Content uploaded by J. Urrutia Fucugauchi
Author content
All content in this area was uploaded by J. Urrutia Fucugauchi on Dec 18, 2013
Content may be subject to copyright.
Available via license: CC BY 4.0
Content may be subject to copyright.
Earth Planets Space, 58, 1061–1069, 2006
Lava identification by paleomagnetism: a case study and some problems
surrounding the 1631 eruption of Mount Vesuvius, Italy
Gennaro Conte-Fasano
1∗
, Jaime Urrutia-Fucugauchi
1
, Avto Goguitchaichvili
1
, Alberto Incoronato
2
, and Pasquale Tiano
2
1
Laboratorio de Paleomagnetismo y Geofisica Nuclear, Instituto de Geof
´
ısica, UNAM, Ciudad Universitaria, 04510 Mexico, D.F., Mexico
2
Dipartimento di Scienze della Terra, Universit
`
a degli Studi di Napoli Federico II, Largo S. Marcellino no. 10, 80138 Naples, Italy
(Received February 7, 2005; Revised February 23, 2006; Accepted March 7, 2006; Online published September 16, 2006)
Detailed rock magnetic, paleomagnetic and absolute paleointensity studies of lava flows from the disputed
1631 Mount Vesuvius eruption are reported. The magnetic carrier consists of pseudo-single domain state Ti-poor
titanomagnetites. Characteristic magnetization directions determined from detailed stepwise alternating field and
thermal demagnetizations provide four new well-defined flow unit mean directions, with α
95
ranging from 0.7
◦
to 2.6
◦
. Paleodirections for 11 lava flows from 24-four flows studied previously appear to be related to the 1631
eruption, as indicated by their correlation to the early 17th century segment of the Italian paleosecular variation
reference curve. This provides new evidence supporting the conclusion that the 1631 episode was an explosive-
effusive eruption. The paleointensity results obtained from this study are the first to be published for Mount
Vesuvius, with virtual dipole moments of 9.24±1.8×10
22
and 13.5±0.4×10
22
Am
2
higher than the present-day
geomagnetic field strength.
Key words: Paleosecular variation, Vesuvius volcano, 1631 eruption, paleointensity, Italy.
1. Introduction
The Somma-Vesuvius complex is localized in the Cam-
pania plain, a region of southern Italy bordered by Meso-
zoic carbonate platforms. Its formation was related to the
tectonic activity that commence in the upper Pliocene, fol-
lowing formation of the Apennine belt. During the Qua-
ternary, an intense and complex volcanic activity, in asso-
ciation with the subduction of the African plate beneath
the Apennines chain, formed the large Roman-Campana
Province (Di Girolamo, 1978). The main eruptive cen-
ters are localized in the Lazio (Vulsini, Vico, Sabatini) and
Campania regions (Roccamonfina, Campi Flegrei, islands
of Ischia and Procida, and Somma-Vesuvius).
The Somma-Vesuvius complex is one of the most dan-
gerous volcanic complexes in southern Italy and is char-
acterized mainly by explosive activity. Its eruptive history
has been well documented in historical documents since 79
A.D. The Somma-Vesuvius complex is formed by two su-
perposed volcanic structures: Vesuvius volcano, with a con-
ical shape truncated at the summit, and the older Somma
volcano, which represents the caldera of the volcanic com-
plex. Stratigraphical observations and radiometric dating
of some of the volcanic products intercalated between pa-
leosoils have provided evidence of several cycles of erup-
tive activity, which begin usually with Plinian eruptions and
have durations of less than 2000 years. The oldest prod-
∗
Also at: Instituto de Investigaciones Antropologicas, UNAM, Ciudad
Universitaria, 04510 Mexico, D.F., Mexico.
Copyright
c
The Society of Geomagnetism and Earth, Planetary and Space Sci-
ences (SGEPSS); The Seismological Society of Japan; The Volcanological Society
of Japan; The Geodetic Society of Japan; The Japanese Society for Planetary Sci-
ences; TERRAPUB.
ucts of the Somma volcano are estimated to date from about
25 ka ago. In general, the activity of the Somma-Vesuvius
complex can be divided into three long periods (Delibrias et
al., 1979; Arn
´
o et al., 1987; Santacroce, 1987). The first is
the longest and oldest period of the volcanic complex, ex-
tending from 25 ka ago to 79 A.D. (the large historic Plinian
eruption);, this period is characterized by the formation of
the caldera. The second period includes the period of post-
caldera activity, during which time there were several erup-
tions (at least 11), with the most famous being the 79 A.D.
and 1631 ‘big’ eruptions. The third period comprises the
recent eruptive activity between 1631 and 1944.
It is generally agreed that the ‘big eruption’ of Vesuvius
in 1631 apparently occurred after 500 years of quiescence.
The eruption started in early morning of 16 December 1631
with the formation of a tall Plinian column and culminated
with the emission of pyroclastic flows and the generation of
secondary lahars. This event is considered to be one of the
most destructive episodes in the volcanic history of Mount
Vesuvius and the last in which large pyroclastic flows were
emitted. As far as the nature of this eruption is concerned, it
was initially accepted that lava flows were associated to this
eruption, as also supported by the first paleomagnetic study
(Hoye, 1981) of two lava flows. The nature of this eruption
was later questioned by Rosi and Santacroce (1984) and
Rosi et al. (1993) who considered this event to be an explo-
sive eruption, suggesting that lavas previously attributed to
this event were much older and had probably been erupted
during the 968–1039 period. Rolandi et al. (1991) and
Rolandi and Russo (1987, 1989, 1993) considered the 1631
event to be an explosive-effusive eruption, with the lavas al-
ready having been mapped by Le Hon (1865). To provide
new data that would contribute towards solving this dispute,
1061
1062 G. CONTE-FASANO et al.: LAVA IDENTIFICATION BY PALEOMAGNETISM
40°50’N
40°45’N
14°22’E
14°28’E
|
|
Tanguy et al., 1999
Carracedo et al., 1993
Gialanella et al., 1993
This study
Presumably 1631 eruption
Undifferentiated pyroclastic products of the Vesuvius and Mt. Somma.
Air-fall deposits and pyroclastic flows, surges and lahar deposits younger
than 17,000 years.
Lava flows of the A.D. 1760 eruption. Rocks with association ranging in
composition between tephritic and phonolitic leucitites.
Lavas of the period A.D. 1855-1872.
Lava flows before A.D. 1631 and after A.D. 79, frequently covered by
thick pyroclastic deposits of the A.D. 1631 eruption.
Main fragmental flow deposits (glowing clouds and associated ash clouds,
lahars) of the A.D. 1631 eruption.
Lavas of the period A.D. 1913-1944; flows and exogenous tholoids showing
generally ìaaî and ropy “pahoehoe” surfaces. Rocks of the high potassium.
|
|
Fig. 1. Simplified geologic/volcanological map (modified from Principe et al., 1987) showing the location of paleomagnetic sampling sites (the location
of sites studied by Tanguy et al. (1999) are not shown because these flows are reported in Carracedo et al. (1993) while those of Hoye (1981) are
unknown). Stars refer to the presumed 1631 eruption.
new paleomagnetic studies were initiated in the early 1990s
onwards, but these have provided conflicting results. For
instance, Carracedo et al. (1993) did not recognize any flow
associated to this eruption, a conclusion also reached by
Tanguy et al. (1999), who re-sampled at the same sites the
flows studied by Carracedo et al. (1993), and by Tanguy et
al. (2003). In contrast, Gialanella et al. (1993) and Incoro-
nato (1996) concluded that the 1631 eruption was character-
ized by an effusive phase. In particular, Incoronato (1996)
showed that the disputed lava flows were characterized by
mean directions of magnetization that clearly differed from
those of the Medieval flows, the age of which had been pa-
leomagnetically evaluated (968–1039) in a regional study
(Incoronato and Del Negro, 2004).
The problem of the nature of the 1631 eruption is of ac-
tual relevance in the evaluation of potential volcanic hazards
in the Vesuvius area. In this study, we re-examined previ-
ously published paleomagnetic data and assessed these data
in combination with new, high-quality rock magnetic, pale-
omagnetic, and absolute paleointensity results that we ob-
tained from lava flows that are presumed to be associated
with the 1631 eruption. Nine to fifteen standard paleomag-
netic cores were obtained from each site using a portable
electric-powered drill that was oriented with a sun compass
prior to removal. A total of 49 samples from four sites were
collected (Fig. 1; modified from Principe et al., 1987). The
cores were later sliced into standard specimen cores (diam-
eter: 2.5 cm; length: 2.1 cm) for laboratory measurement.
It cannot be ascertained that there are any sites obviously
belonging to the same lava flow. In our sampling area, there
was no evidence of previous samplings (i.e., with the field
drilling technique).
2. Rock Magnetic, Paleomagnetic and Paleointen-
sity Measurements
2.1 Continuous susceptibility curves
Three samples per lava flow were heated in air up to
600
◦
C at a heating rate of 10
◦
C/min and then cooled at
the same rate in order to identify magnetic minerals re-
sponsible for any remanent magnetization. Measurements
were carried out using the Bartington susceptibility system
MS2 equipped with a furnace. Curie temperatures were de-
termined using the method of Pr
´
evot et al. (1983). Most
samples yielded evidence of a single ferrimagnetic phase
(Fig. 2(a), sample V47) with a Curie point compatible with
those of Ti-poor titanomagnetites. The heating and cool-
ing curves are reasonably reversible, which attests to the
high thermal stability of samples. A few samples (Fig. 2(a),
sample V48) yielded evidence of two ferrimagnetic phases
during heating and cooling. The lower Curie points range
between 220 and 300
◦
C, with the highest one being about
580
◦
C. Both Ti-rich and Ti-poor titanomagnetites seem to
co-exist in these lava flows.
G. CONTE-FASANO et al.: LAVA IDENTIFICATION BY PALEOMAGNETISM 1063
NRM
5mT
10
15
20
25
30
40
50
100
88va
V51
N
S
EUp
W Dn
70
90
NRM
300
450
475
500
520
540
560
NS
EUp
W Dn
31va
V47
400
350
250
100°C
0 200 400 600
Temperature (°C)
0
100
200
300
400
500
Susceptibility (Arbitrary Units)
V47
0 200 400 600
Temperature (°C)
0
100
200
300
400
Susceptibility (Arbitrary Units)
V48
-2 -1 0 1 2
Applied Field (T)
-8E-6
-4E-6
0
4E-6
8E-6
Magnetic Moment (Am2)
M=29.3 mg
Uncorrected
V47
Mr/Ms=0.25
Hcr/Hc=1.35
0 0.1 0.2 0.3 0.4 0.5
Applied Field (T)
0
4E-7
8E-7
1E-6
Magnetic Moment (Am2)
I
00.10.20.3
Applied Field(T)
0
1E-6
2E-6
3E-6
Magneti c Moment (Am2)
-2 -1 0 1 2
Applied Field (T)
-4E-5
-2E-5
0
2E-5
4E-5
Magnetic Moment (Am2)
M=25.2 mg
Uncorrected
V48
Mr/Ms=0.13
Hcr/Hc=1.64
IRM
IRM
B)
C)
A)
Fig. 2. (A) Susceptibility versus temperature (in air) curves of representative samples. The arrows indicate the heating and cooling curves. (B) Typical
examples of hysteresis loops (uncorrected) and associated isothermal remanence acquisition curves of small chip samples from the studied volcanic
flows. (C) Orthogonal vector plots of stepwise thermal or alternating field demagnetization of representative samples (stratigraphic coordinates). The
numbers refer either to the temperatures in
◦
C or to peak alternating fields in mT. o. Projections into the horizontal plane; x, projections into the
vertical plane.
1064 G. CONTE-FASANO et al.: LAVA IDENTIFICATION BY PALEOMAGNETISM
Table 1. Mean paleodirections of cleaned remanence for Vesuvius lava flows that presumably erupted in 1631. N, Number of treated samples; n,
number of specimens used for calculation; Dec, Declination; Inc, Inclination; k and α
95
, precision parameter and radius of 95% confidence cone of
Fisher statistics, respectively. Note: in Tanguy et al. (1999) the sampling was performed on the same sites as those used by Carracedo et al. (1993),
but the data are different.
Site Age n/N Dec (
◦
) Inc (
◦
) α
95
(
◦
)k
This Study
V-45 1631? 7/9 7.4 66.7 2.4 611
V-47* 1631? 9/13 11.5 61.6 1.5 1124
V-48* 1631? 10/15 13.8 65.1 2.6 340
V-51* 1631? 11/12 15.9 64.6 0.7 3863
Tanguy et al. (1999)
V8 1631? 10 12.0 66.9 1.2 1454
V9 1631? 15 20.1 59.4 1.2 943
V10 1631? 9 16.8 60.4 1.4 1040
V11 1631? 14 15.9 60.7 1.1 1198
V12 1631? 10 19.0 60.4 1.8 634
Carracedo et al. (1993)
V8 1631? 5 23.9 57.5 4.1 350
V9 1631? 11 5.1 61.1 2.4 371
V10 1631? 9 4.6 60.6 2.8 343
V11 1631? 7 20.6 63.7 2.3 709
V12 1631? 7 14.3 57.5 2.0 930
V14 1631? 6 10.2 61.2 1.4 2376
V16* 1631? 9 11.8 65.3 2.7 358
V19 1631? 5 11.3 60.3 2.1 1383
V20* 1631? 5 14.0 62.3 1.5 2477
V27 1631? 4 6.4 66.1 2.4 1473
V31 1631? 5 359.3 68.7 2.2 1171
Gialanella et al. (1993)
1* 1631? 5/9 12 65 2.3 1100
2* 1631? 6/8 14 64 1.6 1675
8* 1631? 6/7 14 64 1.2 3120
9 1631? 4/12 16 57 3.2 559
10* 1631? 6/7 13 63 2.2 961
11 1631? 3/12 4 66 4.8 666
12* 1631? 11/15 14 63 1.9 585
Hoye (1981)
A* 1631? 5/5 14.2 63.4 1.5 2553
B 1631? 5/5 15.6 66.2 3.1 627
ALL* N=11, Dec=13.5
◦
, Inc=63.7
◦
, α
95
= 0.7
◦
,k=3738
*Presumably 1631
2.2 Hysteresis measurements
All samples were subjected to magnetic hysteresis exper-
iments using an AGFM “Micromag” apparatus in fields up
to 1.4 Tesla. The hysteresis parameters (saturation rema-
nent magnetization M
r
, saturation magnetization M
s
, and
coercive force H
c
) were calculated after correction for the
paramagnetic contribution. The coercivity of remanence
(H
cr
) was determined by applying a progressively increas-
ing back-field after saturation. Typical hysteresis plots are
shown in Fig. 2(b); as can be seen, the curves are quite
symmetrical in all cases. Near the origin no potbellied and
wasp-waisted behaviors were detected (Tauxe et al., 1996),
which probably reflects the very restricted ranges of the
opaque mineral coercivities. In the ratio plot of hystere-
sis parameters, samples fall in the pseudo-single-domain
(PSD) grain size region (Day et al., 1977; Dunlop, 2002a,
b). Isothermal remanent magnetization (IRM) acquisition
curves indicate that saturation is reached in moderate fields
of 100–200 mT, which points to some spinels as remanence
carriers (most probably titanomagnetites).
2.3 Paleodirections
The intensity and direction of natural remanent magne-
tization (NRM) of 9–15 samples from each cooling unit
were measured with both JR-5A and JR6 spinner magne-
tometers (sensitivity: approx. 10
−9
Am
2
). The coercivity
and unblocking temperature spectra, stability, and vectorial
composition of NRM were investigated by detailed progres-
sive alternating field (AF) or stepwise thermal demagneti-
zations. AF demagnetization was carried out in eight to ten
steps up to maximum fields of 70 or 100 mT using a Mol-
spin AF demagnetizer, while thermal demagnetization was
carried out in 12–14 steps up 580
◦
C using a non-inductive
Schonstedt thermal demagnetizer.
In total, 39 specimens from four lava flows were demag-
netized, with stable univectorial components being isolated
in most cases (Fig. 2(c)). A small component, probably of
G. CONTE-FASANO et al.: LAVA IDENTIFICATION BY PALEOMAGNETISM 1065
NRM Remaining
TRM(max) = 2.24 A/m
NRM(max) = 3.61 A/
m
Flab=30 microT
200°C
250
300
350
400
450
475
500
520
0.1
0.1
0.2
0.2
0.3
0.3
0.4
0.4
0.5
0.5
0.6
0.6
0.7
0.7
0.8
0.8
0.9
0.9
92vc
V51
200°C
300
350
NRM
NS
EUp
W Dn
400
450
475
500
520
540
560
92vc
V51
NRM Remaining
TRM(max) = 1.40 A/m
NRM(max) = 3.86 A/m
200°C
0.1
0.1
0.2
0.2
0.3
0.3
0.4
0.4
0.5
0.5
0.6
0.6
0.7
0.7
0.8
0.8
0.9
0.9
29vc
V47
Flab=30 microT
300
400
450
475
500
520
540
560
570
580
200∞C
NS
EUp
WD
n
29vc
V47
580
570
560
540
520
500
475
450
400
350
NRM
TRM Gainning
TRM Gainning
Fig. 3. The representative NRM-TRM plots and associated orthogonal diagrams from Vesuvius samples. The notations in the orthogonal diagrams are
the same as in Fig. 2.
viscous origin, was easily removed during the first steps of
demagnetization. The alternating field and thermal treat-
ments, carried out on the same cores, generally yielded
similar results. The greater part of the remanent magne-
tization was removed at temperatures between 520
◦
and
560
◦
C, which indicates low-Ti titanomagnetites as the main
remanence carriers. The median destructive fields ranged
between 30 and 40 mT, which is in agreement with data
from the hysteresis experiments. Few samples showed more
complex or unstable remanence behavior and no primary
magnetization was determined; these samples were there-
fore rejected for further paleomagnetic analysis. The prin-
cipal component analysis (PCA) of remanent magnetization
for individual specimens was carried out using the linearity
test of Kirschvink (1980) using 6–11 points for the mean
direction determination. Site-mean directions were calcu-
lated by vector addition, which gives unit weight to sample
directions. Site-mean directions and their statistical Fishe-
rian parameters are summarized in Table 1 (Fisher, 1953).
2.4 Thellier paleointensity experiments
The method of Thellier (1959) as modified by Coe et al.
(1978) was used to determine absolute paleointensities. For
the experimental study, heating and cooling steps were car-
ried out using a MDT80 furnace (Magnetic Measurement
Ltd.) with the laboratory field set to 30 microTesla. Thir-
teen temperature steps were distributed between room tem-
perature and 580
◦
C. Temperature reproducibility between
two steps was in general better than 2
◦
C. During the experi-
ment five control heating measurements (so-called “pTRM”
checks) were performed after every second step to check the
thermal alteration (Coe et al., 1978).
Thirty-four samples were pre-selected for the Thellier
paleointensity experiment on the basis of stable, one-
component magnetization accompanied with relatively high
median destructive field (MDF) values, elevated blocking
temperature, and reasonably reversible k-T curves. Some
typical Arai-Nagata curves are shown in Fig. 3 (Nagata et
al., 1963). We only accepted determinations that fulfilled
1066 G. CONTE-FASANO et al.: LAVA IDENTIFICATION BY PALEOMAGNETISM
-10
0
10
20
30
This Study
V45
V47 V48
V51
V8 V9 V10 V11
V12 V14 V16 V19 V20 V27
V31
Flows
12
8
912AB10 11
Gialanella et al. 1993 Hoye 1981
DECLINATION (°)
52
56
60
64
68
72
This Study
V45
V47 V48
V51
V8 V9 V10 V11
V12 V14 V16 V19 V20 V27
V31
12
8
912B10 11
A
Gialanella et al. 1993 Hoye 1981
INCLINATION (°)
Flows
Expectedat 1631
Expectedat 1631
Tanguy et al. 1999 ( )
Carracedo et al.
1993
Tanguy et al. 1999 ( )
Carracedo et al.
1993
Fig. 4. Summary of characteristic unit mean paleodirections obtained from the Vesuvius ‘1631’ lava flows. Stars refer to the presumed 1631 eruption.
the following criteria: (1) obtained from at least six NRM-
TRM (thermoreminant magnetization) points correspond-
ing to a NRM fraction larger than one third, (2) yielded a
quality factor (Coe et al., 1978) of about 5 or more; (3) with
positive ‘pTRM’ checks—i.e., the deviation of the “pTRM”
checks was less that 15% (Table 2).
3. Main Results and Discussion
Characteristic magnetization directions were success-
fully isolated for all samples, and flow unit directions were
precisely determined (Table 1), with α
95
ranging from 0.7
◦
to 2.4
◦
. The thermomagnetic curves show that the rema-
nence is carried by Ti-poor titanomagnetite resulting from
oxy-exsolution of the original Ti-rich titanomagnetite dur-
ing flow cooling; this most probably indicates the thermore-
manent origin of a primary magnetization. The unblocking
temperature spectra and relatively high coercivities point to
PSD grains as being responsible for remanent magnetiza-
tion.
With respect to the paleodirections, mean directions can
be compared with a reference secular variation curve (SVC)
for the area. The reference curve for the last few millen-
nia in certain areas of Western Europe (France and Great
Britain, for example) is reasonably well defined (Bucur,
1994; Gallet et al., 2002), in contrast those available for
other areas, including Italy. There have also been consid-
erable developments in the construction of secular variation
curves and global databases (see Jackson et al., 2000; Le
Goff et al., 2002). Archeomagnetic data are still relatively
scarce in Italy and, consequently, the reference curve re-
lies on volcanic data and on relatively recent historical ob-
servatory observations (Evans and Hoye, 2005; Lanza et
al., 2005). The integration of secular variation data de-
rived from different materials and direct observations has
also enabled evaluation of the precision and spatial valid-
ity (recalculation to near and far away sites) of directional
results. The data from volcanic rocks on such directional
results are usually characterized by high angular precision;
nevertheless, sources of discrepancy and reduced accuracy
have been observed at Italian volcanoes (e.g., Tanguy and
Le Goff, 2004; Lanza et al., 2005) that are similar to results
on other areas (e.g., Urrutia-Fucugauchi et al., 2004).
Two of reference secular variation curves currently being
used for geomagnetic secular variation analyses and dating
are available: the SISVC (Southern Italy Secular Variation
Curve) (Incoronato et al., 2002; Incoronato and Del Negro,
2004) and the SIVC (South Italian Volcanic Curve) (Tanguy
et al., 2003; largely derived from Tanguy et al., 1999). As
these reference curves, derived using different procedures,
have been extensively discussed in Incoronato et al. (2002)
(see also Incoronato and Del Negro, 2004 and Conte et al.,
2006), we will briefly summarize only some of the main
arguments here.
The SISVC relies on:
1) Analysis of the entire coercivity/blocking temperature
spectra by subjecting each specimen to complete de-
G. CONTE-FASANO et al.: LAVA IDENTIFICATION BY PALEOMAGNETISM 1067
This Study
Carracedo et al. 1993
Tanguy et al. 1999
Gialanella et al. 1993
Hoye 1981
Presumably 1631 eruption
1037
1885
1651
1646
1631
1537
1536
1139
968
1689
-30
30
50
70
1037
0
1595
1910
1983
1858
1885
1804
1760
1689
1669
1651
1646
1631
1537
1536
1301
1284/1285
113 9
968
Fig. 5. Site-mean characteristic paleodirections from Vesuvius lava flows presumably related to the 1631 eruption with reference to the Southern
Italy secular variation curve (Incoronato et al., 2002). The statistical uncertainties were used to draw the swath of error. Squares refer to historical
measurements of the geomagnetic field in Rome and the stars refer to the presumed 1631 eruption.
magnetization in order to check the presence of mag-
netization components.
2) Information on the magnetic carriers.
3) Separation of the different magnetization components
achieved by PCA. In dealing with lava flows emplaced
a few hundred years ago only, very low maximum
angular deviation have been used in order to separate
primary components of magnetization from secondary
ones that have been acquired in such a small time
interval of a few hundred years only, although this
caused many specimens or sites to be rejected.
4) Relocation of geomagnetic/magnetization directions,
via the inclined geocentric dipole, performed within
a few hundred kilometers only.
The bulk of the data on which the SIVC relies derives
from measurements of partial progressive alternating mag-
netic field demagnetization (PAFD); there are no PCA anal-
ysis, Zijderveld diagrams, and information on magnetic car-
riers available.
We will take the different procedures leading to these
SVCs when using the SISVC in our discussion of the re-
sults. To draw the swath of error on the curve we used the
statistical uncertainties.
Mean directions from 11 flows fall close to the early 17th
century segment. The difference found between the pa-
leodirections retrieved from the reference curve and those
from the pyroclastic lava (unambiguously pertaining to this
eruption) seems to be related to the mechanism of rema-
nence acquisition in the pyroclastic flow. The evidence of
the unsuitability of pyroclastic lava as a basis for determin-
ing a reference direction comes directly from Carracedo et
al.’s (1993) paper. The main problem is that the temperature
of emplacement of the 1631 pyroclastic flow is unknown.
These rocks most likely carry a partial thermoremanence
instead of a full thermoremanent magnetization. Moreover,
rather unstable magnetic behavior was detected during the
thermal stepwise demagnetization due to the some lithic
(clasts) inclusion in the juvenile pumice. Strictly speaking,
only lava flow V48 from the present study matches the cor-
responding segment of the reference curve (Figs. 4 and 5),
although two other lavas, V47 and V51, also show similar
paleodirections close to this reference curve. However, the
absolute paleointensities of the latter differ (Table 2). Site
V45 has paleodirections that differ considerably from the
reference curve. The lava flows from the Carracedo et al.
(1993) study yield highly scattered directional results. Two
sites (V16 and V20) may be also related to the 1631 erup-
tion, while all flows (excepting two sites 9 and 11) studied
by Gialanella et al. (1993) give directions concordant with
the reference curve, as far as is reported in his work. The
same applies for site A of Hoye (1981).
The use (and usefulness) of the reference curves also
presents problems, including correct age correlation, when
distinct curve segments are relatively close—which is the
case for the 10–12th century and 16–17th century segments
of the Italian data. However, the general angular distribu-
tion and tendency of the 11 sites presumably related to the
1631 eruption correspond to the 16–17th century segment
of the Italian data set (Fig. 5). With respect to the other sites,
it seems that only few of them may be directly attributed to
the medieval ages (sites V8 and V11 of Carracedo et al.,
1068 G. CONTE-FASANO et al.: LAVA IDENTIFICATION BY PALEOMAGNETISM
Table 2. First paleointensity results from the Vesuvius. N, The number of NRM-TRM points used for paleointensity determination; T
min
–T
max
, the
temperature interval used; f , g, and q, the fraction of extrapolated NRM used, the gap factor and quality factor (Coe et al., 1978), respectively. Fe is
the individual paleointensity estimate with associated error, VDM and VDMm are individual and average virtual dipole moments.
Site Sample N T
min
–T
max
fg qF
E
± σ(F
E
) VDM VDMm
◦
C(µT) 10
22
Am
2
10
22
Am
2
V47 V29 14 20–580 0.96 0.90 69.24 79.5 ± 0.9 13.7 13.5 ± 0.4
V31 14 20–580 0.96 0.90 46.69 83.2 ± 1.5 14.0
V32 14 20–580 0.95 0.90 97.08 77.6 ± 0.7 13.3
V33 14 20–580 0.96 0.89 60.40 73.9 ± 1.1 12.9
V34 14 20–580 0.93 0.90 82.86 79.5 ± 0.8 13.7
V35 14 20–580 0.94 0.90 79.39 79.6 ± 0.8 13.5
V51 V88 5 350–500 0.51 0.74 4.09 51.3 ± 4.7 8.34 9.24 ± 1.8
V90 5 350–500 0.47 0.74 3.83 53.4 ± 4.9 8.48
V91 6 350–520 0.53 0.78 5.13 57.8 ± 4.7 9.24
V92 9 200–520 0.63 0.85 12.31 76.4 ± 3.3 12.3
V95 6 350–520 0.56 0.77 6.65 48.8 ± 3.2 7.89
Genevey and Gallet, 2002
Chauvin et al., 2000
Jackson et al., 2000
(models)
Age (AD)
Intensity ( T)
(V47)
(V51)
Fig. 6. Evolution of the geomagnetic field intensity in Western Europe
during the last two millennia as assessed from archeomagnetic data
(Chauvin et al., 2000; Genevey and Gallet, 2002) and geomagnetic field
models (Jackson et al., 2000). All data are reduced to the latitude of
Paris, France.
1993 and sites V9, V10, V11, V12 of Tanguy et al., 1999).
In summary, the paleodirections of 11 of the lava flows
from the 24 flows studied coincide within associated un-
certainties with the corresponding segment of the SISVC.
Thus, this major event in Vesuvius volcanic history—the
1631 eruption—can be considered to be an explosive-
effusive eruption.
Eleven samples yielded acceptable paleointensity esti-
mates; these estimates are the first published to date for
Mount Vesuvius (Table 2). The fraction of the NRM f
used for paleointensity determination ranges between 0.47
and 0.96, and the quality factor q from 3.8 to 97.1 (gen-
erally being greater than 5). NRM end points obtained
from the Thellier experiments at each step are reason-
ably linear and point to the origin (Fig. 3); no deviation
of the NRM left direction towards the applied laboratory
field was observed, indicating the high technical quality of
the paleointensity determinations. Obtained VDMs (vir-
tual dipole moments)—9.24±1.8×10
22
Am
2
for site V51
and 13.5±0.4×10
22
Am
2
for site V47—are higher than the
present-day geomagnetic field strength.
Absolute paleointensity studies on recent Italian lava
flows are mainly devoted to the Etna volcano. However, the
geomagnetic significance of the data has recently been seri-
ously questioned by Calvo et al. (2002) and, consequently,
these data cannot be used as reference material. However,
an abundance of good archeointensity data has been re-
cently reported for Western Europe (mainly from France).
Chauvin et al. (2000) and Genevey and Gallet (2002) com-
piled archeointensity results for the last two millennia. In
addition, these authors obtained numerous intensity deter-
minations on well-dated potteries, which has greatly im-
proved the archeointensity reference curve for Western Eu-
rope (Fig. 6). The data derived from the Jackson et al.
(2000) model are also incorporated in this curve. Two pa-
leointensity determinations obtained from Mount Vesuvius
lava flows (V47 and V51, see also Table 2) are reduced to
the latitude of Paris (Fig. 6) as well as all data from Chau-
vin et al. (2000) and Genevey and Gallet (2002). There is
a good agreement with the value obtained from flow V51
and the historical geomagnetic field model (Jackson et al.,
2000) at about 1631 (Fig. 6). It is probable that this lava also
belongs to the 1631 eruption even though it shows slightly
different magnetic declination with respect to the SISVC
(Fig. 4). In contrast, the intensity retrieved from lava flow
V47 differs significantly from the curve, showing an un-
usually high geomagnetic field strength that better matches
with segment on the 8th to 10th centuries. It may be specu-
lated that this lava erupted during the 968–1039 period. Al-
ternatively, this unusually high absolute intensity value may
be due to the problems related to the Thellier paleointensity
determination.
4. Conclusion
The paleodirections of 11 lava flows out of 24 (this
study; Hoye, 1981; Carracedo et al., 1993; Gialanella et
al., 1993; Tanguy et al., 1999) coincide—within associ-
ated uncertainties—with the corresponding segment of the
Southern Italy Secular Variation Curve. Thus, we conclude
that this major event in Vesuvius volcanic history repre-
sented an explosive-effusive eruption.
Acknowledgments. We acknowledge the comments by Dr. Lau-
rie Brown of the journal review board. AG acknowledges the fi-
nancial support of a Conacyt grant 42661.
G. CONTE-FASANO et al.: LAVA IDENTIFICATION BY PALEOMAGNETISM 1069
References
Arn
´
o, V., C. Principe, M. Rosi, R. Santacroce, A. Sbrana, and M. F. Sheri-
dan, Eruptive history, in “Somma-Vesuvius” Quaderni de “La Ricerca
Scientifica” C. N. R., edited by R. Santacroce, 114, 243 pp., 1987.
Bucur, I., The direction of the terrestrial magnetic field in France, during
the last 21 centuries. Recent progress, Earth Planet. Sci. Lett., 87,95–
109, 1994.
Calvo, M., M. Pr
´
evot, M. Perrin, and J. Riisager, Investigating the reasons
for the failure of paleointensity experiments: A study on historical lava
flows from Mt. Etna (Italy), Geophys. J. Int., 149,44–63, 2002.
Carracedo, J., C. Principe, M. Rosi, and V. Soler, Time correlation by
palaeomagnetism of the 1631 eruption of Mount Vesuvius. Volcanolog-
ical and volcanic hazard implications, J. Volcanol. Geothermal Res., 58,
203–209, 1993.
Chauvin, A., Y. Garcia, P. Lanos, and F. Laubenheimer, Paleointensity of
the geomagnetic field recovered on archeomagnetic sites from France,
Phys. Earth Planet. Int., 120, 111–136, 2000.
Coe, R., S. Gromme, and E. A. Mankinen, Geomagnetic paleointensity
from radiocarbon-dated lava flows on Hawaii and the question of the
Pacific non-dipole Low, J. Geophys. Res., 83, 1740–1756, 1978.
Conte, G., J. Urrutia-Fucugauchi, A. Goguitchaichvili, and A. Incoronato,
Paleomagnetic dating of lava flows of uncertain age, Somma-Vesuvius
volcanic complex (Southern Italy), Int. Geol. Rev., 48(4), 349–359,
2006.
Day, R., M. Fuller, and V. A. Schmidt, Hysteresis properties of titanomag-
netites: Grain size and compositional dependence, Phys. Earth Planet.
Int., 13, 260–267, 1977.
Delibrias, G., G. Di Paola, M. Rosi, and R. Santacroce, La storia eruttiva
del complesso vulcanico Somma-Vesuvio ricostruita nelle successioni
piroclastiche del M. te Somma, Rend. Soc. It. Mineral. Petrol., 35, 411–
438, 1979.
Di Girolamo, P., Geotectonic setting of Miocene-Quaternary volcanism in
and around the Eastern Tyrrhenian sea border (Italy) as deduced from
major element geochemistry, Bull. Volcanol., 41, 229–250, 1978.
Dunlop, D. J., Theory and application of the Day plot (Mrs/Ms versus
Hcr/Hc) 1: Theoretical curves and tests using titanomagnetite data, J.
Geophys. Res., 107(B3), 10.1029/2001JB000486, 2002a.
Dunlop, D. J., Theory and application of the Day plot (Mrs/Ms versus
Hcr/Hc) 2: Application to data for rocks, sediments, and soil, J. Geo-
phys. Res., 107(B3), 10.1029/2001JB000487, 2002b.
Evans, M. E. and G. S. Hoye, Archaeomagnetic results from southern Italy
and their bearing on geomagnetic secular variation, Phys. Earth Planet.
Int., 151, 155–162, 2005.
Fisher, R. A., Dispersion on the sphere, Proc. R. Soc. London Ser. A, 217,
295–305, 1953.
Gallet, Y., A. Genevey, and M. Le Goff, Three millennia of directional
variation of the earths magnetic field in western Europe as revealed by
archeological artifacts, Phys. Earth Planet. Int., 131,81–89, 2002.
Genevey, A. and Y. Gallet, Intensity of the geomagnetic field in western
Europe over the past 2000 years: New data from ancient French poteries,
J. Geophys. Res., 107(B11), 2285, 10.1029/2001JB000701, 2002.
Gialanella, R., A. Incoronato, F. Russo, and G. Nigro, Magnetic stratigra-
phy of Vesuvius products. I-1631 lavas, J. Volcanol. Geothermal Res.,
58, 211–215, 1993.
Hoye, G. S., Archaeomagnetic secular variation record of Mount Vesuvius.
Nature, 291, 216–218, 1981.
Incoronato, A., Magnetic stratigraphy procedures in volcanic areas: the
experience at Vesuvius. Palaeomagnetic and Tectonics of the Mediter-
ranean Region. Geological Soc. Special Pub., 105, 367–371, 1996.
Incoronato, A. and C. Del Negro, Magnetic Stratigraphy Procedures at
Etna, in The Mt. Etna Volcano, edited by A. Bonaccorso, S. Calvari,
M. Coltelli, C. Del Negro, and S. Falsaperla, AGU Geophys Monogr.
Ser., 143, 263–271, 2004.
Incoronato, A., A. Angelino, R. Romano, A. Ferrante, R. Sauna, G. Vana-
core, and C. Vecchione, Retrieving geomagnetic secular variations from
lava flows: evidence from Mount Arso, Etna and Vesuvius (southern
Italy), Geophys. J. Int., 149, 724–730, 2002.
Jackson, A., A. Jonkers, and M. Walker, Four centuries of geomagnetic
secular variation from historical records, Philos. Trans. R. Soc. London,
Ser. A, 358, 957–990, 2000.
Kirschvink, J. L., The least-squares line and plane and the analysis of
palaeomagnetic data, Geophys. J. R. Astr. Soc., 62, 699–718, 1980.
Lanza, R., A. Meloni, and E. Tema, Historical measurements of the Earth’s
magnetic field compared with remanence directions from lava flows in
Italy over the last four centuries, Phys. Earth Planet. Int., 148,97–107,
2005.
Le Hon, H., Histoire complete de la grande eruption du Vesuve de 1631
(With a topographic map of Vesuvius lavas from 1631 till 1861), Bull.
R. Acad. Sci. Lett. Beaux Arts, Belg., 20, 483–538, 1865.
Le Goff, M., Y. Gallet, A. Genevey, and N. Warm
´
e, On archeomagnetic
secular variation curves and archeomagnetic dating, Physics of the Earth
and Planetary Interiors, 134, 203–211, 2002.
Nagata, T., R. A. Fisher, and K. Momose, Secular variation of the geo-
magnetic total force during the last 5000 years, J. Geophys. Res., 68,
5277–5281, 1963.
Pr
´
evot, M., R. S. Mankinen, S. Gromme, and A. Leccaille, High paleoin-
tensity of the geomagnetic field from thermomagnetic studies on rift
valley pillow basalts from the middle Atlantic ridge, J. Geophys. Res.,
88, 2316–2326, 1983.
Principe, C., M. Rosi, R. Santacroce, and A. Sbrana, Explanatory Notes to
the Geological Map of Vesuvius. Quaderni de “La Ricerca Scientifica”,
CNR, Roma, 114, 8, 11–51, 1987.
Rolandi, G. and F. Russo, Contributo alla conoscenza dell’attivit
`
a storica
del Vesuvio. La stratigrafia di Villa Inglese (Torre del Greco), Rend.
Acad. Sci. Fis. Mat. Napoli, 54, 123–127, 1987.
Rolandi, G. and F. Russo, Contributo alla conoscenza dell’attivit
`
a storica
del Vesuvio: dati stratigrafici e vulcanologici nel settore meridionale tra
Torre del Greco, localit
`
a Villa Inglese, e Torre Annunziata, Boll. Soc.
Geol. Ital., 108, 521–536, 1989.
Rolandi, G. and F. Russo, L’eruzione del Vesuvio del 1631, Boll. Soc. Geol.
Ital., 112, 315–332, 1993.
Rolandi, G., A. Barrella, A. Borrelli, and G. D’Alessio, The 1631 Vesuvian
eruption, Int. Conf. on Active Volcanoes and Risk Mitigation. Abstr., 91,
1991.
Rosi, M. and R. Santacroce, The famous AD 1631 eruption of Vesuvius:
a revised interpretation in light of historical and volcanological data,
Workshop on Volcanic Blast, Mount St. Helens, abstr., 1984.
Rosi, M., C. Principe, and R. Vecchi, The 1631 eruption of Vesuvius
reconstructed from the review of chronicles and study of deposits, J.
Volcanol. Geotherm. Res., 58, 151–182, 1993.
Santacroce, R., Somma-Vesuvius. Quaderni de “La Ricerca Scientifica”,
CNR, Roma 114, 8, 243 pp., 1987.
Tanguy, J. C. and M. Le Goff, Distortion of the geomagnetic field in vol-
canic terrains: an experimental study of the Mount Etna stratovolcano,
Phys. Earth Planet. Int., 141,59–70, 2004.
Tanguy, J. C., M. Le Goff, V. Chillemi, A. Paiotti, C. Principe, S. La
Delfa, and G. Patan
`
e, Variation s
´
eculaire de la direction du champ
g
´
eomagn
´
etique enregistr
´
ee par les laves de l’Etna et du V
´
esuve pendant
les deux derniers mill
´
enaires, C. R. Acad. Sc. Paris Sci. Terre Plan
`
et.,
329, 557–564, 1999.
Tanguy, J. C., M. Le Goff, V. Chillemi, A. Paiotti, C. Principe, S. La Delfa,
and G. Patan
`
e, Archeomagnetic dating of Mediterranean volcanics of
the last 2100 years: validity and limits, Earth Planet. Sci. Lett., 211,
111–124, 2003.
Tauxe, L., T. A. T. Mullender, and T. Pick, Pot-bellies, wasp-waists and su-
perparamagnetism in magnetic hysteresis, J. Geophys. Res., 95, 12337–
12350, 1996.
Thellier, E. and O. Thellier, Sur l’intensite de champ magnetique terrestre
dans le passe historique et geologique, Ann. Geophys., 15, 285–376,
1959.
Urrutia-Fucugauchi, J., L. Alva-Valdivia, A. Goguitchaichvili, M. L. Ri-
vas, and J. Morales, Palaeomagnetic, rock-magnetic and microscopy
studies of historic lava flows from Paricutin volcano, Mexico: Impli-
cations for the deflection of palaeomagnetic measurements, Geophys. J.
Int., 156, 431–442, 2004.
G. Conte-Fasano (e-mail: gennaro@geofisica.unam.mx), J. Urrutia-
Fucugauchi, A. Goguitchaichvili, A. Incoronato, and P. Tiano
