1
Aeromagnetic anomalies reveal hidden tectonic and volcanic structures in the central
2
sector of Aeolian Islands, Southern Tyrrhenian Sea, Italy
3
4
5
6
7
8
9
10
11
12
13
14
Riccardo De Ritis1, Guido Ventura2, Massimo Chiappini1
15
Abstract Salina Island (Italy) is located in the central sector of the Aeolian Islands and
16
represents the northernmost volcanic structure of an elongated ridge emplaced on a regional
17
shear zone characterized by NNW-SSE strike-slip faults and by second order N-S and NE-
18
SW faults. High-resolution, low-altitude aeromagnetic data collected in 2003 and 2005 allow
19
us to study the subsurface structure of Salina Island. The magnetic data show a pattern with a
20
wide range of wavelengths and intensities. Magnetic modeling constrained by volcanological
21
data allow us to reconstruct the inner structure of the Salina volcanoes and surrounding
22
marine regions. Long wavelength negative anomalies overlap E-W elongated sedimentary
23
basin related to the Early Pliocene-Pleistocene opening of the Southern Tyrrhenian Sea back-
24
arc basin. The shorter wavelength positive magnetic anomalies are related to the Late
25
Pleistocene-Holocene volcanic (conduits, dikes) and tectonic (faults) structures. The magnetic
26
and volcanological data indicate that the early (168- about 100 ka), basaltic to basaltic
27
andesitic Salina volcanism developed along N-S and NE-SW tectonic structure, whereas the
28
more recent basaltic andesitic to rhyolitic products (about 100-13ka) were emitted by vents
29
related to the main NW-SE fault tectonic structures. The tectonic structures also control the
30
location of the seamounts around the island and the geometry of the volcano-tectonic
31
collapses. The Salina volcanism emplaced on a NNW-SSE regional discontinuity that
32
represents a tear fault of the present-day roll-backing slab in the Southern Tyrrhenian Sea.
1
Istituto Nazionale di Geofisica e Vulcanologia, Department of Geomagnetism, Aeronomy and Environmental Geophysics, Viale del
Pinturicchio 23e, 00196 Roma, Italy
2
Istituto Nazionale di Geofisica e Vulcanologia, Department of Seismology and Tectonophysics, Via di Vigna Murata 605, 00143 Roma,
Italy
*Corresponding author: Riccardo De Ritis, Istituto Nazionale di Geofisica e Vulcanologia, Department of Geomagnetism, Aeronomy and
Environmental Geophysics, Viale del Pinturicchio 23e, 00196 Roma, Italy - Phone (+39) 06-36915627; Fax (+39) 06-36915617; e-mail:
deritis@ingv.it
2
1
2
3
4
5
6
7
Keywords: aeromagnetic survey, volcanoes, submarine vents, sector collapses, Aeolian
Islands
8
1. Introduction
Index terms: 0900 Exploration geophysics; 1517 Magnetic anomalies: modeling and
interpretation; 3075 Submarine tectonics and volcanism; 8185 Volcanic arcs; 8485 Remote
sensing of volcanoes.
9
Aeromagnetism is widely employed in volcanic areas due to the recent improvements
10
in data acquisition, precise GPS measurements, and accurate data processing [e.g., Finn et al.,
11
2001; Lénat el al., 2001]. In these areas, the aeromagnetic survey is particularly useful to
12
detect buried or submerged eruptive centers, alignments of vents, faults [i.e. Rollin et al.,
13
2000; Blanco-Montenegro et al., 2003], and, more in general, to analyze subsurface structural
14
features at temperatures below the Curie isotherm [De Ritis et al., 2005].
15
In the last five years, Istituto Nazionale di Geofisica e Vulcanologia (Italy) carried out
16
various high-resolution, low-altitude aeromagnetic surveys of the Aeolian volcanoes
17
(Southern Tyrrhenian Sea, Italy). In this study, we present an aeromagnetic anomaly map of
18
Salina Island (Aeolian Islands, Southern Tyrrhenian Sea; Fig. 1) and offshore using data
19
collected in 2003 and 2005. Salina last erupted at 13 ka and fumaroles and seismicity
20
characterizes its present activity. These features suggest that the magmatic system below the
21
island is still active. The structural importance of Salina is related to its central location within
22
the Aeolian Arc, in the area where the three main tectonic structures affecting the volcanic arc
23
converge (Fig. 1a,b,c). In addition, the largest eruption of the Aeolian volcanoes occurred in
24
this island [Pollara I eruption, 24 ka; Calanchi et al., 1993].
25
The results of the aeromagnetic surveys here are discussed in light of the available
26
volcanological, structural, and marine geology information. A high correlation between the
27
local and regional structural trends and the direction and shape of the magnetic anomalies is
28
found. The aeromagnetic data also constraint the extent of the Salina sector collapse
29
structures, and allow us to recognize previously unknown eruptive centers presently buried by
3
1
products of more recent activity and/or partly eroded. The data and results presented in this
2
study clarify the complex relationships among volcanism, tectonics and morphology in the
3
central sector of the Aeolian Islands.
4
5
2. Regional setting and geology of Salina Island
6
Salina Island (26 km2; 962 m a.s.l.) is located in the central sector of the Aeolian
7
volcanic Arc (Southern Tyrrhenian Sea, Italy). The calcalkaline, shoshonitic and alkaline
8
potassic character of the Aeolian volcanites, and the occurrence of a deep (up to 500 km of
9
depth) seismicity in the Southern Tyrrhenian Sea and Calabrian Arc reflect the active roll-
10
back of a NW-dipping slab related to the subduction of the Ionian plate beneath the Calabrian
11
Arc [Fig. 1a; Carminati et al., 1998; Faccenna et al., 2001; De Astis et al., 2003].
12
The central Aeolian Islands (Salina, Lipari and Vulcano) form a NNW-SSE elongated
13
volcanic ridge that emplaces along a major NNW-SSE dextral shear zone, the so called
14
Tindari-Letojanni (hereafter TL) fault system, which extends from Salina southward to the
15
northern Sicily [Fig. 1b,c; Mazzuoli et al., 1995; Ventura et al., 1999]. This shear zone and
16
the associated second-order N-S tensional and NE-SW striking faults separate the Aeolian
17
volcanoes (1.3 Ma-active) in two sectors characterized by different crustal thickness: the
18
western sector (Moho depth= 20-25 km) include the islands of Filicudi and Alicudi and the
19
western Aeolian seamounts; the eastern sector (Moho depth= 15-20 km) include the islands of
20
Stromboli and Panarea [De Astis et al., 2003]. The volcanoes of the western sector emplace
21
along WNW-ESE striking dextral to reverse faults and those of the eastern sector are affected
22
by NE-SW striking normal faults (Fig. 1b). Salina, Lipari and Vulcano are located at the
23
intersection of the TL shear zone and these two WNW-ESE and NE-SW fault systems (Fig.
24
1). Geodynamic models of the evolution of the Tyrrhenian Sea [e.g., Mantovani et al., 1996;
25
Gvirtzman and Nur, 2001; Goes et al., 2004] indicate that the WNW-ESE fault system acted
26
in the past (from Early Pliocene to Pleistocene) as a strike-slip structure bounding to the south
4
1
the southeastward opening of the Tyrrhenian Sea back-arc basin and the E-W Aeolian
2
sedimentary basins, whereas the TL faults act as tears of the roll-backing slab from
3
Pleistocene-Holocene times.
4
The evolution of the volcanism and the structural setting of Salina are well known and
5
only the main features are summarized here [Keller, 1980; Bernasconi and Ferrini, 1989;
6
Romagnoli et al., 1989; Barca and Ventura, 1991; Rossi et al., 1987; Calanchi et al. ,1993;
7
Critelli et al., 1993; Mazzuoli et al., 1995; Gertisser and Keller, 2000; Quarerni et al., 2001;
8
De Rosa et al., 1989, 2002, 2003), De Astis et al., 2003; Favalli et al., 2005; Ventura et al.,
9
2006; Donato et al., 2006]. Salina includes five main volcanic structures (Fig. 1d): the Rivi-
10
Capo volcanic complex, which consists of a NE-SW aligned eruptive fracture (Capo) and a
11
younger central vent (Rivi), the Fossa delle Felci cone, the Porri cone, the Corvo volcano and
12
the Pollara depression. The activity on the island starts with the formation of the basaltic-
13
andesitic Corvo volcano and of the Rivi-Capo volcanic complex. The age of the Corvo
14
volcano, which is largely dismantled by the erosion, is unknown but the compositional
15
features of its products (basalts) relate it to the early stages of activity on the island. The Rivi-
16
Capo volcanic complex developed between 87 and 168 ka in the northeastern sector of Salina
17
(Fig. 1d). The Rivi-Capo northern flank is cut by NE-SW striking normal faults, and consists
18
of dikes striking NE-SW, and of hydrothermalized volcanics that represent the remnants of
19
the upper portions of the Rivi-Capo conduits. The southern flank consists of lava flows and
20
scoria layers of basaltic-andesitic composition. The two flanks of the volcano are separated by
21
a NW-SE to NE-SW-oriented rim delimiting a north to northwest facing scar. This represents
22
the morphological expression of a sector collapse that occurred along a NE-SW structural
23
discontinuity. This collapse dissected the inner parts of the volcanic complex (Fig. 1d) and
24
occurred after 87 ka, which is the age of the younger outcropping lava flows.
25
The Fossa delle Felci volcano (962 m a.s.l.; Fig. 1d) formed between 108 and 59 ka in
26
the eastern sector of the island. The products consist of lava flows and minor pumice and
5
1
scoria deposits, and range in composition from basaltic-andesites to dacites. Lava flows
2
outcrop on the western and southern flanks, whereas pyroclastics and minor lavas mantle the
3
eastern flank. These latter products fill a pre-existing, amphitheatre-like morphological low of
4
uncertain origin. A well-preserved summit crater characterizes the top of the edifice. Minor
5
dome-like structures, which represent parasitic vents, outcrop at the base of the southeastern
6
flank of the edifice and between the summit crater and the Rivi-Capo morphological rim (Fig.
7
1d). The eastern sector of Fossa delle Felci is affected by a N-S striking fault with oblique
8
(dextral/normal) movements.
9
The Porri volcano (860 m a.s.l.) occupies the western sector of the island. Its activity
10
developed between 83 and 43 ka and started with the emplacement of lava flows that outcrop
11
at the base of the southern flank. Wet surges and breccia deposits overlies these lavas. This
12
basal succession of deposits is covered by scoria deposits and lava flows that form the
13
present-day Porri cone and summit crater (Fig. 1d). The Porri rocks range in composition
14
from basaltic-andesites and andesites to minor dacites. NW-SE striking oblique slip faults
15
affect the southern and northern sector of the Porri volcano as well as the valley separating
16
Porri from Fossa delle Felci.
17
The last activity on Salina developed in the northwestern corner of the island between
18
30 ka and 13 ka. A 1 km-wide amphitheatre-like depression (Pollara depression) affects the
19
northwestern sector of Porri (Fig. 1d). A 30 ka old dacitic lava flow (Perciato lava) delimits
20
the northern corner of the depression, which is filled by lacustrine and reworked deposits
21
related to two explosive eruptions occurred at 24 ka (Pollara I eruption) and 13 ka (Pollara II
22
eruption). The Pollara I sub-plinian eruption is the highest intensity [mass discharge rate up to
23
8
24
reported. The Pollara pyroclastics consist of pumice fall and flow deposits, and range in
25
composition from basaltic andesites to rhyolites (Pollara I), and from high-K2O dacites to
26
rhyolites (Pollara II). The northern and eastern rims of the Pollara depression are partly
.
106 kg/s; Calanchi et al., 1993] event recorded at the Aeolian Islands, as previously
6
1
covered by these pyroclastics, whereas the southeastern and southwestern rims cut the
2
northwestern flank and the crater of the Porri volcano, as well as the older lavas of the Corvo
3
edifice. NE-SW striking, northwest dipping normal faults affect the inner sector of the Pollara
4
depression [Critelli et al., 1993], which originated by the collapse of the northwestern flank of
5
the Porri volcano. This event occurred between 43 ka, which is the youngest age of the Porri
6
lava flows, and 30 ka, which is the age of the oldest lava flows infilling the depression.
7
Marine geology data reveal hummocky-like morphologies in the sea sectors in front of
8
the Rivi-Capo and Pollara sector collapses [Favalli et al., 2005]. These morphologies
9
represent the deposits related to these collapses. Six major seamounts have been recognized
10
around the island [Fig. 1c; Romagnoli et al., 1989]. The two larger seamounts (cone radius of
11
about 2 km) are located at about 5 km north of the Rivi-Capo volcanic complex. Other two
12
(cone radius of about 0.8-1 km) occur offshore the southeastern corner of the island. The last
13
two seamounts locate about 4 km east of the Fossa delle Felci and Rivi-Capo volcanoes.
14
Submarine fumaroles occur offshore the southeastern corner of the Porri volcano and offshore
15
the Pollara depression (Fig. 1d). Wide erosional shelves, which correspond to low stands of
16
the sea level, surround Salina Island and the other Aeolian islands.
17
18
3. Rock magnetic properties
19
Magnetic anomalies arise from sources located at all depths above the Curie isotherm
20
so that the magnetic anomaly field can provide an integrated areal and subsurface view that
21
can be used to infer 3D geological information. Because of the non-uniqueness of the
22
distribution of potential field sources, the interpretation of magnetic anomalies has to be
23
constrained with the available information about magnetic properties and geology of the
24
studied area. In particular, the directions and intensity of the magnetization of the lithological
25
units are necessary to provide a reliable geological interpretation of the subsurface structures
26
[Clark, 1997].
7
1
Magnetization of rocks depends on the chemical composition, mineral content,
2
texture, and it may be highly variable even inside the same lithologic unit. With respect to
3
rocks from other environments (e.g. sedimentary basins), volcanic rocks generally have a
4
higher magnetization due to the higher concentration of ferromagnetic minerals such as
5
titanomagnetite. The magnetic properties of the outcropping rocks can be evaluated through
6
direct measurements, but they may not reflect the characteristics of the subsurface rocks.
7
At Salina Island, the susceptibility and the natural remanent magnetization (NRM)
8
measurements were carried out on rocks representative of the main outcropping volcanic
9
units. The total magnetization, which is obtained through the integration of the collected field
10
measurements and already available data [Barberi et al., 1994; Zanella, 1995], allow us to
11
roughly evaluate the shallower magnetic property. An average magnetization value for each
12
volcanic edifice of Salina is considered taking into account the stratigraphy reported in Barca
13
and Ventura [1991] and Critelli et al. [1993]. Table 1 shows the values of remanent, induced
14
and total magnetization for the sampled units, as well as their Koenigsberger ratio, which is
15
the ratio between remanent and induced magnetizzation.
16
Due to the lack of paleomagnetic measurements on Salina Island, we consider the
17
available data on the directions of remanence from the neighbouring Lipari Island [Zanella,
18
1995]; these data are collected on rocks that cover the same age span and compositional range
19
as the rocks from Salina Island. The paleomagnetic samples of the andesites of Porri and
20
Fossa delle Felci volcanoes span in the time between 43 and 108 ka and have on the average I
21
= 51.6° and D = – 5.5°. The mean directional values of the magnetic Earth Field measured on
22
the Lipari samples, averaged on the last 223 ka, are I = 52.1° and D = -3.0°; these values are
23
not far from the present magnetic field directions (I = 54°, D = 2°) provided by the IGRF
24
model [IAGA, 2005]. The total magnetization direction, which is the vector sum of the
25
induced and remanent components and is the parameter required for most of the analysis in
26
the paper, will be even closer to the Earth’s inducing field direction. Therefore, most of the
8
1
analysis has been done assuming the induced magnetic parameters. Taking into account the
2
few available data, the total magnetization (Table 1) and the inclination and declination data
3
must be considered as a guidance to constrain the modelling. The magnetization values are,
4
on average, high and reflect the high basaltic andesite-andesite (~80 vol.%) to rhyolite-dacite
5
(~20 vol.%) ratio of the outcropping rocks on the island [Barca and Ventura, 1991]. The
6
range of the total magnetization varies between 5.0 A/m (basaltic lavas and dikes of Corvo
7
and Rivi Capo) to 2.0 A/m (dacitic lavas and pyroclastics of Fossa delle Felci and rhyolitic
8
pumices of Pollara). Magnetization data on the offshore volcanic units are unavailable.
9
However, the early phases of activity on Salina and the other Aeolian Islands were
10
characterized by the emission of basaltic to basaltic andesitic magmas [De Astis et al., 2003].
11
Therefore, the average magnetization of the offshore rocks is supposed to be that of the older
12
Salina rocks. Finally, the basement below the Aeolian volcanics consists of crystalline rocks
13
belonging to the Calabrian Arc, as also inferred from the xenoliths found in lavas and
14
pyroclastics. The magnetic susceptibility of these rocks is close to zero [K< 7× 10-7 SI;
15
Barberi et al., 1994]. Therefore, the magnetic sources in the central sector of the Aeolian
16
Islands are located in the uppermost 3 km of the crust, which is the maximum thickness of the
17
volcanic blanket overlaying the crystalline basement [Barberi et al., 1994; Ventura et al.,
18
1999].
19
20
4. Data acquisition and processing
21
Two low-altitude airborne surveys were carried out on Salina Island and its
22
surrounding sea sectors at constant barometric height. The first helicopterborne campaign was
23
conducted in October 2003 with NE-SW oriented profiles with a line spacing of 1500 m. The
24
second campaign was carried out in December 2005 with the same line spacing, but infilling
25
the 2003 survey in the middle. The two digital data sets were integrated and reduced to the
26
same geomagnetic epoch, so that the resulting profile spacing was 750 m (Fig. 2a).
9
1
The airborne platform was equipped with an optically pumped cesium magnetometer
2
to acquire the total intensity of the Earth’s magnetic field at a sampling rate of 10 Hz. The
3
survey was conducted mainly at constant barometric altitude (500 m), except for the few
4
profiles that intercepted the topographic relief exceeding the planned flight altitude. These
5
few lines were flown draped maintaining a terrain clearance of 200 m. A GPS receiver in a
6
differential configuration was used for an accurate survey navigation, whereas flight altitude
7
was recorded using a laser altimeter at 0.3 Hz. The magnetic base station consisted of a proton
8
precession magnetometer to record the Earth magnetic field variations, and it was located on
9
the neighboring Vulcano Island recording data at 1 Hz.
10
The data processing can be summarized as follows. The external magnetic field
11
variations were removed using the base station data. Subsequently, both datasets were
12
reduced to the same geomagnetic epoch (2003.0) using the L’Aquila geomagnetic
13
Observatory (Central Italy) datum as reference value. The magnetic anomaly field was then
14
obtained by removing the appropriate Geomagnetic Reference Field (IGRF) [IAGA, 2005].
15
Finally, the merged datasets were gridded at a 200 m spacing using the minimum curvature
16
algorithm. The obtained aeromagnetic map of Salina Island and the surrounding marine
17
regions is shown in Fig. 2b.
18
19
5. Magnetic modeling
20
With the aim to shift anomalies to positions directly above their sources, the magnetic
21
anomaly grid was reduced-to-the-pole [RTP; Baranov and Naudy, 1964]. This analytical
22
transformation converts a dipolar anomaly into a single positive or negative anomaly placed
23
above its source (Fig 3). In volcanic environments, the RTP tool is particularly useful since
24
the effects of different neighboring sources overlap. Here we used induced field directions
25
because they are nearly similar to the total magnetization directions.
10
1
We emphasize the long wavelength components using the Upward Continuation
2
algorithm with the aim to gather information about deeper sources. Since most of the survey
3
is at constant barometric altitude, except for the Salina areas above the 500 m, we applied the
4
continuation between flat surfaces [Jacobsen, 1987] and between uneven surface-to-flat
5
surface [Xia et al., 1993]. We verified that for this survey there are not substantial differences
6
between the results of the two methods. The upward continued field is finally Reduced-to-the
7
Pole as displayed in Fig. 4.
8
9
5.1 Long wavelength anomalies modelling
10
Figs. 2 and 3 show that negative values affect a wide area surrounding the
11
northeastern side of the volcanic ridge. The area centred on Salina is characterized by a 24 km
12
long, E-W elongated area marked by several positive, high intensity anomalies. This sector is
13
bounded to the North by NW-SE and E-W elongated negative anomalies. To the South,
14
several negative anomalies are aligned E-W. These negative anomalies could be related to the
15
occurrence of not magnetized bodies. To verify this hypothesis, we carried out a synthetic
16
model with the following constraints: (a) three, 10 km wide bands elongated E-W; (b) the
17
northernmost and southernmost bands have a constant magnetization value of 0.0 A/m,
18
whereas the central band has constant magnetization of 2.7 A/m. This configuration has been
19
selected to analyze the effect of the superimposition of a magnetic volcanic body, i.e. Salina
20
Island, on E-W elongated, not magnetized bodies, i.e. the Plio-Pleistocene sedimentary
21
basins. Results are reported in Fig. 5.
22
23
5.2 Terrain effects
24
The high topographic relief of Salina (962 m a.s.l.) and the high magnetization of the
25
outcropping units may cause a terrain effect. This effects is considered the consequence of the
26
contact between the air and an irregular magnetized surface [Grauch, 1986] and the resulting
11
1
anomalies mimic the topographic shapes. We are interested to isolate the terrain effect to
2
emphasize anomalies not related to the topography. Here, we model the terrain effect using
3
the Blakely [1981] and Grauch [1987] algorithms. We compute the terrain effect for a
4
topographic slab whose upper surface is the digital elevation model and whose bottom is the
5
depth of the surrounding seafloor (-1400 m). Salina was subdivided in ten polygons that
6
match the main physiographic and volcanological units. Each polygon is characterized by a
7
constant total magnetization value based on the data available for the reference unit (Table 1).
8
For the Pollara units, whose maximum thickness is only 70 m, we consider magnetization
9
higher than the reference value due to the presence of the high magnetized Porri lavas below
10
the Pollara pyroclastics. In the marine areas, an estimation of terrain magnetization was
11
conducted using a statistical correlation method [Grauch, 1987]. The final residual map is
12
obtained subtracting the terrain effect at variable magnetization from the aeromagnetic data
13
and reducing-to-the-pole the resulting grid (Fig. 6).
14
15
5.3 Forward Modelling
16
The goal of the forward modelling of the Salina magnetic anomaly field is to obtain
17
some quantitative information about the geometry and depth of magnetic anomalies. We
18
apply a 2.75D forward modelling based on Talwani and Heirtzler (1964), Rasmussen and
19
Pedersen (1979), and on the algorithm proposed by Won and Bevis (1987). In the modelling,
20
we used the reduced-to-the-pole anomalies since the difference between the remanence and
21
the induced magnetization can be considered negligible. Therefore, as reported above, the
22
total magnetizzation for the last 223 ka is assumed parallel to the present magnetic field
23
directions and the reduced-to-the-pole map can be considered reliable. We constrain the
24
model using seismic profiles and gravity models [Barberi et al., 1994; Pepe et al., 2000;
25
Finetti et al., 2005]. The depth and the shape of the crystalline basement (~0 A/m) and the
26
thickness of the volcanic and sedimentary blanket of the Aeolians are inferred from these
12
1
data. The outcropping units are constrained by the surface geology map of Barca and Ventura
2
[1991]. The magnetization values (Table 1) are used for the properties of the outcropping
3
rocks. These values define the order of magnitude of the average magnetization of the
4
volcanological units. An E-W profile is chosen for the modelling as shown in Fig. 7
5
This profile crosses the high intensity anomalies centred on the Rivi, Porri and Corvo
6
edifices, as well as some offshore volcanic structures. The profile is orthogonal to the graben-
7
like structure on which Vulcano, Lipari and Salina Island emplaced.
8
9
6. Discussion
10
The RTP anomaly map of Salina Island and the surrounding marine areas (Fig. 3)
11
displays magnetic anomalies covering a wide range of amplitudes and wavelengths. A
12
correlation between the anomalies and some morphological features can be easily recognized,
13
e.g. the positive anomalies located on the Rivi-Capo and Porri volcanoes. In the RTP upward
14
continued map (Fig. 4), the island and its neighboring regions are characterized by positive
15
values and the anomaly pattern is roughly elongated in the E-W direction. Negative values
16
occur north and south of the island, where long wavelength (≥ 5 km), N100°E (to the north)
17
and E-W (to the south) elongated anomalies prevail. These negative features have the same
18
elongation of the longer wavelength anomalies that characterize the sedimentary basins
19
located between northern Sicily and the Aeolian islands [Fig. 1b; Caratori et al., 2004]. In the
20
Salina sector, the northern negative anomaly (NA1 in Fig. 4) overlaps a basin bounded by a
21
N100°E striking fault. This magnetic feature includes the two seamounts located north of
22
Salina (Fig. 1c). The E-W elongated negative anomaly south of the island (NA2 in Fig. 4)
23
overlaps the western flank of the Salina-Lipari-Vulcano volcanic ridge and includes the two
24
seamounts located 2-3 km south of the southeastern corner of the island (Fig. 1c). In this
25
framework, NA1 and NA2 could be due to the high magnetization contrast between the low
26
or not magnetized basin infill (evaporites and sediments), which, in the above mentioned
13
1
basins, have a thickness up to 3 km, and the highly magnetized surrounding, basaltic
2
volcanics. This interpretation is supported by the results of the modeling shown in Fig. 5. The
3
good agreement between the computed (Fig. 5) and measured (Fig. 2b) long wavelength
4
anomalies suggests that the negative anomaly pattern north and south of Salina could be due
5
to the presence of sedimentary bodies with negligible magnetization. In this frame, the
6
negative signature of the seamounts located within the above mentioned basins (Fig. 3) could
7
be related to (a) a low magnetization related to the evolved, i.e., dacitic or rhyolitic,
8
composition of the volcanites, or (b) an reverse magnetization of the rocks. This latter
9
interpretation implies that the seamounts have ages older than about 0.7 Ma (Bruhnes epoch).
10
We have no data to unequivocally select one of these two hypotheses since geochronological
11
data on the seamounts are lacking. However, the well preserved dome-like shape of the
12
seamounts, the occurrence on Salina and neighboring Lipari Island of dacitic and rhyolitic
13
volcanites, and the young age of the Salina-Lipari-Vulcano ridge [age<223 ka; De Rosa et al.,
14
2003], suggest that the negative anomaly of the seamounts could reflect the evolved
15
composition of their products.
16
In the following, we focus on the smaller size magnetic structure of Salina Island (Fig.
17
3). The positive anomalies A and B represent the top of a larger scale anomaly elongated
18
NW-SE. Anomaly A extends from the northwestern corner of the island 7 km offshore and its
19
northwestern tip overlaps a NW-SE elongated, topographic ridge [Favalli et al., 2005]. Near
20
the coastline, the southeastern tip of this anomaly overlaps the erosional shelf located in front
21
of the Pollara depression. Anomaly A could be interpreted as a NW-SE striking, dike-like
22
body, probably representing the crystallized plumbing system of a submarine eruptive fissure.
23
Anomaly B overlaps the vent area of the 30 ka old Perciato lavas and of the Pollara I sub-
24
plinian eruption (see Fig. 1).
25
Anomaly C overlaps the floor of the Pollara depression, and, in particular, the
26
lacustrine deposits and vent area of the Pollara II eruption (Fig. 3). The shape of this anomaly
14
1
well depicts the sub-circular shape of the Pollara depression [De Rosa et al., 1989]. The
2
anomaly low might represent the magnetic signature of chaotic material related to the sector
3
collapse of the Porri northwestern sector. Anomaly E includes the summit crater of the Porri
4
volcano, the southern boundary of the Pollara depression and the Corvo vent area (see also
5
Fig. 1d). Therefore, this anomaly may be interpreted as the superimposition of the Porri and
6
Corvo plumbing systems. Anomaly F appears to be separated from anomaly E. Topographic
7
structures overlapping anomaly F are lacking. This observation suggest that this anomaly
8
reflect an intrusive body or, alternatively, a completely covered eruptive center.
9
The positive anomaly G represents the southern tip of a larger, N-S elongated anomaly
10
that extends from the Porri crater 5 km southward (Fig. 3). In the offshore, the anomaly G
11
overlaps a N-S elongated morphological ridge (see Fig. 1c for a weak signature). The
12
elongated shape of this anomaly suggests that it may be due to the occurrence of a submarine,
13
possibly N-S elongated, fissure or a dike-like shallow intrusive body. The negative anomaly
14
H extends 4 km offshore from the northern sector of Salina. This anomaly is N-S elongated
15
and it is bounded, to the west, by the positive anomalies B and E, and to the east, by the
16
anomalies I, J and K. The southern tip of anomaly H overlaps the collapsed sector of the Rivi-
17
Capo volcanic complex, whose magnetic signature is characterized by the NE-SW elongated
18
anomalies I (Rivi) and J (Capo). Therefore, the anomaly H appears to be the magnetic
19
expression of the Rivi-Capo sector collapse. The negative values of the anomaly field (Figs.
20
2b and 3) could be due to an accumulation of chaotic material, as suggested by the
21
occurrence of hummocky surfaces in the sea sector in front of the Rivi-Capo scar [Favalli et
22
al., 2005]. The N-S elongated positive anomaly K locates just to the north of Rivi-Capo,
23
where the bathymetry shows any evidence of volcanic relief. However, hydrothermalized
24
rocks and dikes testifying to the occurrence of an eroded vent outcrop along the northern
25
coastline of Salina. Indeed, K partly overlaps these volcanics [Keller, 1980; Barca and
26
Ventura, 1991]. Therefore, we propose that this anomaly is related to the shallow plumbing
15
1
system of this vent, which, according to the occurrence of an erosional shelf, has been
2
dismantled by the sea erosion. The positive features L and M form a NW-SE elongated
3
magnetic high that overlaps the Fossa delle Felci crater (L) and the secondary vent (M)
4
located at the southeastern corner of the volcano (Fig. 1d). Therefore, the two anomalies can
5
be interpreted as shallow conduits of the Fossa delle Felci volcano. The NW-SE strike, along
6
which L and M are aligned, and which is also the strike of the major TL fault segments
7
affecting the island, cannot be interpreted as the magnetic expression of a basaltic lava flow
8
pile because most of the outcropping lavas have a dacitic composition. We conclude that the
9
NW-SE elongated anomaly affecting the southeastern sector of Fossa delle Felci is probably
10
due to a buried dike system striking NW-SE. The above observations led us to conclude that
11
the Fossa delle Felci plumbing system is controlled by the TL tectonic structures.
12
The boundary between the negative and positive magnetic values, which is located
13
south of anomaly M, follows a N-S strike. This is also the strike of the fault affecting the
14
eastern flank of the Fossa delle Felci volcano (Fig. 3). Therefore, the magnetic boundary
15
could represent the offshore prolongation of this fault.
16
Anomalies N and O align along a N-S trend. N overlaps the two seamounts located
17
offshore the Fossa delle Felci volcano [Romagnoli et al., 1989]. Therefore, this anomaly
18
could be interpreted as the shallow plumbing system of these two submarine volcanoes. O
19
coincides with a submarine lava flow field [Favalli et al., 2005], possibly related to effusions
20
from the northernmost seamount. We infer that the anomaly O is due to this lava flow field.
21
Results of the forward model (Fig.7, profile A-A’) support the above discussed
22
interpretation of the anomalies of Salina Island and of its offshore sectors. The model crosses
23
the anomalies F, E, I and N depicted in Fig. 3. In particular, the anomalies F is consistent with
24
a submerged eruptive center (total magnetization=5 A/m) located West of the Corvo and Porri
25
volcanoes. Anomaly N is also consistent with the occurrence of a submerged center (total
26
magnetization=6.7 A/m) located to the East of the Capo edifice. The results from the forward
16
1
model reveal that the anomalies E and I are mainly due to the high magnetized rocks of the
2
inner structure of the Porri and Corvo volcanoes (anomaly E) and of the basaltic Rivi edifice
3
(anomaly I). This is indirectly demonstrated by the good correlation between the magnetic
4
anomalies of the magnetic anomaly map (Fig. 2 ) and those of the terrain-corrected map (Fig.
5
6). According to the geological data, the model well depicts the superimposition of the less
6
magnetic andesitic to dacitic Porri lavas on the higher magnetic basaltic-andesitic Corvo lavas
7
(Fig.7). In addition, the model shows the trace of the Pollara and Corvo collapses. The
8
resulting depression is filled by submarine lavas with total magnetization=7 A/m. As concerns
9
the Rivi edifice, its top is characterized by rocks with low magnetization. This feature is
10
consistent with the occurrence of the altered and hydrothermalized rocks that diffusely
11
outcrop on the top of Rivi. Also, the model is able to recognize the superimposition of the
12
younger (87 ka) Rivi edifice on the older (168 ka) Capo volcano. In the model (Fig.7), the
13
anomaly field decreases moving away from Salina. This feature is consistent with the
14
sedimentary successions overlapping the crystalline rocks of the Calabrian Arc, as also
15
suggested from Fig. 5. However, on the short wavelength scale, the geometry of the anomaly
16
field away from Salina Island cannot be explained by the sediments alone, and the occurrence
17
of highly magnetized surface bodies is required (Fig. 7). According to marine geology data
18
[Favalli et al., 2005], these bodies consist of submarine lavas.
19
The model depicted along the B-B’ profile in Fig. 7 crosses the anomalies NA1, H, I,
20
M and NA2 of Fig. 3. According to geological data (Figs. 1 and 3) and the previously
21
discussed interpretation, the anomaly NA1 the sediments filling the E-W elongated basin
22
located north of Salina. The anomaly H is due to the altered rocks filling of the Rivi-Capo
23
collapse structure. The anomaly I is fully justified by the high magnetized rocks of the
24
shallow plumbing (dike) system of Rivi, whereas the positive anomaly M is related to a
25
secondary vent located 1 km southeast of the present-day crater. The profile B-B’ allow us to
26
reconstruct the structure of the Fossa delle Felci volcano. In particular, the top of the edifice is
17
1
characterized by a cone-like body with low magnetization (0.3 A/m). This body corresponds
2
to the altered material (hydrothermalized scorias) forming the top of the edifice. In addition,
3
the anomaly pattern between M and NA2 in Fig. 7 (profile B-B’), is well justified by the high
4
magnetized, basaltic-andesitic lava flows affecting the southern flank of Fossa delle Felci
5
(Fig. 1). The anomaly NA2 corresponds to the sediments filling the E-W elongated basin
6
located between the islands of Salina and Lipari. Here, the profile clearly shows, according to
7
the available marine geology data [Favalli et al., 2005], that the relationships between the
8
Salina and Lipari products are controlled by tectonic faults.
9
Taking into account the approximations of the forward modelling (see section 5.4), the
10
results shown in Fig. 7 represent a reasonable, geologically consistent interpretation of the
11
volcanic structure of Salina Island.
12
13
7. Concluding remarks
14
The data and results from the magnetic modeling presented here show that the long
15
wavelength negative anomalies are related to sedimentary basins, whereas the shorter
16
wavelength positive anomalies are due to volcanic and tectonic structures. The combined
17
analysis of magnetic and volcanological data indicate that the Salina volcanism older than 100
18
ka developed along N-S and NE-SW tectonic structure, whereas the volcanism younger than
19
100 ka was related to the main NW-SE faults. In particular, the NW-SE elongated positive
20
anomalies are caused by the dike-like structures of the southeastern sector of Fossa delle Felci
21
and of the Pollara marine region. (Figs. 1b and 4b). The Rivi-Capo plumbing system, on the
22
contrary, align NE-SW. The other offshore positive anomalies O, N, G and K, align N-S.
23
Taking into account the age of the Salina volcanoes, the early (~168 ka) phases of activity
24
(Rivi-Capo volcanic complex) were characterized by volcanoes with plumbing systems
25
controlled by the antithetic NE-SW and tensional N-S striking structures, whereas the more
26
recent activity, i.e. the 43 to 108 old Fossa delle Felci and Porri volcanoes and the 13-30 ka
18
1
old Pollara vents, developed along the TL, NW-SE striking main structures. This tectonic
2
scheme, which is highlighted by the elongation and/or alignment of the positive anomalies, is
3
in agreement with the results from experimental models on the evolution of shear zones
4
[Bartlett et al., 1981]. These models show that the early structures that develop in a shear
5
zone are the antithetic (R’) and tensional (T) fractures, while the main fault appears at a later
6
time.
7
On Salina, the main tectonic structures control the Pollara and Rivi-Capo sector
8
collapses, whose magnetic signature is well defined by negative anomalies possibly related to
9
the chaotic nature of the collapse deposits (Rivi-Capo), or by the geometry of the
10
morphological depression (Pollara). In both cases, the depicted geometry is, according to the
11
surface geology [Critelli et al., 1993; Favalli et al., 2005], that of circular-like magnetic lows.
12
Finally, the overall magnetic pattern of Salina Island and offshore regions reflects the
13
superimposition of the volcanoes on the Calabrian Arc crystalline rocks and on the
14
sedimenary basins related to the Pliocene-Pleistocene subsidence of northern Sicily.
15
16
Acknowledgments We thank Elena Zanella for providing us some magnetic measurements,
17
and Iacopo Nicolosi for discussions about the Aeolian magnetism. The Salina airborne
18
magnetic survey has been funded by INGV. We thanks an anonymous reviewer, D. Ravat, an
19
anonymous Associate Editor, and Patrick Taylor for the in-depth comments and suggestions
20
that greatly improved the manuscript, and for the editorial handling.
21
22
References
23
24
25
Baranov, V., H. Naudy (1964), Numerical calculation of the formula of reduction to the
magnetic pole, Geophysics, 29, 67-79.
19
1
Barberi, F., A. Gandino, A. Gioncada, P. La Torre, A. Sbrana, C. Zenucchini (1994), The
2
deep structure of the Eolian arc (Filicudi-Panarea-Vulcano sector) in light of gravity,
3
magnetic and volcanological data, J. Volcanol. Geotherm. Res., 61, 189-206.
4
5
Barca, D., G. Ventura (1991), Volcano-tectonic evolution of Salina island (Aeolian
Archipelago, South Tyrrhenian Sea), Mem. Soc. Geol. It., 47, 401-415.
6
Bartlett, W.L., M. Friedman, J.M. Logan, (1981), Experimental folding and faulting of rocks
7
under confining pressure. IX. Wrench fauls in limestone layers, Tectonophysics, 79,
8
255-277.
9
10
11
12
13
14
Bernasconi, M.P., G. Ferrini (1989), I terrazzi marini dell'isola di Salina (Isole Eolie,
Tirrenomeridionale): un'ipotesi di datazione, AIQUA, 2, 49-55.
Blakely, R. J. (1981), A program for rapidly computing the magnetic anomaly over digital
topography, U.S. Geol. Surv. Open-file rep. 81-298.
Blakely, R. J., V. J. S. Grauch (1983), Magnetic models of crystalline terrain: Accounting for
the effect of topography, Geophysics, 48, 1551 – 1557.
15
Blanco-Montenegro, I., J.M. Torta, A. Garcia, V. Araña (2003), Analysis and modelling of
16
the aeromagnetic anomalies of Gran Canaria (Canary Islands), Earth Placet. Sci. Lett.,
17
206, 601–616
18
Calanchi, N., R. De Rosa, R. Mazzuoli, P.L. Rossi, R. Santacroce, G.Ventura (1993), Silicic
19
magma entering a basaltic magma chamber: eruptive dynamics and magma mixing- an
20
example from Salina (Aeolian Islands, Southern Tyrrhenian Sea), Bull. Volc., 55, 504-
21
522.
22
Caratori, F., P. Stefanelli, I. Giori, O. Faggioni, C. Carmisciano (2004), Aeromagnetic
23
anomaly map of Italy (scale 1:1.500.000), supplement to Annals of Geophysics, 47.
24
Carminati, E., M.J.R., Wortel, W. Spakman, R. Sabadini (1998), The role of slab detachment
25
processes in the opening of the western-central Mediterranean basins: some geological
26
and geophysical evidence, Earth and Planet. Sci. Letters, 160, 651-665.
27
Clark, D.A. (1997), Magnetic petrophysics and magnetic petrology: aids to geological
28
interpretation of magnetic surveys, AGSO Journal of Australian Geology and
29
Geophysics, 17, 83-103.
30
Critelli, S., M. Sorriso-Valvo, G. Ventura (1993), Relazioni tra attività vulcanica,
31
sedimentazione epiclastica ed evoluzione geo-morfologica nell'Isola di Salina (Isole
32
Eolie), Boll. Soc. Geol. It., 112, 447-470.
33
De Astis, G., G. Ventura, G. Vilardo (2003), Geodynamic significance of the Aeolian
34
volcanism (Southern Tyrrhenian Sea, Italy) in light of structural, seismological and
35
geochemical data. Tectonics, doi:10.1029/2003TC001506.
20
1
De Ritis R., I. Blanco-Montenegro, G. Ventura, M. Chiappini (2005), Aeromagnetic data
2
provide new insights on the tectonics and volcanism of Vulcano Island and offshore
3
areas
4
doi:10.1029/2005GL023465
(Southern
Tyrrhenian
Sea,
Italy),
Geoph.
Res.
Lett.,
32,
L15305,
5
De Rosa, R., H. Guillou, R. Mazzuoli, G. Ventura (2003), New Unspiked K-Ar ages of
6
volcanic rocks of the central and western sector of the Aeolian Islands: reconstruction of
7
the volcanic stages, J. Volcanol. Geotherm. Res., 120, 161-178.
8
De Rosa, R., P. Donato, G. Ventura (2002), Fractal analysis of mingled/mixed magmas: an
9
example from the Upper Pollara eruption (Salina Island, Southern Tyrrhenian Sea,
10
Italy), Lithos, 65, 299-311.
11
De Rosa, R., R. Mazzuoli, P.L. Rossi, R. Santacroce, G. Ventura (1989), Nuovi dati ed ipotesi
12
per la ricostruzione della storia eruttiva dell'Isola di Salina (Isole Eolie), Boll. GNV-
13
CNR, 2, 791-809.
14
Donato, P., H. Behrens, R. De Rosa, F. Holtz, F. Parat (2006), Crystallization conditions in
15
the Upper Pollara magma chamber, Salina Island, Southern Tyrrhenian Sea, Mineralogy
16
and Petrology, doi: 10.1007/s00710-005-0105-5
17
Faccenna, C., T.W. Becker, P. Lucente, L. Jolivet, F. Rossetti (2001), History of subduction
18
and back-arc extension in the Central Mediterranean. Geoph. J. Int., 145, 809-820.
19
Favalli, M., D. Karatson, R. Mazzuoli, M.T. Pareschi, G. Ventura (2005), Volcanic
20
geomorphology and tectonics of the Aeolian Archipelago (Southern Italy) based on
21
integrated DEM data, Bull. Volcanol., doi: 10.1007/s00445-005-0429-3.
22
Finetti et al., 2005
23
Finn, C., T. W. Sisson, and M. Deszcz-Pan (2001), Aerogeophysical measurements of
24
collapse-prone hydrothermally altered zones at Mount Rainer volcano, Nature, 409,
25
600–603.
26
27
Gertisser, R., J.Keller, (2000), From basalt to dacite: origin and evolution of the calc-alkaline
series of Salina, Aeolian Arc, Italy, Contrib. Mineral. Petrol., 139, 607-626.
28
Goes, S., D. Giardini, S. Jenny; C. Hollenstein, H.G. Kahle, A. Geiger (2004), A recent
29
tectonic reorganization in the south-central Mediterranean, Earth Planet. Sci. Lett., 226,
30
335-345.
31
Grauch, V. J. S. (1986), VARMAG: a FORTRAN program to implement the variable-
32
magnetization terrain correction method for aeromagnetic data, U.S. Geol. Surv. Open-
33
file rep. 86-268.
34
35
Grauch, V. J. S. (1987), A new variable-magnetization terrain correction method for
aeromagnetic data, Geophysics, 52, 94– 107.
21
1
2
3
4
5
6
Gvirtzman, Z., and Nur, A. (2001), Residual topography, lithospheric thickness, and sunken
slabs in the central Mediterranean, Earth Planet. Sci. Lett., 187, 117-130.
IAGA (2005) The 10th-Generation International Geomagnetic Reference Field, Geoph. J.
Int., 161, 561-565.
Jacobsen, B. H. (1987) A case for upward continuation as standard separation filter for
potential field maps, Geophysics, 52, 1138-1148.
7
Keller, J. (1980), The Island of Salina., Rend. Soc. It. Miner. Petrol., 36, 489-524.
8
Lanza, R., E. Zanella (2003), Paleomagnetic secular variation at Vulcano (Aeolian Islands)
9
during the last 135 kyr, Earth Planet. Sci. Lett., 213, 321– 336.
10
Lénat, J.F., B. Gibert-Malengreau, A. Galdeano (2001), A new structural model for the
11
evolution of the volcanic island of Réunion (Indian Ocean), J. Geophys. Res., 106(B5),
12
8645–8663
13
Mantovani, E., D. Albarello, C. Tamburelli, D. Babbucci, (1996), Evolution of the Tyrrhenian
14
basin and surrounding regions as a result of the Africa-Euroasia convergence, J.
15
Geodynamics, 21, 35-72.
16
Mazzuoli, R., L. Tortorici, G.Ventura (1995), Oblique rifting in Salina, Lipari and Vulcano
17
Islands (Aeolian Islands, Southern Tyrrhenian Sea, Italy), Terra Nova, 7, 444-452.
18
Pasquale, V., M. Verdoya, P. Chiozzi (1999), Thermal state and deep earthquakes in the
19
Southern Tyrrhenian, Tectonophysics, 306, 435-448.
20
Pepe, F., G. Bertotti, F. Cella, E. Marsella (2000), Rifted margins formation in the Southern
21
Tyrrhenian Sea: A high- resolution seismic profile across the north Sicily continental
22
margin, Tectonics, 19, 241-257.
23
Quareni, F., G. Ventura, F. Mulargia (2001), Numerical modelling of the transition from
24
fissure-type to central-type activity on volcanoes: a case study from Salina Island, Italy,
25
Phys. Earth Planet. Int., 124, 213-221.
26
Rollin, P. J., J. Cassidy, C. A. Locke, H. Rymer (2000), Evolution of the magmatic plumbing
27
system at Mt Etna: new evidence from gravity and magnetic data, Terra Nova, 12, 193-
28
198.
29
Romagnoli, C., N Calanchi., G. Gabbianelli, G. Lanzafame, P.L. Rossi (1989), Contributi
30
delle ricerche di geologia marina alla caratterizzazione morfostrutturale ed evolutiva dei
31
complessi vulcanici di Salina, Lipari e Vulcano (Isole Eolie), Boll. GNV-CNR, 2, 971-
32
978.
33
34
Rossi, P.L, N. Calanchi, G. Gabbianelli, G. Lanzafame (1987), Nuovi dati strutturali su Salina
e sull'area sottomarina circostante, Boll. GNV-CNR, 2, 599-611.
22
1
Ventura, G., G. Vilardo, G. Milano, N.A. Pino (1999), Relationships among crustal structure,
2
volcanism and strike-slip tectonics in the Lipari-Vulcano volcanic complex (Aeolian
3
Islands, Southern Tyrrhenian Sea, Italy), Phys. Earth Planet. Int., 116, 31-52.
4
Ventura, G., P. Del Gaudio, G. Iezzi (2006), Enclaves provide new insights on the dynamics
5
of magma mingling: a case study from Salina Island (Southern Tyrrhenian Sea, Italy),
6
Earth Planet. Sci. Lett., doi: 10.1016/j.epsl.2006.01.004
7
Zanella, E. (1995), Studio delle variazioni paleosecolari del campo magnetico terrestre
8
registrate nelle vulcaniti Quaternarie dell’area tirrenica e del Canale di Sicila. Ph.D.
9
Thesis, Universita’ di Torino, 198 pp..
10
11
12
13
14
15
16
17
Figure captions
18
19
Figure 1. (a) Geodynamic setting of Italy and Southern Tyrrhenian Sea (data from:
20
Mantovani et al., 1996; Carminati et al., 1998; Gvirtzman and Nur, 2001; De Astis et
21
al., 2003); (b) Simplified structural scheme of the Aeolian Islands (data from: Mazzuoli
22
et al., 1995; Ventura et al., 1999; Pepe et al., 2000; De Astis et al., 2003). (c) Structural
23
scheme of the Salina-Lipari-Vulcano volcanic ridge (data from: Romagnoli et al., 1989;
24
Ventura et al., 1999; Favalli et al., 2005). (d) Geological map of Salina Island (data
25
from: Keller, 1980; Rossi et al., 1987; Barca and Ventura, 1991; Mazzuoli et al., 1995).
26
27
Figure 2. (a) Aeromagnetic survey flight path. The 2003 profiles are reported in red, and
28
those of 2005 are black. (b) Shaded relief aeromagnetic anomaly map of Salina Island
29
and offshore.
30
23
1
Figure 3. Reduced-to-the-pole aeromagnetic anomaly map (Geomagnetic inclination = 54°;
2
declination = 2°). Labels refer to the anomalies discussed in the text. Structural and
3
morphological features of Salina Island and offshore are from Romagnoli et al, [1989],
4
Ventura et al. [1999], and Favalli et al. [2005] (see also Fig. 1c). The figure covers the
5
available Digital Terrain Model [Favalli et al., 2005].
6
7
Figure 4. Reduced-to-the-pole upward continued (500 m) aeromagnetic anomaly map of
8
Salina Island and offshore. NA1 is the East-West oriented negative anomaly located
9
North of Salina Island and NA2 is the East-West elongated negative anomaly located
10
South of Salina Island.
11
12
Figure 5. Synthetic anomaly magnetic field generated by the Blakely and Grauch [1983] and
13
Grauch [1987] algorithms by adopting the following constant magnetization values: 2.7
14
A/m in a 10 km wide, E-W elongated band including Salina Island and its offshore; 0.0
15
A/m in two 10 km wide, E-W elongated bands located North and South of the previous
16
band. The area covered by the data (see Fig. 2), is also reported.
17
18
Figure 6 Reduced-to-the-pole terrain-corrected aeromagnetic anomaly map obtained with
19
variable magnetization [Blakely, 1981; Grauch, 1986]. The reference values are reported
20
in Table 1.
21
22
Figure 7 Forward 2.75D magnetic model of Salina and its surrounding marine areas along a
23
E-W striking cross section (A-A’ in the inset). The labels identify the anomalies
24
reported in Fig. 3 and discussed in the text.
25
26
24
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Table 1. Mean rock magnetic properties at Salina Island compiled using the paleomagnetic data provided by Barberi et al. [1994], Zanella
20
[1995] and field measurement. JI is the induced magnetization (calculated assuming an ambient magnetic field of 35 A m-1), JR is the natural
21
remanent magnetization, Q is the Koenigsberger ratio (JR/JI) and JTOT is total magnetization (calculated on the assumption that JR and JI are
22
parallel).
23
24
Volcano
Corvo
Age (ka)
Rivi - capo
Fossa delle felci
Litology
Deposits
JI (Am-1)
JR (Am-1)
Q
JToT (Am-1)
Basalts, basaltic andesites
Lava flows and scorias
1
4.0
4
5.0
87 - 168
Basalts, basaltic andesites
Lava flows and scorias
0.7
2.8
4
3.5
59 -108
Basaltic andesites, andesites, high-K dacites
Lava flows, scorias, pumice
falls and flows, domes
0.9
2.7
3
3.6
Porri
43 - 83
Basaltic andesites, andesites, dacites
Lava flows and scorias
0.9
2.1
2.3
3.0
Pollara
30 - 13
Basaltic andesites, andesites, dacites , highK dacites, rhyolites
Lava flows, pumice falls and
flow
0.3
1. 7
5.6
2.0