite. Hence, the aeromagnetic anomalies might represent the horizontal
556
intrusion zones where mafic magma accumulates as a consequence of the upper crustal
23
557
deformation produced by the collapse (Fig. 2). These intrusion zones serve as heating elements of
558
magma reservoirs (that eventually might evolve into small magma chambers) and yield the
559
energy that sustain the Acoculco geothermal system (Fig. 2). When magma did not migrate
560
horizontally, it flowed upwards, sometimes ggetting ot tapped on its way to the surface evolving
561
to silicic compositions and in other cases got trapped as the aplitic dike dated at 183 ± 36 ka
562
found in EAC-1 (Sosa-Ceballos et al., 2018) (Table 1; Fig. 2). These young felsic intrusions have
563
reactivated the system providing heat for the hydrothermal activity and Holocene hydrothermal
564
explosions (Canet et al., 2015a and b). López-Hernández et al. (2009) also suggested the possible
565
presence of an intrusive body, whose top is at > 1000 m depth and that was not reached by EAC-
566
1 well. Although, we believe these intrusions also contribute to the geothermal system, their
567
volume and abundance are not yet known and probably can be resolved estimated by geophysical
568
methods to constraint their energy input.
569
A 3D aeromagnetic model of the area shows four anomalies (shallow intrusive bodies
570
likely mafic sills) intruded inside the Cretaceous limestones at depths >1 km (Fig. 7B). The CFE
571
exploratory wells were drilled between two of these anomalies where the wells cutting limestones
572
and skarns at different depths. These wells were drilled at a local structural high separated by two
573
depressions with NE-SW orientations (Fig. 6B). These depressions may be occupied by represent
574
intrusive bodies (laccoliths, sills) injected of basic and intermediate composition rocks
575
corresponding toduring the syn-caldera and early post-caldera rocks phase, in comparison with
576
the presence of more silicic rocks and sediments outside the caldera. The aplitic dike dated at 183
577
± 36 ka supports the presence of young mafic intrusions that have reactivated the system
578
providing the heat for hydrothermal activity and Holocene hydrothermal explosions (Canet et al.,
579
2015b). López-Hernández et al. (2009) also suggested the possible presence of an intrusive body,
24
580
whose top is at > 1000 m depth and that was not reached by EAC-1 well. The distribution of
581
magnetic anomalies inside the caldera also reveals the presence of tectonic lineaments with NE-
582
SW and NW-SE orientations, likely associated to horst and grabens blocks and coinciding with
583
those intra-caldera and caldera ring faults observed and measured in the field, suggesting that
584
such faults were used as vent zones for post-caldera volcanism (Figs. 2, 76 and 87 ). These faults
585
may have also favored the ascent of gas with isotopic compositions (N2/He, 3He/4He, 13C, 15N) of
586
both MORB- and arc-type signatures (Peiffer et al., 2014), and high values of CO2 and 3He/4He
587
ratio (R/Rair = 8.5), suggesting that suggest an active magmatic source at depth under the ACC
588
(Polak et al., 1982).
589
Volcanic evolution
590
Based in our synthetic map, updated refined volcanic stratigraphy, structural analysis, and
591
interpretation of the aeromagnetic anomalies interpretation is clear that ACC was emplaced
592
through the Cretaceous limestones, the Peñuela dacitic domes (~13 to 11 Ma), and the Zacatlán
593
basaltic plateau (Figs. 2 and 3). Rocks of Tthe ACC consists were divided intoof 30 units
594
corresponding formed into different stages of the caldera evolution from pre, syn, early-post, late-
595
post caldera and extra-caldera volcanism of the Apan-Tezontepec Volcanic Field (Avellán et al.,
596
2018; Sosa-Ceballos et al., 2018).
597
Caldera formation (~2.7 Ma)
598
Prior to the formation of the caldera, the Puente and Terrerillos lava domes had been
599
extruded in the area (Fig. 89A). These domes of andesitic and dacitic composition and calc-
600
alkaline affinity were emplaced during the Pliocene between 3.9 and 3 Ma. After a few hundred
601
thousand years (0.3 Ma), a large amount of andesitic magma stagnated at depth preparing for a
602
major eruption. This calc-alkaline magma overpressured the encasing rock initiating the emission
603
of an andesitic pyroclastic density current (Acoculco ignimbrite). The continuous emission of the
25
604
ignimbrite eventually diminished the magmatic pressure and weakened the roof of the magma
605
chamber triggering its collapse (Fig. 89B). The collapse followed older structures as the NW-SE
606
Manzanito fault in the ignimbrite was dispersed from different sectors of the caldera structural
607
rim, along the Manzanito structure at the western and- southwestern parts and generated , the
608
Atotonilco structure at the northwest, and the new-formeda semi-circular caldera ring (Atotonilco
609
scarp) sector of the topographic caldera rim at the nNorth (Figs. 2 and 67B). The ignimbrite only
610
crops out in at the western- and southwestern and northern parts of the caldera interior because in
611
other locations it has beenwas covered by younger deposits. The collapse used pre-existing fault
612
systems that control the formed a 18 x 16 km asymmetric depression calderaof the ring fault
613
(Figs. 2 and 76). The used of older pre-existing faults (i.e., a long dike conduit) and fractures as
614
magma feeders facilitated a high discharge rate at the beginning of the eruption as discussed by
615
Costa and Martí (2016). This developed immediately into massive proportions, thus precluding
616
the formation of a vertical eruption column and the deposition of fallout deposits, which have not
617
been recognized neither outside the caldera nor in the intra-caldera wells. The occurrence of a co-
618
ignimbrite breccia found in relative small, isolated areas at the north, west and southwestern parts
619
of the caldera rim points to emissions sites of the ignimbrite (Fig. 67B). In fact, the ring fault
620
scarps and the topographic caldera rim are clearly visible at the western half of the caldera, but
621
they are hidden at the eastern side, suggest that caldera-collapse could have been more intense in
622
the western sector than in the eastern sector. We cannot inform on how the caldera-forming
623
eruption progressed, as most if its corresponding deposits have been eroded outside the caldera or
624
are now totally hidden by post-caldera rocks at the interior of the caldera depression. However,
625
the presence of a large number of normal faults with different orientations affecting intra-caldera
626
rocks suggests that caldera-collapse could have occurred in a piecemeal-trapdoor fashion through
627
several stages of volcanism.
26
628
Early-post caldera phase (2.6-2.2 Ma)
629
After the caldera formation followed a relatively short (~0.1 Ma) quiescence period
630
during which an intra-caldera shallow lake developed with lacustrine sedimentation (Fig. 56B).
631
The early post-caldera volcanism is represented by eruptions that occurred mainly in the interior
632
of the caldera with a minimum total volume of 27 km3 (Fig. 89C). This volcanism vented along
633
the ring faults and intra-caldera normal faults that formed or acted during caldera collapse to then
634
be reactivated by the extensional tectonics that have affected the area until present (García-
635
Palomo et al., 2018). These modifications of the local stress field allowed the ascent of
636
peralkaline basaltic and basaltic trachyandesite magmas that were generated by partial melting of
637
a metaszomatized mantle, genetically unrelated to the calc-alkaline magmas that dominated the
638
pre-caldera and caldera activities (Sosa-Ceballos et al., 2018). Both suites of magma mixed and
639
formed the early post-collapse dome complexes on the ring of the caldera.
640
Late-post caldera phase (~2.0 - ~0.016 Ma)
641
After a quiescent period of around 0.2 Ma the caldera entered a new phase of volcanism
642
along the caldera rim dominated by the emission of domes and lava flows, two ignimbrites, and
643
four scoria cones (Fig. 89D). An extended period of intermittent volcanism occurred between 2
644
and 1.3 Ma with the occurrence of at least 13 effusive eruptions along the caldera rim. Between 2
645
and 1.6 Ma eruptions emitted bimodal andesitic (Coal, Tal, Cual units) and rhyolitic (Alrd, Lrd
646
units) products to then erupted rhyolitic effusive magmas until 1.28 Ma (Ard unit) at the
647
southwestern part of the caldera rim. At about the same time (1.27 Ma) a rhyolitic magma
648
explosively flared explosion occurred at the eastern part of the caldera dispersing to the through
649
the eastern part of the caldera dispersing to the east and northeast the Encimadas ignimbrite (26
650
km3) Encimadas ignimbrite (26 km3). Encimadas was apparently erupted from a previous sealed
651
portion of the caldera (tilted blocked to the west of the piece meal collapse) since no Acoculco
27
652
ignimbrite has been found eastern of the caldera. We assume that magmatic overpressure in a
653
shallow magma chamber triggered the rhyolitic eruption that represents the second largest event
654
of the caldera. Immediately after this major eruption occurred four rhyolitic eruptions in the S-
655
SW parts of the caldera between 1.27 and 1.06 Ma. These eruptions particularly emitted the
656
voluminous Ailitla coulée dome to the south (Arcd; 12.4 km3), the Cabezas lava to the southwest
657
(Crl), a minor explosive event that dispersed small-volume pyroclastic density currentsPDCs
658
(Msf), and the Minilla coulée come to the west (Mrcd). These effusive eruptions precluded
659
preceded the occurrence of another explosive rhyolitic event that occurred some 0.2 Ma later.
660
This eruption vented at the southwestern edge of the caldera rim (0.6-0.76 Ma) at the intersection
661
with the NW-SE Manzanito fault. This eruption dispersed the Tecoloquillo rhyolitic ignimbrite
662
(11 km3) to the southwest of the caldera rim ending with the extrusion of a rhyolitic summit
663
dome. The Tecoloquillo ignimbrite represents the third largest explosive event of the ACC after
664
which activity associated to the caldera apparently declined for a period of circa 0.5 Ma.
665
However, renewed activity of the caldera started 0.071 Ma with the emplacement of the La Paila
666
(Plc) at the southeast, the Tuliman scoria cone and lava flow (0.063 Ma) at the northwest, and the
667
Cuatzitzinguito (16 ka) and Cuatzitzingo ca. 0.01 Ma (16 ka) scoria cones at the southeast. These
668
magmas have erupted basaltic andesite to basaltic trachyandesite products (Sosa-Ceballos et al.,
669
2018).
670
During this late post-caldera period the peralkaline magma suite gradually dominated
671
over the calc-alkaline suite (Sosa-Ceballos et al., 2018) (Figs. 89D-C). Thise mixing process
672
resulted in an undistinguishable unique source for the early post-collapse magmas, whereas trace
673
elements geochemistry suggests a relatively homogeneous source for the late post-collapse
674
rhyolites (Sosa-Ceballos et al., 2018). The total volume of magma erupted during late post-
675
caldera phase was ~90 km3. We believe that during this phase of the caldera evolution began the
28
676
resurgence of the caldera floor as supported suggested by uplifted and tilted lacustrine sediments
677
(250°/30SE) exposed inside the caldera (Fig. 2) covering the Acoculco ignimbrite. The
678
morphology of the caldera floor suggests that the resurgence was more important in the
679
southwestern portion, which is in accordance with shallow intrusions of magma revealed by
680
aeromagnetic data that could cause the caldera resurgence. According to Sosa-Ceballos et al.
681
(2018) these shallow magma intrusions were originated by post-caldera deformation that
682
promoted the formation ofa swarm of dikes and sills and dykes above the collapsed reservoir and
683
might be the heat source of the Acoculco geothermal system.
684
Finally, the area covered by the AAC products is ca. 856 km2 with a total minimum
685
volume ca. 143 km3. By assuming that the present extension of deposits is minimum, we
686
calculated that activity was dominated by nearly 80 % of effusive volcanism with three major
687
bursts of explosive volcanism that represent 20% of the ACC. As summarized above Acoculco
688
has been a site of persistent volcanism since 2.7 Ma that include the presence of a young
689
intrusion (183 ± 36 ka) at the bottom of the EAC-1 well. This intense magmatism, as well as
690
Holocene hydrothermal explosions (Canet et al., 2015a; 2015b), and geothermal manifestations
691
(Peiffer et al., 2014) indicate that the complex is still active and could represent a site to develop
692
a geothermal energyfield.
693
CONCLUSIONS
694
The Acoculco collapse caldera was originated at about ~2.7 Ma ago in response to the
695
eruption of 127 km3 of the andesitic ignimbrite of the Acoculco unit. The caldera collapse episode
696
occurred in part along a pre-existing regional NE-SW trending faults (southwestern, western and
697
northwestern sectors), but also along newly formed ring faults (northern sector and probably the
698
eastern sector, but this is not visible). The collapse occurred in a piecemeal-trapdoor fashion, in
699
which intra-caldera blocks bounded by pre-existing or newly formed normal faults collapsed
29
700
gravitationally collapsed in a partly emptied to the associated magma chamber. After the
701
formation of the caldera, volcanic activity stopped for a while permitting the formation of an
702
intra-caldera lake as suggested by, which sediments covering part ofed the caldera-forming
703
ignimbrite. The rRenewedal of the volcanic activity produced the emplacedment of several
704
domes and lava flowsvolcanic units along the caldera borders and intra-caldera faults that entirely
705
filledmodified the caldera depression. Post-caldera volcanism also affected the areas around the
706
caldera and was controlled by the main regional fault systems that also influenced the formation
707
of the caldera. The ACC was formed at the intersection of NE-SW and NW-SE fault systems.
708
These fault systems have controlled the formation and evolution of the ACC, the regional
709
subsidence, and venting of syn- and post-caldera volcanism inside and outside the caldera.
710
The presence of positive aeromagnetic anomalies associated to intrusive bodies (sills and
711
dykes) thermal anomaly at the interior of the Acoculco caldera makes it an interesting target for
712
geothermal exploration. This geothermal anomaly is probably related to the existence of more
713
than ~2.7 Ma of post-caldera volcanism and the magma trapped in horizontal intrusions, as
714
suggested by aeromagnetic data.
715
ACKNOWLEDGEMENTS
716
This study was partially funded by the Centro Mexicano de Innovación en Energía
717
Geotérmica (CeMIE-Geo) project P15 and GeMex 4.4. to J.L. Macías. We thank F. Mendiola, G.
718
Reyes-Agustín and S. Cardona for their technical support during the laboratory analyses and D.E.
719
Torres-Gaytan for his support during the aeromagnetic map generation. We appreciate the
720
discussions and input provided by C. Arango, C. Canet and J. Marti is grateful for the MECD
721
(PRX16/00056) grant. We appreciated the constructive comments by J. Aranda and two
722
anonymous reviewers.
723
REFERENCES CITED
30
724
Afanasyev, A., Costa, A., Chiodini, G., 2015. Investigation of hydrothermal activity at Campi
725
Flegrei caldera using 3D numerical simulations: Extension to high temperature
726
processes. Journal of Volcanology and Geothermal Research, 299, 68-77.
727
Aguirre-Dı́az, G.J., López-Martı́nez, M., 2001. The Amazcala caldera, Queretaro, Mexico.
728
Geology and geochronology. Journal of Volcanology and Geothermal Research, 111(1-4),
729
203-218.
730
Aguirre‐Díaz, G.J., Nieto‐Obregón, J., Zúñiga, F.R., 2005. Seismogenic basin and range and
731
intra‐arc normal faulting in the central Mexican Volcanic Belt, Querétaro, México. Journal of
732
Geology, 40(2), 215-243.
733
Aguirre-Díaz, G.J., Labarthe-Hernández, G., Tristán-González, M., Nieto-Obregón, J., Gutiérrez-
734
Palomares, I., 2008. The ignimbrite flare-up and graben caldera of the Sierra Madre
735
Occidental, México. In: Gottsmann J, Martí J. (Eds.), Caldera volcanism: analysis, modeling
736
and response. Elsevier, Amsterdam, 143–174.
737
Alatriste-Vilchis, D., Campos-Enriquez, O., Huizar-Alvarez, R., Marines-Campos, R., 2005. La
738
estructura sub-superficial de la subcuenca de Tecocomulco. La Laguna de Tecocomulco Geo-
739
Ecologia de un desastre. Universidad Nacional Autónoma de México, Instituto de Geología,
740
Publicación especial, 3, 33-48.
741
Angelier, J., 1990. Inversion of field data in fault tectonics to obtain the regional stress—III. A
742
new rapid direct inversion method by analytical means. Geophysical Journal International, 103
743
(2), 363-376.
744
Aranda-Gómez, J.J., McDowell, F.W., 1998. Paleogene extension in the southern Basin and
745
Range province of México: Syndepositional tilting of Eocene red beds and Oligocene volcanic
746
rocks in the Guanajuato Mining District. International Geology Review, 40 (2), 116-134.
747
Avellán, D.R., Macías, J.L., Layer, P.W., Cisneros, G, Sánchez-Núñez, J.M., Gómez-
748
Vasconcelos, M.G., Pola, A., Sosa-Ceballos, G., García-Tenorio, F., Reyes-Agustín, G.,
749
Osorio-Ocampo, S., García-Sánchez, L., Mendiola, I.F., Marti, J., López-Loera H., Benowitz,
750
J., 2018. Geology of the late Pliocene – Pleistocene Acoculco caldera complex, eastern Trans-
751
Mexican Volcanic Belt (México). Journal of Maps, 15(2), 8-18.
752
Bibby, H.M., Caldwell, T.G., Davey, F.J., Webb, T.H., 1995. Geophysical evidence on the
753
structure of the Taupo Volcanic Zone and its hydrothermal circulation. Journal of
754
Volcanology and Geothermal Research, 68 (1-3), 29-58.
31
755
756
Bursik, M., 2009. A general model for tectonic control of magmatism: Examples from Long
Valley Caldera (USA) and El Chichón (México). Geofísica internacional, 48(1), 171-183.
757
Caballero, M., Lozano, S., Ortega, B., Urrutia, J., Macias, J.L., 1999. Environmental
758
characteristics of Lake Tecocomulco, northern basin of México, for the last 50,000
759
years. Journal of Paleolimnology, 22 (4), 399-411.
760
Calcagno, P., Evanno, G., Trumpy, E., Gutiérrez-Negrín, L.C., Macías, J.L., Carrasco-Núñez, G.,
761
Liotta, D., the GEMex T3.1 team, 2018. Preliminary 3-D geological models of Los Humeros
762
and Acoculco geothermal fields (Mexico) – H2020 GEMex Project. Advances in Geosciences,
763
1, 1-13.
764
Campos-Enríquez, J.O., Alatriste-Vilchis, D.R., Huizar-Álvarez, R., Marines-Campos, R.,
765
Alatorre-Zamora, M.A., 2003. Subsurface structure of the Tecocomulco sub-basin
766
(northeastern México basin), and its relationship to regional tectonics. Geofísica Internacional,
767
42 (1), 3-24.
768
Canet, C., Hernández-Cruz, B., Jiménez-Franco, A., Pi, T., Peláez, B., Villanueva-Estrada, R.E.,
769
Salinas, S., 2015a. Combining ammonium mapping and short-wave infrared (SWIR)
770
reflectance spectroscopy to constrain a model of hydrothermal alteration for the Acoculco
771
geothermal zone, Eastern México. Geothermics, 53, 154-165.
772
Canet, C., Trillaud, F., Prol-Ledesma, R.M., González-Hernández, G., Peláez, B., Hernández-
773
Cruz, B., Sánchez-Córdova, M.M., 2015b. Thermal history of the Acoculco geothermal
774
system, eastern México: Insights from numerical modeling and radiocarbon dating. Journal of
775
Volcanology and Geothermal Research, 305, 56-62.
776
777
778
779
Castro-García, A., Córdoba, D., 1994. Estratigrafía del área volcánica de Tulancingo, Hidalgo,
México. Revista de Investigación, Universidad Autónoma del Estado de Hidalgo, 1, 26-39.
Costa, A., Martí, J., 2016. Stress field control during large caldera-forming eruptions. Frontiers
in Earth Science, 4 (92), 1-13.
780
De Cserna, Z.D., De la Fuente-Duch, M., Palacios-Nieto, M., Triay, L., Mitre-Salazar, L. M.,
781
Mota-Palomino, R., 1988. Estructura geológica, gravimetría, sismicidad y relaciones
782
neotectónicas regionales de la cuenca de México. Boletín del Instituto de Geología,
783
Universidad Nacional Autónoma de México, 104, 71.
784
785
De la Cruz-Martínez, V., Castillo-Hernández, D., 1986. Geología de la zona geotérmica de la
caldera de Acoculco, Puebla. Geotermia, Revista Mexicana de Geoenergía, 2 (3), 245-254.
32
786
Di Napoli, R., Martorana, R., Orsi, G., Aiuppa, A., Camarda, M., De Gregorio, S., Candela E.G.,
787
Luzio D., Messina, N., Pecoraino, G., Bitetto, M., de Vita, S., Valenza M., 2011. The structure
788
of a hydrothermal system from an integrated geochemical, geophysical, and geological
789
approach: The Ischia Island case study. Geochemistry, Geophysics, Geosystems, 12 (7), 1-25.
790
De Vicente, G., Muñoz, A., Giner, J.L., 1992. El uso del método de los diedros rectos:
791
implicaciones a partir del modelo de deslizamiento del análisis poblacional de fallas. Revista
792
Sociedad Geológica de España, 5, 3-4.
793
Ellis, R.G., De Wet, B., MacLeod, I.N., 2012. Inversion of magmatic data for remanent and
794
induced sources: Auxiliary Information in Geophysical Inversion. Presented at the ASEG
795
Inversion Forum, extended abstract.
796
797
García-Estrada, G.H., 2000. Gravimetría y magnetometría regional del proyecto Acoculco,
Puebla. CFE-GPG internal report DEX-ACO19/2000, 57.
798
García-Palomo, A., Macías, J.L., Garduño, V.H., 2000. Miocene to Recent structural evolution
799
of the Nevado de Toluca volcano region, central México. Tectonophysics, 318 (1), 281-302.
800
García-Palomo, A., Macías, J.L., Tolson, G., Valdez, G., Mora, J.C., 2002. Volcanic stratigraphy
801
and geological evolution of the Apan región, east-central sector of the Trans-Mexica Volcanic
802
Belt. Geofísica Internacional, 41 (2), 133-150.
803
García-Palomo, A., Macías, J.L., Jiménez, A., Tolson, G., Mena, M., Sánchez-Núñez, J.M.,
804
Arce, J.L, Layer, P.W., Santoyo, M.A., Lermo-Samaniego, J., 2018. NW-SE Pliocene-
805
Quaternary extension in the Apan-Acoculco region, eastern Trans-Mexican Volcanic Belt.
806
Journal of Volcanology and Geothermal Research, 349, 240-255.
807
García-Sánchez, L., Macías J.L., Arce, J.L., Sosa-Ceballos, G., Garduño-Monroy, V.H.,
808
Saucedo, R., Avellán, D.R., Rangel, E., Layer, P.W. Lopez-Loera, H., Rocha, V.S., Cisneros,
809
G., Reyes-Agustín; G., Jiménez, A., 2016. Genesis and evolution of the Cerro Prieto Volcanic
810
Complex, Baja California, Mexico. Bulletin of Volcanology, 79 (44), 1-28.
811
García-Tovar, G.P., Martínez-Serrano, R.G., Solé, J., Correa-Tello, J.C., Nuñez-Castillo, E.Y.,
812
Guilluo, H., Monroy-Rodríguez, E., 2015. Geología, geocronología y geoquímica del
813
vulcanismo Plio-Cuaternario del Campo Volcánico Apan-Tecocomulco, Faja Volcánica
814
Trans-Mexicana. Revista Mexicana de Ciencias Geológicas, 32 (1), 100-122.
815
Geyer, A., Martí, J., 2014. A short review of our current understanding of the development of
816
ring faults during collapse caldera formation. Frontiers in Earth Science, Volcanology, 2, 1-22.
33
817
818
Henry, C.D., Aranda-Gomez, J.J., 1992. The real southern Basin and Range: Mid-to late
Cenozoic extension in México. Geology, 20 (8), 701-704.
819
Hernández-Javier, I., 2008. Geología y geomorfología volcánica de la región de los yacimientos
820
de obsidiana en Otumba en el sector norte de la Sierra Nevada de México. Bacherlor thesis,
821
Universidad Nacional Autónoma de México.
822
Hiriart, G., Gutiérrez, N. L., Quijano, L.L., Ornelas, C.A., Espíndola, S., Hernández, I., 2011.
823
Evaluación de la energía geotérmica en México. Informe para el Banco Interamericano de
824
Desarrollo y la Comisión Reguladora de Energía, México.
825
Huízar-Álvarez, R., Campos-Enríquez, J.O., Lermo-Samaniego, J., Delgado-Rodríguez, O.,
826
Huidobro-González, A., 1997. Geophysical and hydrogeological characterization of the sub-
827
basins of Apan and Tochac (México Basin). Geofísica Internacional, 36, 217-234.
828
Izquierdo, G., Aragón, A., Portugal, E., Arellano, V.M., De León, J., Álvarez, J., 2006.
829
Mineralogía de la zona mineralizada de sílice-epidota (ZMSE) del yacimiento geotérmico de
830
Cerro Prieto, B. C., México. Geotermia, Revista Mexicana de Geoenergia, 19, 2-12.
831
832
Lorenzo-Pulido, C., Armenta-Flores, M., Ramírez-Silva, G., 2010. Characterization of the
833
Acoculco Geothermal Zone as a HDR System. Geothermal Resources Council Transactions,
834
34, 369–372.
835
836
837
838
Johnson, C.A., Harrison, C.G.A., 1990. Neotectonics in central México. Physics of the Earth and
Planetary Interiors, 64, 187-210.
Layer, P.W., 2000. Argon-40/argon-39 age of the El’gygytgyn impactevent, Chukotka, Russia.
Meteorite Planetary Science, 35, 591–599.
839
Layer, P.W., García-Palomo, A., Jones, D., Macías, J.L., Arce, J.L., Mora, J.C., 2009. El
840
Chichón volcanic complex, Chiapas, México: stages of evolution based on field mapping and
841
40Ar/39Ar geochronology. Geofisica Internacional, 48, 33–54.
842
Ledezma-Guerrero, O., 1987. Hoja Calpulalpan 14Q-h (3), con Resumen de la geología de la
843
hoja Calpulalpan, Estados de México, Puebla, Tlaxcala e Hidalgo. Quetzalapa region, southern
844
México. Revista Mexicana de Ciencias Geológicas, 15, 167-180.
845
Lermo, J., Antayhua, Y., Bernal, I., Venegas, S., Arredondo, J., 2009. Monitoreo sísmico en la
846
zona geotérmica de Acoculco, Puebla, México. Geotermia, Revista Mexicana de Geoenergía,
847
22 (1), 40-58.
34
848
849
Lipman, P.W., 2000. Calderas. In: Sigurdsson, H., Houghton, B.F., McNutt, S.R., Rymer, H.,
Stix, J. (Eds.), Encyclopedia of Volcanoes. Academic press, San Diego, 331-343.
850
López-Hernández, A., 2009. Evolución volcánica del complejo Tulancingo-Acoculco y su
851
sistema hidrotermal, Estados de hidalgo y Puebla, México. Ph.D. thesis, Universidad Nacional
852
Autónoma de México.
853
Lopez-Hernandez, A., Castillo-Hernandez, D., 1997. Exploratory drilling at Acoculco, Puebla,
854
México: a hydrothermal system with only nonthermal manifestations (No. CONF-971048).
855
Geothermal Resources Council, Davis, CA (United States).
856
López-Hernández, A., García-Estrada, G., Aguirre-Díaz, G., González-Partida, E., Palma-
857
Guzmán, H., Quijano-León, J., 2009. Hydrothermal activity in the Tulancingo-Acoculco
858
Caldera Complex, central México - Exploratory studies. Geothermics, 38, 279-293.
859
López-Hernández, A., Martínez, E.I., 1996. Evaluación volcanológica y estructural de la zona
860
geotérmica de Acoculco, Puebla, y su relación con la anomalía termal detectada en el pozo
861
EAC-1. CFE-GPG Internal Report OGL-AC-11/96.
862
Macías, J.L., Arce, J.L., García-Tenorio, F., Layer, P.W., Rueda, H., Reyes-Agustin, G.,
863
Avellán, D., 2012. Geology and geochronology of Tlaloc, Telapón, Iztaccíhuatl, and
864
Popocatépetl volcanoes, Sierra Nevada, central México. Field Guides, 25, 163-193.
865
866
867
868
869
870
Martí, J., Geyer, A., Folch, A., Gottsmann, J., 2008. A review on collapse caldera
modelling. Developments in Volcanology, 10, 233-283.
Martí, J., Geyer, A., Aguirre-Diaz, G., 2013. Origin and evolution of the Deception island
caldera (South Shetland Islands, Antarctica). Bulletin of Volcanology, 75 (6), 732.
McDougall I., Harrison, T.M., 1999. Geochronology and thermochronology by the 40Ar/39Ar
method, 2nd edition. Oxford University Press, New York.
871
Molina, F., Martí, J., Aguirre, G., Vega, E., Chavarría, L., 2014. Stratigraphy and structure of the
872
Cañas Dulces caldera (Costa Rica). Geological Society of America Bulletin, 126 (11-12),
873
1465-1480.
874
875
Mooser, F., Ramírez, M.T., 1987. Faja Volcánica Transmexicana: Morfoestructura, tectónica y
vulcanotectónica. Boletín de la Sociedad Geológica Mexicana, 48 (2), 69-73.
876
Pardo, M., Suárez, G., 1995. Shape of the subducted Rivera and Cocos plates in southern
877
México: Seismic and tectonic implications. Journal Geophysical Research, 100 (12), 357-12.
35
878
Peiffer, L., Bernard-Romero, R., Mazot, A., Taran, Y.A., Guevara, M., Santoyo, E., 2014. Fluid
879
geochemistry and soil gas fluxes (CO 2–CH 4–H 2 S) at a promissory Hot Dry Rock
880
Geothermal System: The Acoculco caldera, México. Journal of Volcanology and Geothermal
881
Research, 284, 122-137.
882
Pérez‐Campos, X., Kim, Y., Husker, A., Davis, P.M., Clayton, R.W., Iglesias, A., Pacheco, J.,
883
Singh, S., Constantin, M.V., & Gurnis, M., 2008. Horizontal subduction and truncation of the
884
Cocos Plate beneath central Mexico. Geophysical Research Letters, 35(18).
885
Polak, B.G., Prasolov, E.M., Kononov, V.I., Verkhocskiy, A.B., González, A., Templos, L. A.,
886
Mañón, A., 1982. Isotopic composition and concentration of inert gases in Mexican
887
hydrothermal systems (genetic and applied aspects). Geofísica Internacional, 21 (3), 193-227.
888
Rocha, S., Jiménez, E., Palma, H., 2006. Propuesta para dos pozos exploratorios en el proyecto
889
geotérmico de Acoculco, Puebla. Comisión Federal de Electricidad, internal report OGL-
890
ACO-03/06, unpublished.
891
Renne P.R., Mundil, R., Balco, G., Min, K., Ludwig, K.R., 2010. Joint determination of 40K
892
decay constants and 40Ar*/40K for the Fish Canyon sanidine standard, and improved
893
accuracy for 40Ar/39Ar geochronology. Geochimica et Cosmochimica Acta, 74, 5349–5367
894
Rueda, H., Macías, J.L., Arce, J.L., Gardner, J.E., Layer, P.W., 2013. The~ 31ka rhyolitic
895
Plinian to sub-Plinian eruption of Tlaloc Volcano, Sierra Nevada, central México. Journal of
896
Volcanology and Geothermal Research, 252, 73-91.
897
Ryan, W.B., Carbotte, S.M., Coplan, J.O., O'Hara, S., Melkonian, A., Arko, R., Weissel, R.A.,
898
Ferrini, V., Goodwillie, A., Nitsche, F., Bonczkowski, J., 2009. Global multi‐resolution
899
topography synthesis. Geochemistry, Geophysics, Geosystems, 10 (3).
900
901
Saxby, J., Gottsmann, J., Cashman, K., & Gutiérrez, E., 2016. Magma storage in a strike-slip
caldera. Nature communications, 7, 12295.
902
Siebe, C., Macias, J.L., Abrams, M., Rodriguez, S., Castro, R., Delgado, H., 1995. Quaternary
903
explosive volcanism and pyroclastic deposits in east central México: implications for future
904
hazards. In: Chacko, J., Whitney A. (Eds.), Guidebook of geological excursions, in
905
conjunction with the Annual Meeting of the Geological Society of America. New Orleans,
906
Louisiana, 1-48.
907
Sosa-Ceballos, G., Macías, J.L., Avellán, D.R., Salazar-Hermenegildo, N., Boijseauneau-López,
908
M.E., Pérez-Orozco, J.D., 2018. The Acoculco Caldera Complex magmas: Genesis, evolution
36
909
and relation with the Acoculco geothermal system. Journal of Volcanology and Geothermal
910
Research, 358, 288-306.
911
Steiger, R., Jäger, E., 1977. Subcommission on geochronology: convention on the use of decay
912
constants in geo-and cosmochronology. Earth and planetary science letters, 36 (3), 359-362.
913
Suter, M., 1984. Cordilleran deformation along the eastern edge of the Valles–San Luis Potosí
914
carbonate platform, Sierra Madre Oriental fold-thrust belt, east-central Mexico. Geological
915
Society of America Bulletin, 95(12), 1387-1397.Suter, M., 1991. State of stress and active
916
deformation in México and western Central America. Neotectonics of North America, 1, 401-
917
421.
918
Suter, M., Carrillo-Martínez, M., Quintero-Legorreta, O., 1996. Macroseismic study of shallow
919
earthquakes in the central and eastern parts of the trans-Mexican volcanic belt,
920
México. Bulletin Seismological Society of America, 86 (6), 1952-1963.
921
922
923
924
Tello-Hinojosa, E., 1986. Geoquímica de la zona geotérmica de la caldera de Acoculco, Puebla.
CFE-GPG internal report 34/86, 15.
Urrutia-Fucugauchi, J., Flores-Ruiz, J., 1996. Bouger gravity anomalies and regional crustal
structure in central Mexico. International Geology Review, 38, 176–194.
925
Viggiano-Guerra, J.C., Flores-Armenta, M., Ramírez-Silva, G.R., 2011. Evolución del sistema
926
geotérmico de Acoculco, Pue., México: un estudio con base en estudios petrográficos del pozo
927
EAC-2 y en otras consideraciones. Geotermia, Revista Mexicana de Geoenergia, 24 (1), 14-
928
24.
929
Zúñiga, F.R., Pacheco, J.F., Guzmán-Speziale, M., Aguirre-Dıaz, G.J., Espındola, V.H., Nava,
930
E., 2003. The Sanfandila earthquake sequence of 1998, Querétaro, México: activation of an
931
undocumented
932
Belt. Tectonophysics, 361 (3), 229-238.
fault
in
the
northern
edge
of
central
Trans-Mexican
Volcanic
933
934
FIGURE CAPTIONS
935
Figure 1. A) Regional tectonic configuration of Central México showing the active subduction of
936
the Rivera and Cocos Plates beneath the North American Plate at the Middle-American Trench.
937
The dotted black line depicts the boundary of the Trans-Mexican Volcanic Belt (TMVB) and the
938
location of the Acoculco Caldera Complex and other volcanoes (LP, La Primavera; C, Fuego de
37
939
Colima; T, Tancítaro; Am, Amealco; Hc, Huichapan; To, Nevado de Toluca; P, Popocatepetl; M,
940
Malinche; O, Pico de Orizaba; and Hm, Humeros). The shaded relief map and bathymetric data
941
was acquired and modified from Ryan et al. (2009).
942
Figure 2. Simplified geological map of the Acoculco Caldera Complex after Avellán et al.
943
(2018), Sosa-Ceballos et al. (2018), and Calcagno et al. (2018). The map contains the distribution
944
of 30 stratigraphic units and a composite stratigraphic column supported by 24 new
945
ages and two 14C dates (Fig. 3). This map shows towns, main roads, the location of stratigraphic
946
logs, samples with age determinations, faults, the zone of intense hydrothermal alteration, and
947
the two CFE exploratory wells. The DEM was obtained from the INEGI topographic data with a
948
horizontal resolution of 15-m2. The geologic profile at the bottom correlates the surface units of
949
this map with the CFE stratigraphic logs of the EAC-1 and EAC-2 wells with the location of
950
shallow intrusions.
951
Figure 3. Detailed stratigraphic column of the units and eruptive stages of the Acoculco Caldera
952
Complex composed of the 30 units displayed marked in the geological map of figure 2. To the
953
right appear the plateau age of units, the plateau age sorted by age, and their estimated total
954
volume.
955
Figure 4. Representative
956
the Acoculco Caldera Complex (see Table 1). Plateau and Isochron ages, where calculated, are
957
reported at the 1-sigma level. For Ac90, the 6 single-step fusion results are shows as a single
958
probability density histogram
959
rocks of the Acoculco Caldera Complex (see Table 1).
960
Figure 5. Photos of selected outcrops of the Acoculco andesitic ignimbrite (Aai) with its different
961
lithological facies. A, Aspect of the welded massive structure of the ignimbrite with pumice
962
fragments embedded in a light-yellowish ash matrix. Hand lens for scale. B, Lacustrine deposits
40Ar/39Ar
40Ar/39Ar
age spectra and inverse isochron diagrams of dated rocks of
40Ar/39Ar
age spectrums and inverse isochron diagrams of dated
38
963
that partly covereding the Aai with parallel plane stratification. They are made of interbedded
964
massive fine sandy beds with laminar-stratified clayed beds. Notice that the succession is gently
965
tilted. C, Co-ignimbrite lag breccia (mlBr) of the Acoculco ignimbrite transected by E-W
966
structures and NW-SE faults. Notice that the lava lithics are supported by a light-brown to
967
yellowish ashy matrix with pumice. D, Co-ignimbrite lag breccia (mlBr) at the bottom of the
968
outcrop in contact with the Acoculco ignimbrite. The color change marks a difference in the
969
texture and lithological components of the deposits.
970
Figure 6.
971
Caldera Complex (see Table 1). Plateau and Isochron ages, where calculated, are reported at the
972
1-sigma level.
973
Figure 7. A, Regional structural map modified from García-Palomo et al. (2018). Dark gray areas
974
correspond to horsts and light gray areas correspond to grabens. Dike, scoria cones and
975
slickensides symbols show extension direction. B, Structural map of the Acoculco Caldera
976
Complex showing the main structural features, field sites (outcrops) and the caldera rim (this
977
study). C, The fault planes were plotted on stereographic projections in the lower hemisphere.
978
The trend-frequency polar diagramstereogram (rose diagram) was obtained with Rozeta 2.0, and
979
the fault planes and poles were plotted with Stereonet 9. D, A slip model showing an
980
orthorhombic symmetry of NW- and NE- oriented faults moving under the same strain field.
981
Black zones correspond to 100% compatible compression, and white zones 100% tension areas.
982
Figure 8. A, Image of the Magnetic Field Reduced to the Pole (MFRP) overlaid on a DEM of the
983
ACC region. The MFRP values are represented by the horizontal bar color at the base of the
984
figure with red colors standing for high values and the blue ones for low values of the magnetic
985
intensities. The black dots represent the CFE wells. B, A 3D inversion model of the magnetic
986
susceptibility produced with the Geosoft´s VOXI software by using the magnetic vector inversion
40Ar/39Ar
age spectra and inverse isochron diagrams of dated rocks of the Acoculco
39
987
method (Ellis et al., 2012). In this image four anomalies related to intrusive bodies occur at
988
depths between 1,000 and >2,500 m below the surface.
989
Figure 9. Eruptive stages during the formation of the Acoculco Caldera Complex described as
990
pre-caldera (A), syn-caldera (B), early post-caldera (C) late post-caldera (D), and still late post-
991
caldera and extra-caldera (E). After the caldera collapse two episodes of caldera reactivation
992
generated the Encimadas (Eri), and Tecoloquillo (Tr1) rhyolitic ignimbrites. Outside of the
993
caldera rim, Tthe products of the Acoculco caldera are interbbededinterbedded with deposits of
994
the Apan-Tezontepec volcanic field (extra caldera volcanism).
995
LIST OF TABLES
996
Table 1. List of dated samples from rock units of the Acoculco Caldera Complex. See text for
997
map units. All ages (in ka) are reported at the 1-sigma level. For most samples, the 40Ar/39Ar
998
plateau age is the interpreted age (shown in bold) of the unit. Sample Aco901 showed evidence
999
of excess argon, so the isochron age is the interpreted age. The age of sample Ac90 is a weighted
1000
mean age of 6 single-step fusions, and for sample Ac94a, no plateau or isochron could be
1001
calculated. MSWD: Mean Square Weighted Deviation. See text for analytical methods and
1002
standards used. List of rock units of the Acoculco Caldera Complex dated in this study and their
1003
40Ar/39Ar
1004
were calculated using the constants of Steiger and Jäeger (1977). Analyses performed at the
1005
Geochronology Laboratory of the University of Alaska at Fairbanks, USA.
1006
Table 2. Radiocarbon date of paleosol sample performed during study. The location is in UTM
1007
coordinates. Analyses performed at the International Chemical Analysis Inc. Florida, USA.
ages. , pPreferred results are listed in the plateau age (ka) column in bold. All ages are
1008
40
Highlights
Acoculco caldera complex lies at the intersection of NE-SW and NW-SE fault
systems
The complex has emplaced ~143 km3 of magma since 2.7 Ma.
Acoculco has erupted andesitic to rhyolitic magmas from 2.7 to ca. 0.016 Ma
Young intrusions are likely providing the potential geothermal source
1
Eruptive chronology of the Acoculco caldera complex – a resurgent
2
caldera in the eastern Trans-Mexican Volcanic Belt (México)
3
by
4
5
6
7
Denis Ramón Avellán1, José Luis Macías2, Paul W. Layer3, Giovanni Sosa-Ceballos2,
Martha Gabriela Gómez-Vasconcelos4, Guillermo Cisneros-Máximo2, Juan Manuel
Sánchez-Núñez5, Joan Marti6, Felipe García-Tenorio2, Héctor López-Loera7, Antonio Pola8,
and Jeff Benowitz3
8
1CONACYT
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
– Instituto de Geofísica, Universidad Nacional Autónoma de México, Antigua Carretera a
Pátzcuaro 8701, 58190 Morelia, Michoacán
2Instituto de Geofísica, Universidad Nacional Autónoma de México, Antigua Carretera a Pátzcuaro 8701,
58190 Morelia, Michoacán
3College of Natural Science, Mathematics and Geophysical Institute, University of Alaska Fairbanks,
Fairbanks, AK 99775
4CONACYT – Instituto de Investigaciones en Ciencias de la Tierra, Universidad Michoacana de San
Nicolás de Hidalgo, Francisco J. Mujica s/n, Felicitas del Río, 58040, Morelia, Michoacán
5Instituto Politécnico Nacional-CIIEMAD, Miguel Othón de Mendizábal s/n. Col. La Escalera, C.P.
07320 Del. Gustavo A. Madero, Ciudad de México, México
6Instituto de Ciencias de la Tierra Jaume Almera, CSIC, LLuis Sole Sabaris, s/n, 08028 Barcelona, Spain
7División de Geociencias Aplicadas, Instituto Potosino de Investigación Científica y Tecnológica A.C.;
Camino a la Presa San José 2055, Lomas 4a Sección, C-P. 78216, San Luis Potosí S.L.P.
8Escuela Nacional de Estudios Superiores, Universidad Nacional Autónoma de México, Antigua
Carretera a Pátzcuaro 8701, 58190 Morelia, Michoacán
32
Submitted to Journal of South American Earth Sciences
*Corresponding author e-mail: denisavellan@gmail.com
33
34
November 1, 2019
1
36
ABSTRACT
37
The Acoculco caldera complex (ACC) is located in the eastern part of the Trans-Mexican
38
Volcanic Belt in the northern part of the State of Puebla. The complex sits at the intersection of
39
two regional fault systems with NE-SW and NW-SE orientations. ACC was built atop Cretaceous
40
limestones, the Zacatán basaltic plateau of unknown age, early Miocene domes (~12.7-10.98
41
Ma), and Pliocene lava domes (~3.9-3 Ma). Detailed field mapping and stratigraphy studies
42
complemented by
43
30 volcanic units. Based on all these data and previous studies the ACC eruptive chronology was
44
grouped in four eruptive phases: syn-caldera, early post-caldera, late post-caldera, and extra-
45
caldera. Inception of the ACC volcanism began around 2.7 Ma with the dispersion of an andesitic
46
ignimbrite followed by the collapse of the magma chamber roof as attested by the presence of a
47
lithic breccia in isolated parts of the caldera rim. The collapse produced a 18×16 km caldera
48
depression which was partly filled by the ignimbrite (total volume of ~127 km3) followed by the
49
establishment of an intracaldera lake of unknown total extension. Early post-caldera collapse
50
activity (2.6-2.1 Ma) was restricted within the caldera producing 27 km3 of lava flows and domes
51
dominantly of basaltic trachyandesite to basaltic composition. Late post-caldera collapse activity
52
(2.0-<0.016 Ma) migrated dominantly to the caldera rim and periphery emplacing 90 km3 of
53
magma as rhyolitic domes, lavas, scoria cones, and two younger ignimbrites. The 1.2 Ma
54
Encimadas ignimbrite (26 km3) was vented through the eastern margin of the caldera and
55
dispersed to the northeast, and the 0.6-0.8 Ma Tecoloquillo ignimbrite and dome (11 km3)
56
erupted from the southwestern margin of the caldera. The most recent eruption of this phase was
57
vented close to the southeastern caldera rim producing the Cuatzitzingo (<16,710 ± 50 years BP)
58
scoria cone. Extra-caldera activity (2.4-0.19 Ma) of the Apan–Tezontepec volcanic field
40Ar/39Ar
and
14C
dating allowed to divide the ACC volcanic succession into
2
59
produced scoria cones and lava flows of basaltic trachyandesite to basaltic andesite composition
60
that are interbedded with the products of the caldera complex.
61
Aeromagnetic data further constraint the edge of the caldera rim and is consistent with the
62
presence of at least four intrusive bodies at depths of >1 km hosted in the Cretaceous limestones.
63
These bodies might represent a series of horizontal mafic intrusions located at different depths
64
that provide the energy that maintains active the Acoculco geothermal system.
65
66
Keywords: Geology; geochronology; geothermal energy; Acoculco caldera; Puebla
67
68
INTRODUCTION
69
Silicic collapse calderas are volcanic depressions resulting from the subsidence of the
70
magma chamber roof caused by the rapid withdrawal of magma during an explosive eruption
71
(Lipman, 2000; Martí et al., 2008; Geyer and Martí, 2014). The formation of a collapsed caldera
72
is still an enigmatic geological phenomenon because of the structural complexity involved in
73
such type of volcanic eruptions. Besides, silicic caldera eruptions represent a high risk due to the
74
large amount of magma and eruption rates involved (Costa and Martí, 2016). The resulting
75
caldera depression may represent the site of important ore deposits and high enthalpy geothermal
76
reservoirs. This makes the study of caldera systems of great interest to modern societies. In many
77
active volcanic areas, such as Iceland, New Zealand, Costa Rica, Japan, Indonesia, among several
78
others, an important part of their government energy requirements is covered by the exploitation
79
of geothermal resources associated with collapse caldera systems. In México, geothermal energy
80
has been used since 1959, when a program led by the Comisión Federal de Electricidad (CFE)
3
81
started to explore and exploit some of these potential energy resources, most of them located in
82
calderas of the Trans-Mexican Volcanic Belt (TMVB) (Hiriart et al., 2011). In the particular case
83
of geothermal fields installed in calderas, it is very important to determine the mechanism that
84
formed the caldera and the post-collapse internal structure (Bibby et al., 1995; Di Napoli et al.,
85
2011; Molina et al., 2014; Afanasyev et al., 2015) to better understand fluid paths and the
86
location of reservoirs. It is crucial to determine the exact structural limits of the caldera (e.g.,
87
Molina et al., 2014), the stratigraphy and age of the caldera-forming deposits and the distribution
88
of thermal anomalies. In this respect, the Acoculco caldera located in the eastern TMVB (Fig. 1)
89
represents an excellent case scenario to study the internal structure of the edifice, its eruptive
90
chronology, and surface manifestations of geothermal activity concerning fractures and faults.
91
Acoculco has been considered for years a dry-hot rock reservoir by CFE (Lorenzo-Pulido et al.,
92
2010) and is the site of an on-going European-Mexican effort to develop geothermal energy from
93
non-conventional sources called GeMex Project (Calcagno et al., 2018).
94
In this contribution, we analyzed previous studies combined with new fieldwork to
95
present a simplified volcanological map and a new composite stratigraphic column of Acoculco.
96
The results are assisted by 24 new
97
units that combined with previous information set a refined evolution model of the caldera. This
98
information is crucial to present the chronology of the caldera volcanic complex through time. To
99
gain new insights on the caldera structure we produced aeromagnetic models of the area,
100
analyzed the morphostructural distribution of the volcanic units, and their correlation with
101
subsurface units in geothermal wells. Base on this new information, we propose that the late
102
Pleistocene shallow intrusions are still the heat source of geothermal activity beneath Acoculco.
40Ar/39Ar
ages and one
103
104
GEOLOGICAL SETTING
4
14C
radiometric date of the volcanic
105
The Acoculco Caldera Complex (ACC) is located within the TMVB, a calc-alkaline
106
volcanic arc produced by the subduction of the Cocos and Rivera plates beneath the North
107
American plate at the Middle American Trench (Pardo and Suárez, 1995) (Fig. 1). The Acoculco
108
is located in the eastern sector of the TMVB (Pasquaré et al., 1991). Acoculco occurs ~140 km
109
northeast of Popocatépetl volcano which defines in this region the active front of the TMVB
110
(Siebe et al., 1995; Macías et al., 2012) (Fig. 1). Acoculco is located over a 45-50 km-thick
111
continental crust (Urrutia and Flores-Ruíz, 1996), and sits at 400 km from the trench where the
112
Cocos plate plunges into the mantle (Pérez-Campos et al., 2008). This area is under an NW-SE
113
oriented extensional regime, as deduced from the alignment of volcanic vents, dike orientations,
114
extension fractures, and kinematics of faults at the Apan-Acoculco region (García-Palomo et al.,
115
2002; 2018). The Acoculco caldera is bounded to the west by the Apan-Tezontepec volcanic field
116
(ATVF) (~3–0.2 Ma) (García-Palomo et al., 2002; García-Tovar et al., 2015), to the north and to
117
the east by Cretaceous limestones of the Sierra Madre Occidental (Avellán et al., 2018), and to
118
the south by early Miocene volcanics of the TMVB (García-Palomo et al., 2002).
119
Prior to the 80´s the Acoculco geology has not been extensively studied (e.g., Ledezma-
120
Guerrero, 1987; Castro-García and Cordoba, 1994), in fact, CFE produced the first regional map
121
of the area (De la Cruz-Martinez and Castillo-Hernández, 1986) followed by other detailed
122
studies as rock dating (López-Hernández and Castillo-Hernández, 1997; López-Hernández and
123
Martínez, 1996). Later geophysical studies attempted to understand the internal structure of the
124
caldera (Alatriste-Vilchis et al., 2005; Campos-Enriquez et al., 2003) and the hydrological
125
conditions (Huizar-Álvarez et al., 1997) of the Apan region. The first volcanological study of the
126
region was carried out by López-Hernández et al. (2009) who concluded that Acoculco was an
127
18-km wide caldera active from ~1.7 to ~0.2 Ma. These authors considered that Acoculco was
128
nested within the largest 32-km wide Tulancingo Caldera, which was active between ~3.0 and
5
129
~2.7 Ma. Recently, Avellán et al. (2018) presented the first detailed geologic map and
130
stratigraphy of the caldera supported by nine
131
caldera has a semi-circular shape (18–16 km) and that was active from 2.7 to 0.06 Ma, thus, not
132
corresponding with the previous timing of the caldera formation (López-Hernández et al., 2009).
133
This geology of Acoculco was used to present a simplified geologic map of the caldera to
134
describe its geochemical evolution (Sosa-Ceballos et al., 2018) and a preliminary 3D model of
135
the Acoculco subsurface structure (Calcagno et al., 2018).
136
METHODS
137
40Ar/39Ar
dates. These authors concluded that the
Twenty-four whole-rock samples were crushed, sieved and washed in deionized water for
138
40Ar/39Ar
139
rock) were separated for dating. For sample Ac11, both a plagioclase mineral separate and a
140
whole rock sample were separated and dated, and for Ac69, two different rock types were
141
analyzed. The samples were irradiated in position 5C at the McMaster University Nuclear
142
Reactor in Hamilton, Canada, for 0.75 MWh. The standard mineral TCR-2 with an age of 28.619
143
Ma (Renne et al., 2010) was used to calculate the irradiation parameter, “J”. Samples Ac39 and
144
Ac123, were irradiated with the standard mineral MMHb-1 with an age of 523.5. The standards
145
were fused and except for AC-90. The samples step-heated using a laser dating system consisting
146
of 6W argon-ion laser at the Geophysical Institute, University of Alaska Fairbanks, following the
147
technique described in Layer (2000) and Layer et al. (2009). For Sample Ac90, an obsidian, 7
148
small shards were fused, with 6 of the 7 yielding enough gas to calculate a fusion age. The
149
samples were analyzed in a VG3600 mass spectrometer and the measured Ar isotopes were
150
corrected for system blank, mass discrimination and Ca, K, and Cl interference reactions,
151
according to procedures outlined in McDougall and Harrison (1999) and using the standards
152
reported in Renne et al. (2010). System blanks were 2x10-16 mol
isotopic dating (Table 1). For most samples, phenocrysts-free groundmass chips (whole
6
40Ar
and 2x10-18 mol
36Ar,
153
which are 5 to 50 times smaller than fraction volumes. Mass discrimination was monitored by
154
running calibrated air shots.
155
The structural analysis of the caldera system comprised a review of previous and new
156
structural data, after the evaluation of digital elevation models, aerial photographs, topographic
157
maps, fieldwork and the morphological evaluation of the landforms. The fault geometry was
158
characterized by field measurements and geomorphic analyses of the faults using ArcGIS on a 15
159
m-resolution digital elevation model from INEGI (Instituto Nacional de Estadística y Geografía).
160
Grid references on maps are in the WGS 1984 UTM Zone 14N projection. Rozeta 2.0 software
161
was used for plotting the fault data on stereographic projections (lower hemisphere) and to
162
perform trend-frequency analysis with rose diagrams. Stereonet 9 software was used for plotting
163
the fault planes and poles, from where the compression and tension areas were inferred into the
164
synthetic right dihedral diagram from the two pairs of conjugate fault systems, according to their
165
slip models (e.g., De Vicente et al., 1992).
166
To estimate the volume of the volcanic units we used the geological map, the Digital
167
Elevation Model (DEM), Spot-6 satellite images (1.5-m panchromatic and 6-m multispectral),
168
and the shaded relief DEM (15-m resolution) in the ArcMap 9.3 software. The difference
169
between the actual topography and the geomorphologic element of each unit was used to obtain a
170
z value, and to create a 3D surface geology on the shaded relief DEM and the volume with the
171
Surface Difference tool in ArcMap 9.3.
172
VOLCANIC STRATIGRAPHY
173
In this study we analyzed the regional (García-Palomo et al., 2002; López-Hernández et
174
al., 2009) and local geologic maps of the ACC and its surroundings (Avellán et al., 2018; Sosa-
175
Ceballos et al., 2018; Calcagno et al., 2018), and the geochronological data include in them. We
176
compiled a new simplified geologic map (Fig. 2) supported by 24 new 40Ar/39Ar and one 14C date
7
177
(Tables 1 and 2). The map includes three pre-Acoculco units, two pre-caldera, 30 ACC, and nine
178
extra-caldera units. The ACC units were subdivided in syn-caldera, early post-caldera, and late
179
post-caldera, corresponding to different phases of the caldera formation defined by their
180
distribution, stratigraphic position, age, mineral and chemical composition (Fig. 3). The chemical
181
variations of the ACC rocks have been previously documented and discussed and will not be
182
treated in this contribution (Sosa-Ceballos et al., 2018). Next, we succinctly described all units
183
that span in age from ~2.7 Ma to <0.016 Ma with their location given with respect to the caldera
184
rim.
185
Undifferentiated Cretaceous Limestones (Ksl)
186
The oldest rocks in the region are Cretaceous limestones (Ksl) of the Sierra Madre
187
Oriental that are exposed to the northeastern, eastern, and southeastern parts of the mapped area
188
(Fig. 2). Good outcrops appear in the Tenexapa and Ajajalpan canyons, close to the towns of
189
Chignahuapan and Zacatlán. At the Chignahuapan hot springs these rocks occur as light-gray
190
parallel stratified limestones with chert concretions sometimes affected by vertical fractures filled
191
with hydrothermal minerals (e.g., calcite). Although these rocks are not exposed inside the ACC,
192
they were cut in the two geothermal CFE wells at depths of ~1,200 m in EAC-1 (López-
193
Hernández et al., 2009), and 350-450 m in EAC-2 (Viggiano-Guerra et al., 2011). According to
194
these authors, Ksl was intruded by a light-gray phaneritic granite found at the bottom of EAC-1
195
drill hole at depths from 1800 to 2000 m. The intrusion produced a methamorphic aureole of
196
skarns (Viggiano-Guerra et al., 2011). A sample analyzed of this well resulted to be an aplite
197
made of alkali feldspar, plagioclase, quartz, amphibole, chlorite, and Fe-Ti oxides. An isochron
198
40Ar/39Ar
199
which may correspond to either younger intrusions beneath Acoculco or a reset age of an older
200
regional plutonic body (López-Hernández et al., 2009; Calcagno et al., 2018).
age of this sample yielded an age of 183 ± 36 ka (Table 1, isochron age; Fig. 4A),
8
201
Zacatlán basaltic plateau (Za)
202
Za is a dark-gray, up to 200 m thick basaltic lava plateau that discordantly overlies the
203
Cretaceous limestones near the Chignahuapan and Zacatlán towns (Fig. 2). The lava flow is
204
aphanitic with columnar jointing and spheroidal weathering. The age of this unit is unknown,
205
however, its stratigraphic position indicates that Za is younger than the Cretaceous limestones.
206
Peñuela dacitic dome complex (~13 to 10 Ma) (Mv)
207
These rocks are exposed southern the ACC and belong to the first beginnings of regional
208
volcanism of the eastern part of the TMVB (Fig. 2). The oldest pre-caldera rocks are the Peñuela
209
dacitic (Pdd unit) and Quexnol andesitic dome complexes exposed to the SW and SE parts of the
210
ACC, respectively (Fig. 3). The Peñuela dome was dated with the K-Ar method at 12.7 ± 0.6 Ma
211
(García-Palomo et al., 2002). Another sample of this rock (Pdd) was dated here with the
212
40Ar/39Ar
213
Pre-caldera units (~3.9 – 3.0 Ma) (Pc)
method at 10.98 ± 0.07 Ma (Table 1, plateau age; Fig. 4B).
214
The Puente (Pald) and Terrerillos (Tdld) lava domes are exposed to the N-NW and SW of
215
the geologic map, respectively (Figs. 2 and 3). These rocks are light-gray andesitic to dacitic lava
216
domes with greenish-gray enclaves. Pald yielded in this work an integrated age of 3,620 ± 22 ka
217
(Table 1; Fig. 4C). García-Tovar et al. (2015) reported a K-Ar age of 3.0 ± 0.4 Ma for Tdld,
218
which is consistent with its stratigraphic position.
219
220
Syn-caldera unit (Acoculco andesitic ignimbrite, ~2.7 Ma) (Sc)
221
The Acoculco andesitic ignimbrite (Aai) is a yellow to white massive deposit. It consists
222
of rounded pumice and angular to sub-angular accidental lava fragments (gray, pink and banded)
223
supported by a matrix of coarse to fine ash (Fig. 5A). Tube-banded pumice (greenish-gray to
224
white) with alkali feldspar and amphibole phenocrysts is also present. The Aai crops out in small
9
225
gullies and in some places of high topographic relief where it is greatly covered by younger
226
deposits (e.g., Hbl, Srl, Atal, Fig. 2). In the southwestern and northern parts of the caldera, Aai
227
underlies ≤40 m thick lacustrine deposits (Fig. 5B). The contact between Aai and the underlying
228
Pdd and Pald units was not observed, however, accidental fragments of these units occur in the
229
ignimbrite. At site 24, occurs a massive lithic breccia (mlBr) made of heterolithologic lavas and
230
scarce pumice (Fig. 5C-D). This breccia (5-7 m thick) contains angular accidental lavas set in a
231
coarse to fine ash matrix. Some lava fragments are aphanitic (light-gray and ochre), and other
232
porphyritic (dark-gray, greenish, and ochre) with clinopyroxene and plagioclase phenocrysts. The
233
mlBr matrix contains reddish aphanitic lithics and isolated crystals of plagioclase, pyroxene and
234
amphibole, disseminated pumice and silty minerals. Laterally, this mlBr grades to a massive fine
235
ash layer enriched in pumice and crystals of plagioclase and amphibole. In three sites (24, 65, and
236
114) mlBr is interbedded between flow units of Aai (Fig. 5D).
237
The Aai unit was transected at the EAC-1 exploratory well at depths between 210 and 560
238
m, where it covers the Pald unit (Pre-caldera units) and is overlain by the Pedernal rhyolitic lava
239
unit (Pdl, which is a late post-caldera unit) (Fig. 2). Well EAC-1 shows nearly 800 m of volcanic
240
materials resting atop a skarn which was interpreted as the metamorphosed calcareous basement
241
(López-Hernández et al., 2009; Viggiano-Guerra et al., 2011). These authors interpreted the
242
volcanic column from top to bottom as: Acoculco ignimbrite (0-130 m), Cruz Colorada dacite
243
(130-210 m), Alcholoya ignimbrite (210-580 m), and las Minas rhyodacite (580-790 m).
244
However, based on the new geologic map and revised stratigraphy of Acoculco, we consider that
245
the successions would correspond to our Pdl, Aai and Pald units (Fig. 3). The Pedernal rhyolitic
246
lava (Pdl) crops out around the CFE drill holes as a highly altered vesicular rock with feldspars,
247
plagioclase, quartz and mafic unrecognizable minerals and recrystallized lithic clasts. We
248
consider that these highly altered lavas were mistaken with pumice blocks and erroneously
10
249
described as the Acoculco ignimbrite. Also, we propose that their Alcholoya ignimbrite (~2.7
250
Ma; López-Hernández et al., 2009) described between 210 and 580 m depth would correspond to
251
our Acoculco ignimbrite Aai (Fig. 5A). Unfortunately, Aai was not recognized in the succession
252
of well EAC-2. This well only shows 340 m of volcanic infill (Pdl and Pald units) resting over
253
the skarn including 200 m of the porphyritic unit (Viggiano-Guerra et al., 2011) or our Pdl unit.
254
As mentioned before, EAC-1 and EAC-2 wells are 500 m apart but their stratigraphy suggests
255
that they are separated by a fault that is not recognizable at surface. By averaging these
256
thicknesses in the EAC-1 exploratory well, gullies and outcrops described at the surface (~470 m)
257
we estimate a minimum Aai volume of 127 km3.
258
A pumice fragment separated from Aai was dated with the
40Ar/39Ar
method in
259
plagioclase at 2,732 ±185 ka (Avellán et al., 2018). In this work, we obtained two new whole-
260
rock 40Ar/39Ar dates of Aai pumice samples that yielded younger ages of 2,041 ± 38 and 2,185 ±
261
65 ka (Table 1, plateau age; Figs. 4D-E). However, Aai underlies several early-post caldera lava
262
flows as Aguila (Atal, 2.44 Ma), and Manzanito (Mtal, 2.2 Ma), and the Sayula dome (Srl, ~2.55
263
Ma) (see figures 2 and 3). Based on these stratigraphic relationships, we concluded that the age of
264
Aai must be older than 2.55 Ma for which the 40Ar/39Ar in plagioclase (2,732 ±185 ka) is the best
265
age approximation of the caldera collapse.
266
A sequence of ≥ 40 m thick lacustrine sediments (ls) is exposed inside the caldera rim in
267
the south, southwest, and northern parts. It consists of a tilted alternation of white clayed laminae,
268
and dark-gray cm-thick, volcaniclastic beds. These beds are made of rounded lava fragments set
269
in a fine-grained matrix barren of fossils. At sites 69 and 119, ls overlies the Acoculco ignimbrite
270
(Aai) (Figs. 2 and 3).
271
272
11
273
Early post-caldera units (~2.6-~2.1 Ma) (Epc)
274
Four of these units are exposed inside the caldera depression as basaltic to basaltic
275
trachyandesite lava flows (Hbl, Atal, Vtal and Mtal) that partially cover Aai (Figs. 2 and 3). They
276
are highly eroded and have asymmetric morphologies similar to flatirons with their apex towards
277
the center of the caldera developing a sub-radial exorheic drainage. These units typically appear
278
as light-gray to dark-gray, blocky lava flows with porphyritic to aphanitic textures. These lavas
279
frequently present greenish, yellowish and reddish hydrothermal alteration zones, and host
280
xenoliths of sub-rounded limestones, sandstones, and fine grain granite. Two units were dated
281
with the 40Ar/39Ar method at 2,323 ± 48 ka (Vtal) and 2,199 ± 24 ka (Mtal) (Table 1, plateau age;
282
Figs. 4F-G). Avellán et al. (2018) obtained a
283
Atal. The age of Hbl is unknown; nevertheless, it underlies the Srl dome dated at ~2.55 Ma (see
284
below). The other two early-post-caldera units (Srl and Atad) are exposed to the external
285
northwestern and southeastern rim of the caldera, respectively. The Srl unit (2,553 ± 110 ka;
286
Table 1; Fig. 4H) consists of gray to black and brown, banded obsidian lava flows with
287
holohyaline texture and partially devitrified to light-gray spherulites and lithophysae. The Atad
288
unit (2,179 ± 26 ka; Table 1; Fig. 4I) is a blocky greenish to light-gray, porphyritic lava flow. For
289
all the early post-caldera units we estimated a volume of 27 km3.
40Ar-39Ar
whole-rock age of 2,441 ± 234 ka for
290
291
Late post-caldera units (~2.0 - ~0.016 Ma) (Lpc)
292
These units are represented by 11 domes (Alrd, Lrd, Amrc, Prld, Crcd, Trcd, Crd, Ard,
293
Crl, Arcd and Mrcd), 5 lava flows (Coal, Tal, Cual, Pdl and Prld), 2 ignimbrites (Eri, Tr1), and 4
294
scoria cones (Plc, Tlc, Clc1 and Clc2). The Alrd, Lrd, Amrc, Crcd, Trcd, Crd, Ard, Crl Arcd and
295
Mrcd dome and lava flow units occur on the caldera border and periphery (Fig. 2). These
296
structures have predominant rhyolitic compositions with coulée and asymmetrical morphology
12
297
delimited by very steep levées. The rocks of these units typically appear as light-gray to pinkish-
298
gray, banded to massive obsidian lavas with mottled structure given by abundant spherulites and
299
lithophysae (Fig. 3). López-Hernández et al. (2009) and García-Tovar et al. (2015) reported K-Ar
300
ages in hornblende and whole-rock of the Lrd (1,700 ± 400 ka), Crd (1,300 ± 600 ka), and Crl
301
(1,274 ± 27 ka). In this work, seven new units were dated with the
302
yielding ages of 1,870 ± 36 ka (Alrd) (Fig. 4J), 1,438 ± 24 ka (Amrc) (Fig. 4K), 1,394 ± 8 ka
303
(Crcd) (Fig. 4L), 1,360 ± 15 ka (Trcd) (Fig. 6A), 1,283 ± 88 ka (Ard) (Fig. 6B), 998 ± 36 to
304
1,145 ± 14 ka (Arcd) (Fig. 6C-D) and 1,066 ± 42 ka (Mrcd) (Fig. 6E), that are in agreement with
305
their stratigraphic position. On the other hand, on the western edge the undifferentiated Maguey
306
unit formed of a sequence of pyroclastic surges and fall with a
307
(Table 1; Fig. 6F).
40Ar/39Ar
40Ar-39Ar
method (Table 1)
age of 1,084 ± 22 ka
308
The Coal, Tal, Cual, Pdl and Prld units are situated inside the caldera partially overlying
309
the syn-caldera and early post-caldera units (Fig. 2). They appear as a stack of bedded lava flows
310
with andesitic (Coal, Tal, and Cual) and rhyolitic (Pdl and Prld) compositions. The andesitic lava
311
flows are porphyritic, black to dark-gray in color, with reddish to yellowish intense hydrothermal
312
alteration (Fig. 3). The Coal unit was dated by Avellán et al. (2018) with the 40Ar-39Ar method at
313
2,027 ± 40 ka. A dike 40Ar-39Ar age that cuts this unit yielded 1,600 ± 35 ka (Table 1; Fig. 6G).
314
In this study we dated the Tal unit with the same method at 1,708 ± 54 ka (Table 1; Fig. 6H).
315
López-Hernández et al. (2009) obtained a 40Ar/39Ar age of 1,600 ± 200 ka for the Cual unit.
316
The rhyolitic lava flows (Pdl and Plrd units) are pinkish-gray to pinkish-white porphyritic
317
rocks. These units are white and highly hydrothermally altered lavas that crop out in the vicinity
318
of Pedernal and Acoculco towns. The lava flows are corroded by hydrothermal fluids along
319
highly vesicular breccia structures, where phenocrysts and matrix have been replaced by
13
320
alteration minerals. López-Hernández et al. (2009) reported two K-Ar ages for these lavas of
321
1,600 ± 100 ka (Pdl), and 1,400 ± 200 ka (Prld).
322
The Encimadas unit (Eri) is a rhyolitic ignimbrite widely exposed on the east-northeast
323
external parts of the ACC. It has a moderately dissected plain that mantles the Zacatlán basaltic
324
plateau and partially covers some of the post-caldera units. Eri is a welded ignimbrite with
325
several flow units that appear as massive, light-gray to white, beds. Each bed consists of matrix-
326
supported fine ash particles with feldspar and quartz phenocrysts. It has an approximated volume
327
of 26 km3. López-Hernández et al. (2009) reported a 40Ar/39Ar age in sanidine of Eri at 1,300 ±
328
200 ka. A similar 40Ar/39Ar age of 1,278 ± 14 ka was obtained in this study (Table 1, plateau age;
329
Fig. 6I).
330
The Tecoloquillo ignimbrite (Tr1) is a rhyolitic ignimbrite widely exposed to the south-
331
southwest parts of the caldera (Fig. 2). South of the caldera, Tr1 partially covers some units
332
belonging to the pre-caldera, syn-caldera and post-caldera. The Tr1 unit consists of two main
333
beds; the lowermost part is massive, monolithologic, brittle and matrix-supported with highly
334
friable pumice fragments embedded in a medium to fine ash matrix. Both pumice and matrix
335
contain bipyramidal quartz and alkali feldspar phenocrysts. The upper part is massive with pink-
336
gray, corroded lava blocks, supported by a crumbly medium ash matrix. The upper part of Tr2 is
337
a rhyolitic dome made of angular light-pink to gray lava blocks. The rock is moderately vesicular,
338
fibrous and porphyritic with quartz, alkali feldspar and amphibole phenocrysts. One pumice
339
sample collected from the basal part of the Tr1 unit yielded a 40Ar/39Ar age in plagioclase of 611
340
± 72 ka (Table 1; Figs. 6J). Other pumice sample that yielded older age of 762 ± 9 ka (Table 1;
341
Fig. 6K). López-Hernández et al. (2009) reported a 40Ar/39Ar age in sanidine of 0.8 ± 0.1 Ma for
342
this unit.
14
343
The last four late-post-caldera units are cinder cones (Paila, Tulimán, Cuatzitzinguito and
344
Cuatzitzingo) and associated lava flows of basaltic andesite composition. These scoria cones
345
occur above or close to the topographic caldera rim (Fig. 2). The Paila (Plc) and Tulimán (Tlc)
346
cinder cones are exposed on the southeastern and northwestern parts of the caldera rim,
347
respectively. The Paila unit directly overlies the Atad and Eri post-caldera units and Tulimán unit
348
lies discordantly on top of Srl post-caldera unit. The Cuatzitzinguito (Clc1) and Cuatzitzingo
349
(Clc2) scoria cones lies on the southern flank of the Paila unit (Fig. 3). All these scoria cones are
350
composed of massive poorly-sorted fallout beds with dense blocks and bread-crust scoria and
351
spatter bombs, as well as, black to dark-gray blocky lava flows associated with effusive activity.
352
The Tulimán scoria cone was dated with the 40Ar/39Ar at 63 ± 9 ka (Avellán et al., 2018). A 40Ar-
353
39Ar
354
underneath a fallout tephra of the Cuatzitzinguito scoria cone yielded a
355
BP (Table 2). We did not obtain an age for the Cuatzitzingo scoria cone, however, its
356
stratigraphic position indicates that it should be younger than 16 ka.
age of 71 ± 17 ka was obtained here for the Paila scoria cone (Table 1; Fig. 6L). A paleosol
14C
age of 16,710 ± 50
We estimated a total volume for all the late-post caldera units including Eri and Trl of 90
357
358
km3.
359
Extra-caldera units (~2.4-0.19 Ma) (ATVF)
360
Nineteen scoria cones and four small-shield volcanoes of the ATVF occur around the
361
caldera complex (Figs. 2). Four of these scoria cones, known as Amanalco (Asc; 2,408 ± 58 ka;
362
Avellan et al., 2018), Huixtepec (Hsc), Tecolote (Tsc) and Apapasco (Asc) are located ca. 7 km to
363
the southeast of the caldera border. Five scoria cones called Buenavista (Bsc), Comal (Csc),
364
Calandria (Csc), Toronjil (Tsc), and Tezontle (Tsc) and the Coatzetzengo small-shield-volcano
365
are located at ca. 4 km to the northwest. Three scoria cones named Moxhuite (Msc; 239 ± 34 ka;
366
Avellán et al., 2018), Matlahuacala (Msc), and Cazares (Csc) lie discordantly on top of the
15
367
Encimadas ignimbrite at ca. 6 km to the east. Two small-shield volcanoes, Camelia (Clc; 2,033 ±
368
84 ka; Avellán et al., 2018) and Tetelas (Tlc, 1,060 ± 84 ka; Avellán et al., 2018) are located at 10
369
km to the south of the caldera border. Three scoria cones are aligned in a NW-SE direction, these
370
are Tecajete (Tsc) (1,235 ± 62 ka), Blanco (Bsc; 1,274 ±62 ka; Avellán et al., 2018) and Hermosa
371
(Hsc). Another, four scoria cones, Coliuca (Clc; 188 ± 6 ka K-Ar age; García-Tovar et al., 2015),
372
Colorado (Csc), El Conejo (Csc) and Tezoyo (Tsc), and the Coyote small-shield volcano are
373
situated at ca. 7 km to the southwest of the caldera border.
374
CHARACTERIZATION OF FAULT SYSTEMS
375
The ACC is in a highly tectonized region affected by NW-SE, NE-SW and E-W
376
structures (Fig. 7A). The NW-SE and NE-SW structures belong to two regional fault systems.
377
The NW-striking (NNW to NW) fault system contains the oldest regional structures. Some of
378
these faults follow the trend of older structures, such as fold axes and thrust faults from the
379
Laramide orogeny in the Sierra Madre Oriental (Suter, 1984; Rocha et al., 2006; Lermo et al.,
380
2009), exposed just north of the study area. The NE-oriented Cenozoic extension (Henry and
381
Aranda-Gómez, 1992) was active in this region from middle to late Miocene creating NW-
382
striking normal faults and NE-striking strike-slip faults synchronous with regional volcanism
383
(Andreani et al., 2008). In the ACC, the NW-SE fault system is represented by several major
384
normal and oblique (right-lateral component) faults and numerous minor fault strands with
385
lengths between 2 and 5 km, with an average azimuth of 130˚ dipping mainly to the NE at
386
average angles of 50˚, creating horst and graben-like structures (García-Palomo et al., 2002; Fig.
387
7A). The Manzanito structure, also known as the Tulancingo-Tlaxco fault system (López-
388
Hernández et al., 2009), is the most representative fault pertaining to this system in the ACC
389
(Avellán et al., 2018). It has an en échelon array of normal faults, which extends for ca. 30 km
390
(Fig. 6B) and displaces the 1.7 Ma Lobera rhyolitic dome by at least 145 m, and the 1.07 Ma
16
391
Minilla rhyolitic dome by 120 m (this study). Also, at around 1.6 Ma, the ~160˚-striking and ~2.5
392
km long Colorada basaltic dikes were emplaced at the southeastern part of the caldera (Fig. 2;
393
Coal; Avellán et al., 2018). But there is no evidence of recent (Holocene) activity in the local
394
NW-striking faults.
395
The NE-striking (NNE to NE) Apan-Tlaloc Fault System (García-Palomo et al., 2018),
396
Tenochtitlan-Apan Fault System (García-Palomo et al., 2002) or Tenochtitlan Shear Zone (De
397
Cserna et al., 1988), has been active since the Miocene and is still active (García-Palomo et al.,
398
2002; 2018; Fig.7). This is the most important fault system in the region, consisting of normal
399
faults with a left-lateral component with an average azimuth of 040˚ dipping both to the NW and
400
SE with an average dip angle of 75˚ that present two generations of striae (~30 and 80°). This
401
system has created regional horst and graben-like structures, obeying a NW-oriented extension
402
regime that affects eastern México (this study and García-Palomo et al., 2018; Fig. 7A). In
403
general, this fault system shows a good geomorphic expression represented by several major
404
faults (e.g. Apan-Tlaloc Fault and Chignahuapan Fault) and numerous minor fault strands and
405
fractures with 1-4 km individual lengths. Field exposures show prominent scarps; the ~1.3 Ma
406
Canoas rhyolitic dome is displaced by the Atotonilco fault (northern caldera rim) by 150 m, and
407
the ~2.2 Ma Ajolotla trachyandesitic dome is displaced by the Chignahuapan fault by 200 m (this
408
study). The most recent volcanic structures of the ACC (Plc with 71 ka, and Tlc with 63 ka units)
409
are cut by these faults (this study; Figs. 2 and 7B).
410
The NE- and NW-striking normal fault systems intersect each other (García-Palomo et al.,
411
2002; Lermo et al., 2009) creating an orthogonal arrangement of grabens, half-grabens and
412
horsts. The NE-SW Rosario-Acoculco Horst (García-Palomo et al., 2002) is delimited to the west
413
by the 235°-trending and NW-dipping Apan-Tlaloc Fault (Mooser and Ramírez, 1987; Huizar-
414
Álvarez et al., 1997) and to the east by the 055°-trending and SE-dipping Chignahuapan Fault
17
415
(Avellán et al., 2018). The ACC is a volcano-tectonic depression inside the Rosario-Acoculco
416
Horst, delimited by parallel faults but with opposite dipping. As for the E-W structures, they are
417
represented by at least 10 E-W striking (~N085°) and SE-dipping normal fault strands. They can
418
only be observed inside the ACC, mainly affecting post-caldera volcanism in the central part of
419
the ACC (this study; Figs. 2 and 7).
420
AEROMAGNETIC DATA
421
We produced an aeromagnetic map by reprocessing the airborne data of the Mexican
422
Geological Survey obtained in 2000 (Fig. 8A). This map shows different aeromagnetic anomalies
423
inside or outside of the ACC. The outer domains may be caused by regional and topographic
424
anomalies. Instead, inner domains are possibly associated to four positive local anomalies
425
(subdomains) and conforming a semi-circular shape of the Acoculco caldera. The first subdomain
426
is located in the central part of the ACC; it shows an NE-oriented elongated shape (9.7 km long
427
and 4.8 km wide) with magnetic intensities between -81.8 and 57.8 nT. The second anomaly is
428
located at the southeastern boundary of the ACC, ENE-oriented, 6.6 km long and 4.1 km wide
429
with magnetic intensities between -135.1 and 25.7 nT. The third anomaly is located in the south-
430
central portion of the ACC, NE-oriented, 8.9 km long and 3.5 km wide with magnetic intensities
431
between -64.5 and 36.5 nT. The fourth anomaly appears in the western part of the ACC, NW-
432
oriented, 5.4-km long and 3-km wide with magnetic intensities between 84.5, and -2 nT. The map
433
shows a contrast between the caldera (higher magnetic values) and the surrounding areas (lower
434
magnetic values).
435
DISCUSSION
436
Tectonic implications
18
437
The regional tectonic setting has controlled the formation and evolution of the ACC. The
438
ACC formed on top of the Rosario-Acoculco horst, which is bounded by the Apan and Tlaxco-
439
Chignahuapan grabens (Fig.7). Furthermore, some pre-existing tectonic faults were used as
440
pathways for dike intrusions and volcanism (aligned scoria cones parallel to the NE-trending fault
441
system), but also as part of the ring fault system that controlled the caldera collapse and exerted
442
the main control on the location of post-caldera vents (Fig. 2). The southwestern caldera rim
443
coincides with the NW-SE Manzanito fault, while the northern caldera rim is marked by the
444
Atotonilco scarp (Avellán et al., 2018). The sub-linear to sub-circular collapsed structures form a
445
18 x 16 km rhombohedral-shape, dipping towards the central part of the ACC. In fact, the left-
446
lateral movement in the NE-SW faults during the middle Miocene (García-Palomo et al., 2000;
447
2018) and/or the right-lateral movement of the NW-SE faults could have been responsible for the
448
fracturing and creation of the space necessary to accommodate the magma reservoir beneath the
449
caldera (pull-apart basins in transtensional regimes; e.g., Bursik, 2009; Saxby et al., 2016).
450
Activity of the NE-SW and NW-SE regional systems continued after the caldera formation and
451
modified the trace of the caldera border, causing its displacement at several points (e.g.,
452
northeastern tip of the topographic caldera rim, outcrops 27 and 28; Fig. 7B). Tectonic
453
movements affected the interior of the caldera until very recent time, causing displacements of
454
intra-caldera blocks and disturbing the position of intra-caldera volcanic vents and products. For
455
example, E-W faulting in the central part of the caldera (outcrops 24, 32 and 54; Fig. 7B)
456
Caldera reactivation during the emission of the 26 km3 Encimadas (~1.28 Ma) and 11 km3
457
of the Tecoloquillo ignimbrites (~0.6-0.76 Ma) was possibly originated near or along the E and
458
SW borders of the caldera, respectively. Therefore, it would not be surprising if the caldera used
459
these faults systems to nucleate the rest of the ring fault that controlled the collapse event, as it
460
has occurred in other well documented calderas (see Aguirre-Díaz et al., 2008; Martí et al., 2013;
19
461
Molina et al., 2014). Despite the strong, selective erosion that has affected parts of the area, the
462
northern topographic caldera rim has not retreated significantly from its original position. This
463
may imply that caldera subsidence continued for a long time (±500 ka, between Encimadas and
464
Tecoloquillo eruptive events), dissecting younger rocks emplaced close or through the ring faults.
465
On the contrary, in the eastern side of the ACC, neither the morphological nor the structural
466
border are visible at surface, however, it is depicted in the aeromagnetic map (Fig. 8A). At the
467
surface, the eastern border is not visible because it has been buried by younger deposits and it is
468
also likely an uneven collapse of the caldera. The E-W intra-caldera fault system is limited to the
469
caldera interior and therefore is could represent local deformation related to the resurgence of the
470
caldera.
471
The NE-SW and NW-SE fault systems coexist in the ACC (Fig. 7B). This can be
472
explained by two different constructive phases. The first one took place in the Miocene, ruled by
473
~NE-oriented extension, forming ~NW-SE dip-slip faults and ~NE-SW sinistral strike-slip faults
474
(Henry and Aranda-Gómez, 1992). The second phase took place in the Pliocene-Pleistocene, and
475
it is controlled by ~NW-oriented tectonic extension, forming ~NW-SE dextral strike-slip faults
476
and ~NE-SW dip-slip faults (oblique and normal faults with a minor left-lateral component;
477
García-Palomo et al., 2018). This is consistent with the left-stepping en échelon geometry of the
478
NW-SE Manzanito fault originated by right-lateral slip. Actually, this latter phase of NW-
479
oriented extension is still active (García-Palomo et al., 2018). Therefore, both fault systems occur
480
in the same region and are still active under the same stress regime. This is possible because the
481
NW-SE structures are acting as transfer faults of the NE-SW normal faults. In order to
482
demonstrate that both fault systems are moving under the same strain field, a synthetic right
483
dihedra diagram was created for the two pairs of conjugate fault sets, according to the mean
484
measured planes (Fig. 7B). This slip model implies an orthorhombic symmetry of faults, where
20
485
superposed areas divide the figure into the different compressional and tensional areas in the
486
ACC.
487
Aeromagnetic interpretation and geothermal implications
488
Previous studies used the aeromagnetic and gravimetric data of the Acoculco-Atotonilco
489
region obtained by Petroleos Mexicanos (PEMEX) in 1980 (García-Estrada, 2000; López-
490
Hernández et al., 2009). These authors interpreted a depression bounded by faults coincident with
491
gravimetric data and low magnetic values with much higher values at the interior defining the
492
border of the Acoculco caldera. This coincides with our interpretation of the airborne data
493
displayed in the aeromagnetic map of figure 8A. This map shows an approximated match
494
between the edge of the 18 x 16 km Acoculco caldera as reconstructed from the distribution of
495
the Acoculco ignimbrite, the co-ignimbritic breccia, the Atotonilco scarp, and the Manzanito fault
496
(Avellán et al, 2018; this work) (Fig. 7). The aeromagnetic anomalies suggest a rough elliptical
497
limit of the caldera that may be bounded to the east by the approximated venting location of the
498
Encimadas ignimbrite and to the southeast by the basaltic andesite Paila scoria cone. The positive
499
aeromagnetic anomaly located at the NE portion of the ACC may be produced by the Zacatlán
500
basaltic plateau that is overlaid by the Encimadas ignimbrite and both cover the Cretaceous
501
limestones (Fig. 2). The positive anomalies located to the SW and NW of the ACC may be
502
associated with intrusive bodies that gave place to monogenetic volcanoes of the ATVF. The
503
interior of the caldera is not homogeneous but includes higher and lower magnetic values that
504
might reveal the presence of small-scale horsts and grabens combined with the presence of
505
shallow intrusive bodies (i.e., laccoliths, sills) of basaltic and basaltic and intermediate
506
composition intruded inside the Cretaceous limestones at depths (>1 km; Fig. 8B). These positive
507
anomalies inside the caldera could have fed the ca. 2.6 Ma Huistongo, 2.2 Ma Manzanito, 0.71
508
Ma Paila and <16 ka Cuatzitzingo eruptions (Figs. 2, 3 and 7). The anomalies described in figure
21
509
8A have rough NE-SW orientations separating a local horst on top of which the CFE exploratory
510
wells were drilled, cutting skarns at ~790 m deep (EAC-1), and argilitic limestones at 350 m deep
511
(EAC-2) (Viggiano-Guerra et al., 2011). These positive magnetic anomalies might represent
512
intrusive bodies injected during the post-caldera phase (Huistongo lavas 2.6 Ma), which may
513
have an important role to keep the geothermal system active. Sosa-Ceballos et al. (2018)
514
concluded that after the caldera collapse some 2.7 Ma ago, the stress field on the ACC magma
515
chamber and its immediate surroundings was modified. Such change hindered magma migration
516
through the collapsed reservoir and promoted lateral tension zones where dikes and sills converge
517
and eventually form new magma chambers. Due to the active tectonics in the Acoculco area, a
518
new plumbing system developed after the caldera collapse and gradually favored the ascent of
519
deep seated, independent peralkaline and calcalkaline mafic magmas. In addition, Sosa-Ceballos
520
et al. (2018) found that magma mixing-heating is the main magmatic process that modified the
521
ACC rock suite. Hence, the aeromagnetic anomalies might represent the horizontal intrusion
522
zones where mafic magma accumulates as a consequence of the upper crustal deformation
523
produced by the collapse. These intrusion zones serve as heating elements of magma reservoirs
524
(that eventually might evolve into small magma chambers) and yield the energy that sustain the
525
Acoculco geothermal system. When magma did not migrate horizontally, it flowed upwards,
526
sometimes getting tapped on its way to the surface evolving to silicic compositions es got trapped
527
as the aplitic dike dated at 183 ± 36 ka found in EAC-1 (Table 1). These young felsic intrusions
528
have reactivated the system providing heat for the hydrothermal activity and Holocene
529
hydrothermal explosions (Canet et al., 2015a and b). López-Hernández et al. (2009) also
530
suggested the possible presence of an intrusive body, whose top is at > 1000 m depth and that
531
was not reached by EAC-1 well. Although, we believe these intrusions also contribute to the
22
532
geothermal system, their volume and abundance are not yet known and probably can be estimated
533
by geophysical methods to constraint their energy input.
534
The distribution of magnetic anomalies inside the caldera also reveals the presence of
535
tectonic lineaments with NE-SW and NW-SE orientations, likely associated to horst and grabens
536
blocks and coinciding with those intra-caldera and caldera ring faults observed and measured in
537
the field, suggesting that such faults were used as vent zones for post-caldera volcanism (Figs. 7
538
and 8). These faults may have also favored the ascent of gas with isotopic compositions (N2/He,
539
3He/4He, 13C, 15N)
540
of CO2 and 3He/4He ratio (R/Rair = 8.5), that suggest an active magmatic source at depth under
541
the ACC (Polak et al., 1982).
542
Volcanic evolution
of both MORB- and arc-type signatures (Peiffer et al., 2014), and high values
543
Based in our synthetic map, refined volcanic stratigraphy, structural analysis, and
544
interpretation of the aeromagnetic anomalies ACC was emplaced through the Cretaceous
545
limestones, the Peñuela dacitic domes (~13 to 11 Ma), and the Zacatlán basaltic plateau (Figs. 2
546
and 3). Rocks of the ACC were divided into 30 units formed in different stages of the caldera
547
evolution from pre, syn, early-post, late- post caldera and extra-caldera volcanism of the Apan-
548
Tezontepec Volcanic Field (Avellán et al., 2018; Sosa-Ceballos et al., 2018).
549
Caldera formation (~2.7 Ma)
550
Prior to the formation of the caldera, the Puente and Terrerillos lava domes had been
551
extruded in the area (Fig. 9A). These domes of andesitic and dacitic composition and calc-
552
alkaline affinity were emplaced during the Pliocene between 3.9 and 3 Ma. After a few hundred
553
thousand years (0.3 Ma), a large amount of andesitic magma stagnated at depth preparing for a
554
major eruption. This calc-alkaline magma overpressured the encasing rock initiating the emission
23
555
of an andesitic pyroclastic density current (Acoculco ignimbrite). The continuous emission of the
556
ignimbrite eventually diminished the magmatic pressure and weakened the roof of the magma
557
chamber triggering its collapse forming a 18 x 16 km asymmetric caldera (Fig. 9B). The collapse
558
followed older structures as the NW-SE Manzanito fault in the western and southwestern parts
559
and generated a semi-circular caldera ring (Atotonilco scarp) at the north (Fig. 7B). The
560
ignimbrite crops out in the western-southwestern and northern parts of the caldera interior
561
because in other locations it was covered by younger deposits. The used of older and fractures as
562
magma feeders facilitated a high discharge rate at the beginning of the eruption as discussed by
563
Costa and Martí (2016). This developed into massive proportions, thus precluding the formation
564
of a vertical eruption column and the deposition of fallout deposits, which have not been
565
recognized neither outside the caldera nor in the intra-caldera wells. The occurrence of a co-
566
ignimbrite breccia found in relative small, isolated areas at the north, west and southwestern parts
567
of the caldera rim points to emissions sites of the ignimbrite. In fact, the ring fault scarps and the
568
topographic caldera rim are clearly visible at the western half of the caldera, but they are hidden
569
at the eastern side, suggest that caldera-collapse could have been more intense in the western
570
sector than in the eastern sector. We cannot inform on how the caldera-forming eruption
571
progressed, as most if its corresponding deposits have been eroded outside the caldera or are now
572
totally hidden by post-caldera rocks at the interior of the caldera depression. However, the
573
presence of a large number of normal faults with different orientations affecting intra-caldera
574
rocks suggests that caldera-collapse could have occurred in a piecemeal-trapdoor fashion through
575
several stages of volcanism.
576
Early-post caldera phase (2.6-2.2 Ma)
577
After the caldera formation followed a relatively short (~0.1 Ma) quiescence period
578
during which an intra-caldera shallow lake developed with lacustrine sedimentation (Fig. 5B).
24
579
The early post-caldera volcanism is represented by eruptions that occurred mainly in the interior
580
of the caldera with a minimum total volume of 27 km3 (Fig. 9C). This volcanism vented along the
581
ring faults and intra-caldera normal faults that formed or acted during caldera collapse to then be
582
reactivated by the extensional tectonics that have affected the area until present (García-Palomo
583
et al., 2018). These modifications of the local stress field allowed the ascent of peralkaline
584
basaltic and basaltic trachyandesite magmas that were generated by partial melting of a
585
metasomatized mantle, genetically unrelated to the calc-alkaline magmas that dominated the pre-
586
caldera and caldera activities (Sosa-Ceballos et al., 2018). Both suites of magma mixed and
587
formed the early post-collapse dome complexes on the ring of the caldera.
588
Late-post caldera phase (~2.0 - ~0.016 Ma)
589
After a quiescent period of around 0.2 Ma the caldera entered a new phase of volcanism
590
along the caldera rim dominated by the emission of domes and lava flows, two ignimbrites, and
591
four scoria cones (Fig. 9D). An extended period of intermittent volcanism occurred between 2
592
and 1.3 Ma with the occurrence of at least 13 effusive eruptions along the caldera rim. Between 2
593
and 1.6 Ma eruptions emitted bimodal andesitic (Coal, Tal, Cual units) and rhyolitic (Alrd, Lrd
594
units) products to then erupted rhyolitic effusive magmas until 1.28 Ma (Ard unit) at the
595
southwestern part of the caldera rim. At about (1.27 Ma) a rhyolitic explosion occurred at the
596
eastern part of the caldera dispersing to the east and northeast the Encimadas ignimbrite (26 km3).
597
We assume that magmatic overpressure in a shallow magma chamber triggered the rhyolitic
598
eruption that represents the second largest event of the caldera. Immediately after this major
599
eruption occurred four rhyolitic eruptions in the S-SW parts of the caldera between 1.27 and 1.06
600
Ma. These eruptions particularly emitted the voluminous Ailitla coulée dome to the south (Arcd;
601
12.4 km3), the Cabezas lava to the southwest (Crl), a minor explosive event that dispersed small-
602
volume pyroclastic density currents (Msf), and the Minilla coulée come to the west (Mrcd). These
25
603
effusive eruptions preceded the occurrence of another explosive rhyolitic event that occurred
604
some 0.2 Ma later. This eruption vented at the southwestern edge of the caldera rim (0.6-0.76
605
Ma) at the intersection with the NW-SE Manzanito fault. This eruption dispersed the
606
Tecoloquillo rhyolitic ignimbrite (11 km3) to the southwest of the caldera rim ending with the
607
extrusion of a rhyolitic summit dome. The Tecoloquillo ignimbrite represents the third largest
608
explosive event of the ACC after which activity associated to the caldera apparently declined for
609
a period of circa 0.5 Ma. However, renewed activity of the caldera started 0.071 Ma with the
610
emplacement of the La Paila (Plc) at the southeast, the Tulimán scoria cone and lava flow (0.063
611
Ma) at the northwest, and the Cuatzitzinguito (16 ka) and Cuatzitzingo ca. 0.01 Ma scoria cones
612
at the southeast. These magmas erupted basaltic andesite to basaltic trachyandesite products
613
(Sosa-Ceballos et al., 2018).
614
During this late post-caldera period the peralkaline magma suite gradually dominated
615
over the calc-alkaline suite (Sosa-Ceballos et al., 2018) (Figs. 9C-D). This process resulted in an
616
undistinguishable unique source for the early post-collapse magmas, whereas trace elements
617
geochemistry suggests a relatively homogeneous source for the late post-collapse rhyolites (Sosa-
618
Ceballos et al., 2018). The total volume of magma erupted during late post-caldera phase was
619
~90 km3. We believe that during this phase of the caldera evolution began the resurgence of the
620
caldera floor as suggested by uplifted and tilted lacustrine sediments (250°/30SE) exposed inside
621
the caldera (Fig. 2) covering the Acoculco ignimbrite. The morphology of the caldera floor
622
suggests that the resurgence was more important in the southwestern portion, which is in
623
accordance with shallow intrusions of magma revealed by aeromagnetic data that could cause the
624
caldera resurgence. According to Sosa-Ceballos et al. (2018) these shallow magma intrusions
625
were originated by post-caldera deformation that promoted the formation of sills and dykes above
626
the collapsed reservoir and might be the heat source of the Acoculco geothermal system.
26
627
Finally, the area covered by the AAC products is ca. 856 km2 with a total minimum
628
volume ca. 143 km3. By assuming that the present extension of deposits is minimum, we
629
calculated that activity was dominated by nearly 80 % of effusive volcanism with three major
630
bursts of explosive volcanism that represent 20% of the ACC. As summarized above Acoculco
631
has been a site of persistent volcanism since 2.7 Ma that include the presence of a young
632
intrusion (183 ± 36 ka) at the bottom of the EAC-1 well. This intense magmatism, as well as
633
Holocene hydrothermal explosions (Canet et al., 2015a; 2015b), and geothermal manifestations
634
(Peiffer et al., 2014) indicate that the complex is still active and could represent a site to develop
635
a geothermal field.
636
CONCLUSIONS
637
The Acoculco collapse caldera was originated at about ~2.7 Ma ago in response to the
638
eruption of 127 km3 of the andesitic ignimbrite of the Acoculco unit. The caldera collapse episode
639
occurred in part along a pre-existing regional NE-SW trending faults (southwestern, western and
640
northwestern sectors), but also along newly formed ring faults (northern sector and probably the
641
eastern sector, but this is not visible). The collapse occurred in a piecemeal-trapdoor fashion, in
642
which intra-caldera blocks bounded by pre-existing or newly formed normal faults gravitationally
643
collapsed in a partly emptied magma chamber. After the formation of the caldera, volcanic
644
activity stopped for a while permitting the formation of an intra-caldera lake as suggested by
645
sediments covering part of the caldera-forming ignimbrite. Renewed volcanic activity emplaced
646
several domes and lava flows along the caldera borders and intra-caldera faults that modified the
647
caldera depression. Post-caldera volcanism also affected the areas around the caldera and was
648
controlled by the main regional fault systems that also influenced the formation of the caldera.
649
The ACC was formed at the intersection of NE-SW and NW-SE fault systems. These fault
27
650
systems have controlled the formation and evolution of the ACC, the regional subsidence, and
651
venting of syn- and post-caldera volcanism inside and outside the caldera.
652
The presence of positive aeromagnetic anomalies associated to intrusive bodies (sills and
653
dykes) at the interior of the Acoculco caldera makes it an interesting target for geothermal
654
exploration.
655
ACKNOWLEDGEMENTS
656
This study was partially funded by the Centro Mexicano de Innovación en Energía
657
Geotérmica (CeMIE-Geo) project P15 and GeMex 4.4. to J.L. Macías. We thank F. Mendiola, G.
658
Reyes-Agustín and S. Cardona for their technical support during the laboratory analyses and D.E.
659
Torres-Gaytan for his support during the aeromagnetic map generation. We appreciate the
660
discussions and input provided by C. Arango, C. Canet and J. Marti is grateful for the MECD
661
(PRX16/00056) grant. We appreciated the constructive comments by J. Aranda and two
662
anonymous reviewers.
663
REFERENCES CITED
664
Afanasyev, A., Costa, A., Chiodini, G., 2015. Investigation of hydrothermal activity at Campi
665
Flegrei caldera using 3D numerical simulations: Extension to high temperature
666
processes. Journal of Volcanology and Geothermal Research, 299, 68-77.
667
Aguirre-Dı́az, G.J., López-Martı́nez, M., 2001. The Amazcala caldera, Queretaro, Mexico.
668
Geology and geochronology. Journal of Volcanology and Geothermal Research, 111(1-4),
669
203-218.
670
Aguirre-Díaz, G.J., Labarthe-Hernández, G., Tristán-González, M., Nieto-Obregón, J., Gutiérrez-
671
Palomares, I., 2008. The ignimbrite flare-up and graben caldera of the Sierra Madre
672
Occidental, México. In: Gottsmann J, Martí J. (Eds.), Caldera volcanism: analysis, modeling
673
and response. Elsevier, Amsterdam, 143–174.
674
Alatriste-Vilchis, D., Campos-Enriquez, O., Huizar-Alvarez, R., Marines-Campos, R., 2005. La
675
estructura sub-superficial de la subcuenca de Tecocomulco. La Laguna de Tecocomulco Geo-
28
676
Ecologia de un desastre. Universidad Nacional Autónoma de México, Instituto de Geología,
677
Publicación especial, 3, 33-48.
678
Angelier, J., 1990. Inversion of field data in fault tectonics to obtain the regional stress—III. A
679
new rapid direct inversion method by analytical means. Geophysical Journal International, 103
680
(2), 363-376.
681
Avellán, D.R., Macías, J.L., Layer, P.W., Cisneros, G, Sánchez-Núñez, J.M., Gómez-
682
Vasconcelos, M.G., Pola, A., Sosa-Ceballos, G., García-Tenorio, F., Reyes-Agustín, G.,
683
Osorio-Ocampo, S., García-Sánchez, L., Mendiola, I.F., Marti, J., López-Loera H., Benowitz,
684
J., 2018. Geology of the late Pliocene – Pleistocene Acoculco caldera complex, eastern Trans-
685
Mexican Volcanic Belt (México). Journal of Maps, 15(2), 8-18.
686
Bibby, H.M., Caldwell, T.G., Davey, F.J., Webb, T.H., 1995. Geophysical evidence on the
687
structure of the Taupo Volcanic Zone and its hydrothermal circulation. Journal of
688
Volcanology and Geothermal Research, 68 (1-3), 29-58.
689
690
Bursik, M., 2009. A general model for tectonic control of magmatism: Examples from Long
Valley Caldera (USA) and El Chichón (México). Geofísica internacional, 48(1), 171-183.
691
Caballero, M., Lozano, S., Ortega, B., Urrutia, J., Macias, J.L., 1999. Environmental
692
characteristics of Lake Tecocomulco, northern basin of México, for the last 50,000
693
years. Journal of Paleolimnology, 22 (4), 399-411.
694
Calcagno, P., Evanno, G., Trumpy, E., Gutiérrez-Negrín, L.C., Macías, J.L., Carrasco-Núñez, G.,
695
Liotta, D., the GEMex T3.1 team, 2018. Preliminary 3-D geological models of Los Humeros
696
and Acoculco geothermal fields (Mexico) – H2020 GEMex Project. Advances in Geosciences,
697
1, 1-13.
698
Campos-Enríquez, J.O., Alatriste-Vilchis, D.R., Huizar-Álvarez, R., Marines-Campos, R.,
699
Alatorre-Zamora, M.A., 2003. Subsurface structure of the Tecocomulco sub-basin
700
(northeastern México basin), and its relationship to regional tectonics. Geofísica Internacional,
701
42 (1), 3-24.
702
Canet, C., Hernández-Cruz, B., Jiménez-Franco, A., Pi, T., Peláez, B., Villanueva-Estrada, R.E.,
703
Salinas, S., 2015a. Combining ammonium mapping and short-wave infrared (SWIR)
704
reflectance spectroscopy to constrain a model of hydrothermal alteration for the Acoculco
705
geothermal zone, Eastern México. Geothermics, 53, 154-165.
29
706
Canet, C., Trillaud, F., Prol-Ledesma, R.M., González-Hernández, G., Peláez, B., Hernández-
707
Cruz, B., Sánchez-Córdova, M.M., 2015b. Thermal history of the Acoculco geothermal
708
system, eastern México: Insights from numerical modeling and radiocarbon dating. Journal of
709
Volcanology and Geothermal Research, 305, 56-62.
710
711
712
713
Castro-García, A., Córdoba, D., 1994. Estratigrafía del área volcánica de Tulancingo, Hidalgo,
México. Revista de Investigación, Universidad Autónoma del Estado de Hidalgo, 1, 26-39.
Costa, A., Martí, J., 2016. Stress field control during large caldera-forming eruptions. Frontiers
in Earth Science, 4 (92), 1-13.
714
De Cserna, Z.D., De la Fuente-Duch, M., Palacios-Nieto, M., Triay, L., Mitre-Salazar, L. M.,
715
Mota-Palomino, R., 1988. Estructura geológica, gravimetría, sismicidad y relaciones
716
neotectónicas regionales de la cuenca de México. Boletín del Instituto de Geología,
717
Universidad Nacional Autónoma de México, 104, 71.
718
719
De la Cruz-Martínez, V., Castillo-Hernández, D., 1986. Geología de la zona geotérmica de la
caldera de Acoculco, Puebla. Geotermia, Revista Mexicana de Geoenergía, 2 (3), 245-254.
720
Di Napoli, R., Martorana, R., Orsi, G., Aiuppa, A., Camarda, M., De Gregorio, S., Candela E.G.,
721
Luzio D., Messina, N., Pecoraino, G., Bitetto, M., de Vita, S., Valenza M., 2011. The structure
722
of a hydrothermal system from an integrated geochemical, geophysical, and geological
723
approach: The Ischia Island case study. Geochemistry, Geophysics, Geosystems, 12 (7), 1-25.
724
De Vicente, G., Muñoz, A., Giner, J.L., 1992. El uso del método de los diedros rectos:
725
implicaciones a partir del modelo de deslizamiento del análisis poblacional de fallas. Revista
726
Sociedad Geológica de España, 5, 3-4.
727
Ellis, R.G., De Wet, B., MacLeod, I.N., 2012. Inversion of magmatic data for remanent and
728
induced sources: Auxiliary Information in Geophysical Inversion. Presented at the ASEG
729
Inversion Forum, extended abstract.
730
731
García-Estrada, G.H., 2000. Gravimetría y magnetometría regional del proyecto Acoculco,
Puebla. CFE-GPG internal report DEX-ACO19/2000, 57.
732
García-Palomo, A., Macías, J.L., Garduño, V.H., 2000. Miocene to Recent structural evolution
733
of the Nevado de Toluca volcano region, central México. Tectonophysics, 318 (1), 281-302.
734
García-Palomo, A., Macías, J.L., Tolson, G., Valdez, G., Mora, J.C., 2002. Volcanic stratigraphy
735
and geological evolution of the Apan región, east-central sector of the Trans-Mexica Volcanic
736
Belt. Geofísica Internacional, 41 (2), 133-150.
30
737
García-Palomo, A., Macías, J.L., Jiménez, A., Tolson, G., Mena, M., Sánchez-Núñez, J.M.,
738
Arce, J.L, Layer, P.W., Santoyo, M.A., Lermo-Samaniego, J., 2018. NW-SE Pliocene-
739
Quaternary extension in the Apan-Acoculco region, eastern Trans-Mexican Volcanic Belt.
740
Journal of Volcanology and Geothermal Research, 349, 240-255.
741
García-Sánchez, L., Macías J.L., Arce, J.L., Sosa-Ceballos, G., Garduño-Monroy, V.H.,
742
Saucedo, R., Avellán, D.R., Rangel, E., Layer, P.W. Lopez-Loera, H., Rocha, V.S., Cisneros,
743
G., Reyes-Agustín; G., Jiménez, A., 2016. Genesis and evolution of the Cerro Prieto Volcanic
744
Complex, Baja California, Mexico. Bulletin of Volcanology, 79 (44), 1-28.
745
García-Tovar, G.P., Martínez-Serrano, R.G., Solé, J., Correa-Tello, J.C., Nuñez-Castillo, E.Y.,
746
Guilluo, H., Monroy-Rodríguez, E., 2015. Geología, geocronología y geoquímica del
747
vulcanismo Plio-Cuaternario del Campo Volcánico Apan-Tecocomulco, Faja Volcánica
748
Trans-Mexicana. Revista Mexicana de Ciencias Geológicas, 32 (1), 100-122.
749
Geyer, A., Martí, J., 2014. A short review of our current understanding of the development of
750
ring faults during collapse caldera formation. Frontiers in Earth Science, Volcanology, 2, 1-22.
751
Henry, C.D., Aranda-Gomez, J.J., 1992. The real southern Basin and Range: Mid-to late
752
Cenozoic extension in México. Geology, 20 (8), 701-704.
753
Hernández-Javier, I., 2008. Geología y geomorfología volcánica de la región de los yacimientos
754
de obsidiana en Otumba en el sector norte de la Sierra Nevada de México. Bacherlor thesis,
755
Universidad Nacional Autónoma de México.
756
Hiriart, G., Gutiérrez, N. L., Quijano, L.L., Ornelas, C.A., Espíndola, S., Hernández, I., 2011.
757
Evaluación de la energía geotérmica en México. Informe para el Banco Interamericano de
758
Desarrollo y la Comisión Reguladora de Energía, México.
759
Huízar-Álvarez, R., Campos-Enríquez, J.O., Lermo-Samaniego, J., Delgado-Rodríguez, O.,
760
Huidobro-González, A., 1997. Geophysical and hydrogeological characterization of the sub-
761
basins of Apan and Tochac (México Basin). Geofísica Internacional, 36, 217-234.
762
Izquierdo, G., Aragón, A., Portugal, E., Arellano, V.M., De León, J., Álvarez, J., 2006.
763
Mineralogía de la zona mineralizada de sílice-epidota (ZMSE) del yacimiento geotérmico de
764
Cerro Prieto, B. C., México. Geotermia, Revista Mexicana de Geoenergia, 19, 2-12.
765
Lorenzo-Pulido, C., Armenta-Flores, M., Ramírez-Silva, G., 2010. Characterization of the
766
Acoculco Geothermal Zone as a HDR System. Geothermal Resources Council Transactions,
767
34, 369–372.
31
768
769
770
771
Johnson, C.A., Harrison, C.G.A., 1990. Neotectonics in central México. Physics of the Earth and
Planetary Interiors, 64, 187-210.
Layer, P.W., 2000. Argon-40/argon-39 age of the El’gygytgyn impactevent, Chukotka, Russia.
Meteorite Planetary Science, 35, 591–599.
772
Layer, P.W., García-Palomo, A., Jones, D., Macías, J.L., Arce, J.L., Mora, J.C., 2009. El
773
Chichón volcanic complex, Chiapas, México: stages of evolution based on field mapping and
774
40Ar/39Ar geochronology. Geofisica Internacional, 48, 33–54.
775
Ledezma-Guerrero, O., 1987. Hoja Calpulalpan 14Q-h (3), con Resumen de la geología de la
776
hoja Calpulalpan, Estados de México, Puebla, Tlaxcala e Hidalgo. Quetzalapa region, southern
777
México. Revista Mexicana de Ciencias Geológicas, 15, 167-180.
778
Lermo, J., Antayhua, Y., Bernal, I., Venegas, S., Arredondo, J., 2009. Monitoreo sísmico en la
779
zona geotérmica de Acoculco, Puebla, México. Geotermia, Revista Mexicana de Geoenergía,
780
22 (1), 40-58.
781
782
Lipman, P.W., 2000. Calderas. In: Sigurdsson, H., Houghton, B.F., McNutt, S.R., Rymer, H.,
Stix, J. (Eds.), Encyclopedia of Volcanoes. Academic press, San Diego, 331-343.
783
López-Hernández, A., 2009. Evolución volcánica del complejo Tulancingo-Acoculco y su
784
sistema hidrotermal, Estados de hidalgo y Puebla, México. Ph.D. thesis, Universidad Nacional
785
Autónoma de México.
786
Lopez-Hernandez, A., Castillo-Hernandez, D., 1997. Exploratory drilling at Acoculco, Puebla,
787
México: a hydrothermal system with only nonthermal manifestations (No. CONF-971048).
788
Geothermal Resources Council, Davis, CA (United States).
789
López-Hernández, A., García-Estrada, G., Aguirre-Díaz, G., González-Partida, E., Palma-
790
Guzmán, H., Quijano-León, J., 2009. Hydrothermal activity in the Tulancingo-Acoculco
791
Caldera Complex, central México - Exploratory studies. Geothermics, 38, 279-293.
792
López-Hernández, A., Martínez, E.I., 1996. Evaluación volcanológica y estructural de la zona
793
geotérmica de Acoculco, Puebla, y su relación con la anomalía termal detectada en el pozo
794
EAC-1. CFE-GPG Internal Report OGL-AC-11/96.
795
Macías, J.L., Arce, J.L., García-Tenorio, F., Layer, P.W., Rueda, H., Reyes-Agustin, G.,
796
Avellán, D., 2012. Geology and geochronology of Tlaloc, Telapón, Iztaccíhuatl, and
797
Popocatépetl volcanoes, Sierra Nevada, central México. Field Guides, 25, 163-193.
32
798
799
800
801
802
803
Martí, J., Geyer, A., Folch, A., Gottsmann, J., 2008. A review on collapse caldera
modelling. Developments in Volcanology, 10, 233-283.
Martí, J., Geyer, A., Aguirre-Diaz, G., 2013. Origin and evolution of the Deception island
caldera (South Shetland Islands, Antarctica). Bulletin of Volcanology, 75 (6), 732.
McDougall I., Harrison, T.M., 1999. Geochronology and thermochronology by the 40Ar/39Ar
method, 2nd edition. Oxford University Press, New York.
804
Molina, F., Martí, J., Aguirre, G., Vega, E., Chavarría, L., 2014. Stratigraphy and structure of the
805
Cañas Dulces caldera (Costa Rica). Geological Society of America Bulletin, 126 (11-12),
806
1465-1480.
807
808
Mooser, F., Ramírez, M.T., 1987. Faja Volcánica Transmexicana: Morfoestructura, tectónica y
vulcanotectónica. Boletín de la Sociedad Geológica Mexicana, 48 (2), 69-73.
809
Pardo, M., Suárez, G., 1995. Shape of the subducted Rivera and Cocos plates in southern
810
México: Seismic and tectonic implications. Journal Geophysical Research, 100 (12), 357-12.
811
Peiffer, L., Bernard-Romero, R., Mazot, A., Taran, Y.A., Guevara, M., Santoyo, E., 2014. Fluid
812
geochemistry and soil gas fluxes (CO 2–CH 4–H 2 S) at a promissory Hot Dry Rock
813
Geothermal System: The Acoculco caldera, México. Journal of Volcanology and Geothermal
814
Research, 284, 122-137.
815
Pérez‐Campos, X., Kim, Y., Husker, A., Davis, P.M., Clayton, R.W., Iglesias, A., Pacheco, J.,
816
Singh, S., Constantin, M.V., & Gurnis, M., 2008. Horizontal subduction and truncation of the
817
Cocos Plate beneath central Mexico. Geophysical Research Letters, 35(18).
818
Polak, B.G., Prasolov, E.M., Kononov, V.I., Verkhocskiy, A.B., González, A., Templos, L. A.,
819
Mañón, A., 1982. Isotopic composition and concentration of inert gases in Mexican
820
hydrothermal systems (genetic and applied aspects). Geofísica Internacional, 21 (3), 193-227.
821
Rocha, S., Jiménez, E., Palma, H., 2006. Propuesta para dos pozos exploratorios en el proyecto
822
geotérmico de Acoculco, Puebla. Comisión Federal de Electricidad, internal report OGL-
823
ACO-03/06, unpublished.
824
Renne P.R., Mundil, R., Balco, G., Min, K., Ludwig, K.R., 2010. Joint determination of 40K
825
decay constants and 40Ar*/40K for the Fish Canyon sanidine standard, and improved
826
accuracy for 40Ar/39Ar geochronology. Geochimica et Cosmochimica Acta, 74, 5349–5367
33
827
Rueda, H., Macías, J.L., Arce, J.L., Gardner, J.E., Layer, P.W., 2013. The~ 31ka rhyolitic
828
Plinian to sub-Plinian eruption of Tlaloc Volcano, Sierra Nevada, central México. Journal of
829
Volcanology and Geothermal Research, 252, 73-91.
830
Ryan, W.B., Carbotte, S.M., Coplan, J.O., O'Hara, S., Melkonian, A., Arko, R., Weissel, R.A.,
831
Ferrini, V., Goodwillie, A., Nitsche, F., Bonczkowski, J., 2009. Global multi‐resolution
832
topography synthesis. Geochemistry, Geophysics, Geosystems, 10 (3).
833
834
Saxby, J., Gottsmann, J., Cashman, K., & Gutiérrez, E., 2016. Magma storage in a strike-slip
caldera. Nature communications, 7, 12295.
835
Siebe, C., Macias, J.L., Abrams, M., Rodriguez, S., Castro, R., Delgado, H., 1995. Quaternary
836
explosive volcanism and pyroclastic deposits in east central México: implications for future
837
hazards. In: Chacko, J., Whitney A. (Eds.), Guidebook of geological excursions, in
838
conjunction with the Annual Meeting of the Geological Society of America. New Orleans,
839
Louisiana, 1-48.
840
Sosa-Ceballos, G., Macías, J.L., Avellán, D.R., Salazar-Hermenegildo, N., Boijseauneau-López,
841
M.E., Pérez-Orozco, J.D., 2018. The Acoculco Caldera Complex magmas: Genesis, evolution
842
and relation with the Acoculco geothermal system. Journal of Volcanology and Geothermal
843
Research, 358, 288-306.
844
Suter, M., 1984. Cordilleran deformation along the eastern edge of the Valles–San Luis Potosí
845
carbonate platform, Sierra Madre Oriental fold-thrust belt, east-central Mexico. Geological
846
Society of America Bulletin, 95(12), 1387-1397.Suter, M., Carrillo-Martínez, M., Quintero-
847
Legorreta, O., 1996. Macroseismic study of shallow earthquakes in the central and eastern
848
parts of the trans-Mexican volcanic belt, México. Bulletin Seismological Society of
849
America, 86 (6), 1952-1963.
850
851
852
853
Tello-Hinojosa, E., 1986. Geoquímica de la zona geotérmica de la caldera de Acoculco, Puebla.
CFE-GPG internal report 34/86, 15.
Urrutia-Fucugauchi, J., Flores-Ruiz, J., 1996. Bouger gravity anomalies and regional crustal
structure in central Mexico. International Geology Review, 38, 176–194.
854
Viggiano-Guerra, J.C., Flores-Armenta, M., Ramírez-Silva, G.R., 2011. Evolución del sistema
855
geotérmico de Acoculco, Pue., México: un estudio con base en estudios petrográficos del pozo
856
EAC-2 y en otras consideraciones. Geotermia, Revista Mexicana de Geoenergia, 24 (1), 14-
857
24.
34
858
Zúñiga, F.R., Pacheco, J.F., Guzmán-Speziale, M., Aguirre-Dıaz, G.J., Espındola, V.H., Nava,
859
E., 2003. The Sanfandila earthquake sequence of 1998, Querétaro, México: activation of an
860
undocumented
861
Belt. Tectonophysics, 361 (3), 229-238.
fault
in
the
northern
edge
of
central
Trans-Mexican
Volcanic
862
863
FIGURE CAPTIONS
864
Figure 1. A) Regional tectonic configuration of Central México showing the active subduction of
865
the Rivera and Cocos Plates beneath the North American Plate at the Middle-American Trench.
866
The dotted black line depicts the boundary of the Trans-Mexican Volcanic Belt (TMVB) and the
867
location of the Acoculco Caldera Complex and other volcanoes (LP, La Primavera; C, Fuego de
868
Colima; T, Tancítaro; Am, Amealco; Hc, Huichapan; To, Nevado de Toluca; P, Popocatepetl; M,
869
Malinche; O, Pico de Orizaba; and Hm, Humeros). The shaded relief map and bathymetric data
870
was acquired and modified from Ryan et al. (2009).
871
Figure 2. Simplified geological map of the Acoculco Caldera Complex after Avellán et al.
872
(2018), Sosa-Ceballos et al. (2018), and Calcagno et al. (2018). The map contains the distribution
873
of 30 stratigraphic units and a composite stratigraphic column supported by 24 new
874
ages and two 14C dates (Fig. 3). This map shows towns, main roads, the location of stratigraphic
875
logs, samples with age determinations, faults, the zone of intense hydrothermal alteration, and the
876
two CFE exploratory wells. The DEM was obtained from the INEGI topographic data with a
877
horizontal resolution of 15-m. The geologic profile at the bottom correlates the surface units of
878
this map with the CFE stratigraphic logs of the EAC-1 and EAC-2 wells with the location of
879
shallow intrusions.
880
Figure 3. Detailed stratigraphic column of the units and eruptive stages of the Acoculco Caldera
881
Complex composed of 30 units marked in the geological map of figure 2. To the right appear the
882
plateau age of units, the plateau age sorted by age, and their estimated total volume.
35
40Ar/39Ar
40Ar/39Ar
883
Figure 4.
884
Caldera Complex (see Table 1). Plateau and Isochron ages, where calculated, are reported at the
885
1-sigma level. For Ac90, the 6 single-step fusion results are shows as a single probability density
886
histogram (see Table 1).
887
Figure 5. Photos of selected outcrops of the Acoculco andesitic ignimbrite (Aai) with its different
888
lithological facies. A, Aspect of the welded massive structure of the ignimbrite with pumice
889
fragments embedded in a light-yellowish ash matrix. Hand lens for scale. B, Lacustrine deposits
890
that partly covered the Aai with parallel plane stratification. They are made of interbedded
891
massive fine sandy beds with laminar-stratified clayed beds. Notice that the succession is gently
892
tilted. C, Co-ignimbrite lag breccia (mlBr) of the Acoculco ignimbrite transected by E-W
893
structures and NW-SE faults. Notice that the lava lithics are supported by a light-brown to
894
yellowish ashy matrix with pumice. D, Co-ignimbrite lag breccia (mlBr) at the bottom of the
895
outcrop in contact with the Acoculco ignimbrite. The color change marks a difference in the
896
texture and lithological components of the deposits.
897
Figure 6.
898
Caldera Complex (see Table 1). Plateau and Isochron ages, where calculated, are reported at the
899
1-sigma level.
900
Figure 7. A, Regional structural map modified from García-Palomo et al. (2018). Dark gray areas
901
correspond to horsts and light gray areas correspond to grabens. Dike, scoria cones and
902
slickensides symbols show extension direction. B, Structural map of the Acoculco Caldera
903
Complex showing the main structural features, field sites (outcrops) and the caldera rim (this
904
study). C, The fault planes were plotted on stereographic projections in the lower hemisphere.
905
The trend-frequency polar diagram (rose diagram) was obtained with Rozeta 2.0, and the fault
906
planes and poles were plotted with Stereonet 9. D, A slip model showing an orthorhombic
40Ar/39Ar
age spectra and inverse isochron diagrams of dated rocks of the Acoculco
age spectra and inverse isochron diagrams of dated rocks of the Acoculco
36
907
symmetry of NW- and NE- oriented faults moving under the same strain field. Black zones
908
correspond to 100% compatible compression, and white zones 100% tension areas.
909
Figure 8. A, Image of the Magnetic Field Reduced to the Pole (MFRP) overlaid on a DEM of the
910
ACC region. The black dots represent the CFE wells. B, A 3D inversion model of the magnetic
911
susceptibility produced with the Geosoft´s VOXI software by using the magnetic vector inversion
912
method (Ellis et al., 2012). In this image four anomalies related to intrusive bodies occur at
913
depths between 1,000 and >2,500 m below the surface.
914
Figure 9. Eruptive stages during the formation of the Acoculco Caldera Complex described as
915
pre-caldera (A), syn-caldera (B), early post-caldera (C) late post-caldera (D), and still late post-
916
caldera and extra-caldera (E). After the caldera collapse two episodes of caldera reactivation
917
generated the Encimadas (Eri), and Tecoloquillo (Tr1) rhyolitic ignimbrites. Outside of the
918
caldera rim, the products of the Acoculco caldera are interbedded with deposits of the Apan-
919
Tezontepec volcanic field (extra caldera volcanism).
920
LIST OF TABLES
921
Table 1. List of dated samples from rock units of the Acoculco Caldera Complex. See text for
922
map units. All ages (in ka) are reported at the 1-sigma level. For most samples, the 40Ar/39Ar
923
plateau age is the interpreted age (shown in bold) of the unit. Sample Aco901 showed evidence
924
of excess argon, so the isochron age is the interpreted age. The age of sample Ac90 is a weighted
925
mean age of 6 single-step fusions, and for sample Ac94a, no plateau or isochron could be
926
calculated. MSWD: Mean Square Weighted Deviation. See text for analytical methods and
927
standards used. List of rock units of the Acoculco Caldera Complex dated in this study and their
928
40Ar/39Ar
929
Table 2. Radiocarbon date of paleosol sample performed during study. The location is in UTM
930
coordinates. Analyses performed at the International Chemical Analysis Inc. Florida, USA.
ages.
37
931
38
Table 1 continued
Table 1.
Sample
ACC stage
Unit
Location
North
East
Material
Integrated Age
(ka)
Plateau Age
(ka)
Plateau
information
Isochron Age
(ka)
Isochron information
Fractions=7
Init. 40Ar/36Ar=291.2
±3.2 MSWD=1.83
Fractions=10 of 14, 2
runs
Init. 40Ar/36Ar=303.4
±0.8 MSWD=0.46
Fractions=6
Init. 40Ar/36Ar=297.2
±12.1 MSWD=0.64
Ac42
Late postcaldera
Plc
2199245
593432
Whole rock
28 ± 12
71 ± 17
Fractions=7
79.0% of 39Ar rel.
MSWD=2.23
121 ± 32
Aco901
Late postcaldera
aplitic
dike
2203029
589693
Whole rock
-
-
No plateau
183 ± 36
Ac11
Late postcaldera
Tr1
2192344
586149
Plagioclase
614 ± 82
611 ± 72
Fractions=5
99.2% of 39Ar rel.
MSWD=0.69
633 ± 72
Ac11
Late postcaldera
Tr1
2192344
586149
Pumice
756 ± 10
762 ± 9
Fractions=8
%99.4 of 39Ar rel.
MSWD=0.71
755 ± 13
Fractions=8
Init. 40Ar/36Ar=299.4
±9.0 MSWD=0.77
Ac104
Late postcaldera
Arcd
2194852
591775
Whole rock
921 ± 38
998 ± 36
Fractions=6
92.4% of 39Ar rel.
MSWD=0.49
1,171 ± 81
Fractions=7
Init. 40Ar/36Ar=287.5
±3.5 MSWD=0.28
Ac89
Late postcaldera
Mrcd
2208434
572356
Whole rock
1,363 ± 54
1,066 ± 42
Fractions=5
78.7% of 39Ar rel.
MSWD=2.23
1,036 ± 43
Fractions=8
Init. 40Ar/36Ar=304.4
±2.3 MSWD=2.29
Ac37
Late postcaldera
Msf
2199284
575914
Pumice
2,850 ± 197
1,084 ± 22
Fractions=4
87.1% of 39Ar rel.
MSWD=2.45
1,136 ± 49
Fractions=4
Init. 40Ar/36Ar=290.3
±6.2 MSWD=2.38
Ac82
Late postcaldera
Arcd
2190307
595629
Obsidian
1,139 ± 10
1,145 ± 14
Fractions=8
99.8% of 39Ar rel.
MSWD=2.29
1,130 ± 9
Fractions=8
Init. 40Ar/36Ar=300.5
±5.0 MSWD=1.69
Ac100
Late postcaldera
Eri
2204497
596510
Whole rock
1,240 ± 12
1,278 ± 14
Fractions=5
94.4% of 39Ar rel.
MSWD=1.83
-
No isochron
Ac72
Late postcaldera
Ard
2199502
578889
Whole rock
1,190 ± 54
1,283 ± 88
Fractions=4
78.9% of 39Ar rel.
MSWD=2.08
1,309 ± 526
Fractions=4
Init. 40Ar/36Ar=295.9
±17.6 MSWD=2.92
Ac92
Late postcaldera
Trcd
2213879
583057
Whole rock
1,329 ± 12
1,360 ± 15
Fractions=4
76.0% of 39Ar rel.
MSWD=1.83
1,392 ± 27
Fractions=6
Init. 40Ar/36Ar=285.3
±9.2 MSWD=3.04
Ac94B
Late postcaldera
Crcd
2216042
585241
Whole rock
1,339 ± 8
1,394 ± 8
Fractions=5
71.6% of 39Ar rel.
MSWD=1.22
1,410 ± 12
Fractions=5
Init. 40Ar/36Ar=289.3
±4.6 MSWD=0.97
Ac107
Late postcaldera
Amrc
2213875
590387
Whole rock
1,413 ± 27
1,438± 24
Fractions=6
95.9% of 39Ar rel.
MSWD=0.20
1,462 ± 59
Fractions=6
Init. 40Ar/36Ar=273.2
±58.6 MSWD=0.21
Table 1 continued
Ac98B
Late postcaldera
Cbd
2202522
592368
Whole rock
1,562 ± 42
1,600 ± 35
Fractions=6
97.2% of 39Ar rel.
MSWD=0.57
1,608 ± 32
Fractions=7
Init. 40Ar/36Ar=294.1
±2.3 MSWD=0.63
Ac76
Late postcaldera
Tal
2201609
587830
Whole rock
1,717 ± 40
1,708 ± 54
Fractions=4
90.5% of 39Ar rel.
MSWD=2.64
1,738 ± 26
Fractions=8
Init. 40Ar/36Ar=294.9
±2.8 MSWD=1.55
Ac113
Late postcaldera
Alrd
2196999
584295
Whole rock
1,864 ± 38
1,870 ± 36
Fractions=7
99.6% of 39Ar rel.
MSWD=0.71
1,854 ± 32
Fractions=7
Init. 40Ar/36Ar=285.1
±80.2 MSWD=0.78
Ac69B
Syncaldera
Aai
2200645
584104
Pumice
1,991 ± 33
2,041 ± 38
Fractions=5
96.2% of 39Ar rel.
MSWD=1.59
2,018 ± 56
Fractions=5
Init. 40Ar/36Ar=299.3
±8.9 MSWD=1.98
Ac44
Early postcaldera
Atad
2195570
592761
Whole rock
2,183 ±19
2,179 ± 26
Fractions=4
82.3% of 39Ar rel.
MSWD=2.33
2,207 ± 31
Fractions=8
Init. 40Ar/36Ar=288.9
±12.8 MSWD=4.76
Ac69A
Syncaldera
Aai
2200645
584104
Pumice
2,279 ± 86
2,185 ± 65
Fractions=3
87.7% of 39Ar rel.
MSWD=0.71
2,113 ± 100
Fractions=8
Init. 40Ar/36Ar=308.2
±9.4 MSWD=0.54
Ac80
Early postcaldera
Mtal
2198103
586869
Whole rock
2,113 ± 28
2,199 ± 24
Fractions=4
87.2% of 39Ar rel.
MSWD=0.97
2,316 ± 70
Fractions=4
Init. 40Ar/36Ar=255.6
±26.6 MSWD=0.14
Ac103
Early postcaldera
Vtal
2205459
593663
Whole rock
2,288 ± 53
2,323 ± 48
Fractions=7
99.1% of 39Ar rel.
MSWD=0.35
2,321 ± 47
Ac90
Early postcaldera
Srl
2208919
574826
-
2,553 ± 110
-
-
-
Ac94A
TMVB
Pald
2216042
585241
Whole rock
3,620 ± 22
-
No plateau
-
No isochron
Ac15
TMVB
Pdhd
2190111
579819
Whole rock
10.74 ± 0.06 Ma
10.98 ± 0.07
Ma
Fractions=4
56.4% of 39Ar rel.
MSWD =2.16
-
No isochron
Fractions=7
Init. 40Ar/36Ar=295.9
±4.8 MSWD=0.42
Single shot fusions on 6
chips, MSWD=1.15
Probability=0.33
Table 2
Unit
Clc
Outcrop
Description
(sample)
Ac1
Paleosol
below Clc
Location
North
East
2197465
594866
Lab. Code
Pretreatment
Conventional age
Calibrated Age
18OS/0373
AO
16,710 ± 50 BP
Cal 18410 - 18020
BC