GeolCarp_Vol62_No1_5

www.geologicacarpathica.sk

GEOLOGICA CARPATHICA

, FEBRUARY 2011, 62, 1, 5—16 doi: 10.2478/v10096-011-0001-3

Introduction

When  the  Somma-Vesuvius  volcanic  complex  explosively
erupted  AD 79  its  adjacent  south-eastern  territory  was  cov-
ered by volcanic deposits of some meters thickness in a very
short period of time. Consequently, not only the Roman set-
tlements  of  Pompeii  and  Herculaneum  but  also  almost  the
entire ancient landscape of the Sarno River plain were buried
and thereby preserved to a certain extent. Since the laterally
extensive  volcanic  deposits  of  AD 79  show  a  specific  and
therefore readily identifiable stratigraphy they can be used as
a  chronostratigraphic  marker.  This  stratigraphy  consists  of
white phonolitic and grey tephritic phonolitic pumice lapilli
fallout  on  top  of  the  pre-AD 79  Roman  surface.  The  grey
pumice  is  interrupted  and  overlain  by  several  ash  layers  of
pyroclastic surge and flow deposits (Sigurdsson et al. 1985;
Carey  &  Sigurdsson  1987;  Civetta  et  al.  1991;  Cioni  et  al.
1992; Lirer et al. 1993; Luongo et al. 2003).

Vogel & Märker (2010) made a detailed review of past pa-

leo-environmental  studies  in  the  Sarno  River  plain  including
Cinque & Russo (1986), Livadie et al. (1990), Furnari (1994),
Pescatore et al. (1999), Di Maio & Pagano (2003) and Stefani
& Di Maio (2003). These studies mostly focussed on the de-
lineation  of  the  coastline  before  AD 79  and  the  paleo-course
of  the  Sarno  River  and  its  estuary  mouth.  Vogel  &  Märker

Revised modelling of the post-AD 79 volcanic deposits of

Somma-Vesuvius to reconstruct the pre-AD 79 topography

of the Sarno River plain (Italy)

SEBASTIAN VOGEL

1,3*

, MICHAEL MÄRKER

1,2

and FLORIAN SEILER

Heidelberg Academy of Sciences and Humanities c/o University of Tübingen, Rümelinstraße 19-23, D-72070 Tübingen, Germany;

sv@dainst.de

Department of Vegetation, Soil, Environment and Agroforestry Sciences, University of Florence, P.zza S. Marco 4, I-50121 Florence, Italy

German Archaeological Institute, Germany, Podbielskiallee 69-71, D-14195 Berlin, Germany

(Manuscript received May 10, 2010; accepted in revised form November 16, 2010)

Abstract: In this study the methodology proposed by Vogel & Märker (2010) to reconstruct the pre-AD 79 topography
and paleo-environmental features of the Sarno River plain (Italy) was considerably revised and improved. The method-
ology is based on an extensive dataset of stratigraphical information from the entire Sarno River plain, a high-resolution
present-day digital elevation model (DEM) and a classification and regression tree approach. The dataset was re-evalu-
ated and 32 additional stratigraphical drillings were collected in areas that were not or insufficiently covered by previ-
ous stratigraphic data. Altogether, an assemblage of 1,840 drillings, containing information about the depth from the
present-day surface to the pre-AD 79 paleo-surface (thickness of post-AD 79 deposits) and the character of the
pre-AD 79 paleo-layer of the Sarno River plain was utilized. Moreover, an improved preprocessing of the input param-
eters attained a distinct progress in model performance in comparison to the previous model of Vogel & Märker (2010).
Subsequently, a spatial model of the post-AD 79 deposits was generated. The modelled deposits were then used to
reconstruct the pre-AD 79 topography of the Sarno River plain. Moreover, paleo-environmental and paleo-geomorpho-
logical features such as the paleo-coastline, the paleo-Sarno River and its floodplain, alluvial fans near the Tyrrhenian
coast as well as abrasion terraces of historical and protohistorical coastlines were identified. This reconstruction repre-
sents a qualitative improvement of the previous work by Vogel & Märker (2010).

Key words: AD 79, Sarno River plain, Somma-Vesuvius, modelling, paleo-environment, paleo-topography, landscape
reconstruction.

(2010) demonstrated that the pre-AD 79 paleo-surface of the
entire Sarno River plain can be reconstructed by using sophis-
ticated geostatistical methods. This methodology is based on
an extensive dataset of stratigraphical information, a high-res-
olution present-day digital elevation model (DEM) and a clas-
sification and regression tree approach.

The objective of this paper is to revise and improve the

spatial  model  of  post-AD 79  deposits  established  by  Vogel
& Märker (2010) as well as the deduced pre-AD 79 paleo-to-
pography  and  selected  paleo-environmental  features  by  us-
ing  an  increased  number  of  stratigraphical  data  and  further
preprocessing the input parameters.

Study area

The Sarno River plain is located south of the volcanic

complex  of  Somma-Vesuvius  and  belongs  to  the  southern
part  of  the  Pliocene-Pleistocene  graben  structure  of  the
Campanian  Plain.  It  has  a  surface  area  of  approximately
210 km

and is drained by the Sarno River and its tributaries.

In the south and in the east the basin is flanked by the Lattari
and Sarno Mountains belonging to the Apennine Mountain
chain. In the west it has an approximately 10 km long coast
line with the Tyrrhenian Sea. The Sarno River plain consists

VOGEL, MÄRKER and SEILER

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 1, 5—16

of marine, alluvial and volcanic deposits lying on top of a car-
bonate  platform  that  reaches  down  to  a  maximum  depth  of
2 km b.s.l. (Cella et al. 2007). During the last 25,000 years the
Sarno River plain has been repeatedly influenced by the perio-
dic  activity  of  the  Somma-Vesuvius  volcano.  Several  explo-
sive (Plinian) eruptions have occurred, each one marking the
beginning  of  a  new  eruptive  cycle  after  a  certain  period  of
quiescence that can last some thousand years (Delibrias et al.
1979; Santacroce 1987; Rolandi et al. 1993a,b; Andronico et
al.  1995;  Civetta  et  al.  1998;  Sigurdsson  2002;  Cella  et  al.
2007). Among these Plinian eruptions the event of AD 79 was
one  of  the  most  destructive  and  today  it  is  one  of  the  most
well-known and most studied volcanic eruptions in history.

Methodology

The methodology applied to reconstruct the pre-AD 79 to-

pography of the Sarno River plain was described in detail by
Vogel & Märker (2010). Most fundamentally, it is based on an
extensive dataset of stratigraphic cross-sections from a total of
1,840 core drillings to gain a representative network of strati-
graphical information for the entire Sarno River plain (Fig. 1).
The collected core drillings were carried out during construc-
tion works, as well as past archaeological and geological stud-
ies.  Compared  to  the  previous  model  of  Vogel  &  Märker
(2010) 32 additional sections from drillings were collected in
the  area  of  Boscoreale,  Stabiae,  Angri  and  Longola-Poggio-
marino.  The  new  drillings  north  of  Angri,  but  also  in
Boscoreale,  cover  an  area  where  stratigraphical  data  were  so
far  missing.  Additionally,  between  2007  and  2009  we  con-
ducted 24 further core drillings in areas with insufficient strati-
graphical information and included them in the modelling.

A characteristic feature of the AD 79 explosive eruptions

of Somma-Vesuvius is a thick layer of white phonolitic pum-
ice  that  was  deposited  upon  a  dark  Roman  paleosol.  More-
over,  no  former  or  later  eruption  of  Somma-Vesuvius  is
comparable with that of AD 79 in terms of magnitude, strati-
graphic  appearance  and  spatial  distribution.  Consequently,
the stratigraphic position of the Roman paleo-surface can be
easily  identified  and  measured  even  though  some  drillings
were  carried  out  without  strict  scientific  control  but  during
construction works.

The core drillings were collected, localized and digitized

using geographical information systems (GIS). The stratigra-
phy of the drillings was determined, above all identifying the
volcanic deposits of the AD 79 eruption, the pre-AD 79 pa-
leo-surface  underneath  and  measuring  the  depth  from  the
present-day  surface  to  the  pre-AD 79  surface  (thickness  of
post-AD 79 deposits). Furthermore, the pre-AD 79 layer was
characterized  and  categorized  distinguishing  five  different
environmental classes which later on allows the reconstruc-
tion  of  some  important  paleo-environmental  features  of  the
Sarno River plain (Vogel & Märker 2010):

(i) Terrestrial deposits
Roman paleosol of loamy sand that is characterized by the

formation of a dark humus A horizon and sometimes by an
initial A/Bw weathering horizon. It is composed of fine

weathered ash, rounded pumice, lapilli and lithic clasts.
Sometimes traces of agricultural use and former vegetation
as well as fragments of ceramics can be found.

(ii) Fluvial deposits
Fluvial deposits consist of conglomerates of black fluvial

gravel within a sandy or loamy matrix. They contain round-
ed pumice and lithic clasts mostly sorted in thin layers of dif-
ferent deposition cycles. Sometimes fragments of travertine
can also be found. They identify the alluvium of rivers,
mainly of the Sarno River network.

(iii) Palustrine deposits
Palustrine deposits consist of loamy substrate intercon-

nected with dark greyish peaty layers rich in organic matter
and plant remains. They are often characterized by hydro-
morphic features indicating pedogenetic processes of gleyza-
tion that are caused by the presence of a high ground water
table. Furthermore they can contain pumice, travertine plates
and shells of freshwater gastropods. These deposits represent
wetlands (bogs and swamps) that are mainly linked to the
floodplain of the Sarno River.

(iv) Littoral deposits
Littoral deposits are associated with the ancient coastline.

They contain well-sorted grey littoral gravel and sand with
rounded pumice and lithic clasts.

(v) Marine deposits
Marine deposits consist of grey marine gravel and sand

that can contain remains of marine organisms (e.g. shells of
saltwater gastropods). They indicate the presence of the Tyr-
rhenian Sea or lagoons that are separated from the active
shore by littoral deposits.

Since both, fluvial and palustrine deposits are closely re-

lated  to  the  paleo-Sarno  River  the  two  classes  were  finally
merged to identify the floodplain of the paleo-Sarno River. If
the  pre-AD 79  layer  could  not  be  clearly  related  to  one  of
these  five  classes,  additionally,  mixed  classes  were  defined
to incorporate intermediate forms of deposits.

Our approach is based on the hypothesis that the eruption

AD 79  mantled  the  ancient  topography  of  the  Sarno  River
plain  leaving  ancient  physiographic  elements  still  recogniz-
able  in  the  present-day  topography  (Sigurdsson  &  Carey
2002; Stefani & Di Maio 2003; Vogel & Märker 2010). Con-
sequently, the present-day topography can be utilized for re-
constructing  the  ancient  conditions.  This  is  supported  by
Ohlig  (2001)  who  states  that  the  general  structure  of  the
Pompeiian  hill  must  have  remained  constant  after  the  erup-
tion of AD 79 apart from a different absolute altitude.

To model the pre-AD 79 topography of the Sarno River

plain  a  classification  and  regression  tree  approach  was  ap-
plied  using  a  high-resolution  present-day  digital  elevation
model (DEM) and 16 deduced primary and secondary topo-
graphic  indices.  Neotectonic  phenomena  that  occurred  be-
fore  AD 79  are  taken  into  account  in  this  modelling
approach because their resulting landforms are incorporated
into  the  present-day  topography.  However,  the  present-day
DEM can also contain structures of recent neotectonic activi-
ties  (after  AD 79)  which  may  by  mistake  be  transferred  to
the  reconstructed  pre-AD 79  topography.  By  means  of  the
stratigraphical information from the drilling data, such as the

MODELLING OF VOLCANIC DEPOSITS TO RECONSTRUCT PRE-AD 79 TOPOGRAPHY OF SARNO RIVER PLAIN

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 1, 5—16

elevation of the pre-AD 79 littoral deposits, major elevation
discrepancies can be identified.

Classification and regression trees is an algorithm used for

exploratory data analysis and predictive modelling to discover
features and structural patterns in large databases by determin-
ing the correlation between predictor variables and a response
variable  (Breiman  et  al.  1984;  Myles  et  al.  2004).  For  the
model generation we used the TreeNet software (Salford Sys-
tems).  As  predictor  variables  we  used  the  present-day  DEM
and  16  deduced  primary  and  secondary  topographic  indices
and  as  the  response  variable  we  took  the  depth  to  the  pre-
AD 79  surface  (thickness  of  post-AD 79  deposits)  (Table 1).

Before entering the input parameters into TreeNet the quality
of  all  1,840  stratigraphical  drilling  data  were  re-evaluated.
Plausibility  checks  were  carried  out  on  every  drilling  point
by  comparison  with  the  surrounding  nearest  neighbours  to
identify  drillings  where  the  AD 79  volcanic  deposits  were
not correctly determined and by that also the pre-AD 79 pa-
leo-surface. Furthermore, the original drilling documentation
was double-checked for drillings whose location was incor-
rect. By means of a statistical evaluation of the stratigraphical
data outliers among the drilling points were detected and elim-
inated. Moreover, the aspect (A) was transformed according to
Beers  et  al.  (1966)  following  A

’= cos (45—A)+ 1. Instead of

Fig. 1. Present-day digital elevation model (DEM) and fluvial network of the Sarno River plain with the location of 1,840 stratigraphical
core drillings. New collected or conducted drillings are indicated in red. Letters mark sites of interest.

Predictor variables

Method/Reference

Elevation

Topo to Raster (ArcGIS 9.2)/SIT 1:5,000

Altitude above channel network SAGA terrain analysis module (Olaya & Conrad 2008)
Analytical hillshading

SAGA terrain analysis module (Olaya & Conrad 2008)

Aspect

SAGA terrain analysis module (Zevenbergen & Thorne 1987)

Catchment area

SAGA terrain analysis module (Olaya & Conrad 2008)

Channel network

SAGA terrain analysis module (Olaya & Conrad 2008)

Channel network base level

SAGA terrain analysis module (Olaya & Conrad 2008)

Convergence index

SAGA terrain analysis module (Köthe & Lehmeier 1993)

Curvature

SAGA terrain analysis module (Zevenbergen & Thorne 1987)

Curvature classification

SAGA terrain analysis module (Dikau 1988)

Plan curvature

SAGA terrain analysis module (Zevenbergen & Thorne 1987)

Profile curvature

SAGA terrain analysis module (Zevenbergen & Thorne 1987)

LS-factor

SAGA terrain analysis module (Olaya & Conrad 2008)

Slope

SAGA terrain analysis module (Zevenbergen & Thorne 1987)

Stream power

SAGA terrain analysis module (Olaya & Conrad 2008)

Watershed subbasins

SAGA terrain analysis module (Olaya & Conrad 2008)

Topographic wetness index

SAGA terrain analysis module (Olaya & Conrad 2008)

Table 1: Predictor variables used for the modelling process in TreeNet (Salford Systems).

the common range of the aspect where
0° and 360° represent identical values
the  aspect  was  transformed  to  range
of values between 0 and 2. However, it
turned out to not significantly improve
the model performance. Several prelim-
inary  model  runs  were  performed  with
different presettings, such as number of
grown trees and by leaving out different
predictor variables. Thus, the sensitivi-
ty of a particular predictor variable with
respect  to  the  model  performance  was
tested to optimize the prediction.

In order to determine the pre-AD 79

topography of the Sarno River plain the
modelled thickness of post-AD 79 de-
posit was subtracted from the present-
day DEM. The pre-AD 79 Sarno River

VOGEL, MÄRKER and SEILER

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 1, 5—16

network  was  reconstructed  by  delineation  of  linear  thalweg
features of the pre-AD 79 topography using the SAGA terrain
analysis  module  (Olaya  &  Conrad  2008).  The  reconstruction
of specific paleo-environmental features of the pre-AD 79 sur-
face such as the ancient coastline or the floodplain of the pa-
leo-Sarno  River  was  done  by  attributing  the  classified
environmental  character  of  the  pre-AD 79  paleo-layer  to  the
modelled pre-AD 79 topography (Vogel & Märker 2010).

Results and discussion

The statistical evaluation of the thickness of post-AD 79

deposits of the drilling dataset reveals a range of values be-
tween  0  and  19 m  (Fig. 2).  Drillings  having  a  thickness  of
more  than  19 m  can  be  statistically  considered  as  outliers
which  apply  to  10  of  the  1,840  drillings.  Disregarding  the
outliers the arithmetic mean thickness of post-AD 79 depos-
its is 5.8 m, the median is at 5.3 m and the standard deviation
is 3.1 m. Furthermore, 95 % of the drilling data are covering
a range between 1.25 and 15 m.

During preliminary model runs the sensitivity test of the

predictor variables pointed out that the topographic index
‘watershed subbasins’ (SAGA terrain analysis module) pro-

duced  a  lot  of  artifacts  on  the  regionalized  thickness  of  the
post-AD 79  deposits.  Consequently,  it  was  left  out  of  the
modelling process. Fig. 3 shows the performance of the ob-
tained regression model. The minimum mean absolute error
of  the  training  dataset  is  0.98 m  and  that  of  the  test  dataset
1.68 m,  which  is  similar  to  the  previous  model  of  Vogel  &
Märker (2010). 2,094 trees were grown to achieve the small-
est  test  mean  absolute  error  (Fig. 3A).  The  general  model
performance  of  the  training  and  test  dataset  is  illustrated  in
Fig. 3B  and  C,  respectively.  The  gain  charts  show  the  per-
centage  of  population  (x-axis)  related  to  the  percentage  of
the target variable (y-axis). The higher the percentage of the
target  variable  with  respective  low  population  percentages,
the better is the model. In this case the current model perfor-
mance shows a clear positive correlation similar to the previ-
ous model (Vogel & Märker 2010).

The modelled thickness of the post-AD 79 deposits of the

Sarno River plain is illustrated in Fig. 4A. It shows a distinct
spatial distribution of post-AD 79 deposits ranging from 1.1
to 15.6 m. This range of values corresponds well to the cal-
culated range of 95 % of the drilling data.

Apart from the internal model validation that is performed by

TreeNet an external validation of the model results was carried
out by comparing the modelled thickness of the post-AD 79

Fig. 2. Histogram of the drilling data regarding the depth to the pre-AD 79 surface.

Fig. 3. A – Mean absolute error [m] of the depth to the pre-AD 79 surface versus number of trees of the training and test dataset. The mini-
mum mean absolute error of the training dataset of 0.98 m and of the test dataset of 1.68 m was attained at 2,094 grown trees. B – Gain chart
of the training dataset. C – Gain chart of the test dataset (TreeNet).

deposits  with  the  real  thickness  given  by  the
1,840  drilling  data  (Fig. 4B).  Figure 4B  docu-
ments an improvement of the new prediction in
respect to Vogel & Märker (2010). It shows that
79 % of the drilling points match the modelled
thickness with an accuracy of less than 2 m and
50 %  of  the  drilling  points  with  less  than  1 m.
The  respective  percentages  of  the  previous
model  were  44 %  and  24 %.  The  mean  thick-
ness of modelled post-AD 79 deposits is 5.6 m
(6.5 m in Vogel & Märker 2010) which lies ex-
actly between the arithmetic mean of 5.8 m and
the median of 5.3 m obtained from the drilling
data. Best agreement between model and drill-
ing data was attained for the plain areas where
the  post-AD 79  deposits  are  rather  thin.  How-
ever,  with  increasing  thickness  of  post-AD 79
deposits and with increasing relief towards the
slopes  of  Somma-Vesuvius  and  the  Apennine
Mountains  the  difference  between  real  and
modelled thickness also increases (Fig. 4B).

VOGEL, MÄRKER and SEILER

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 1, 5—16

distinct spatial distribution of post-AD 79 volcanic deposits
is  most  notably  controlled  by  two  sets  of  processes:  (i)  the
initial deposition of pyroclastic material during the eruption
of the Somma-Vesuvius volcanic complex and (ii) the subse-
quent redistribution of the volcanic material by processes of
erosion, transport and redeposition (Vogel & Märker 2010).
It  is  noticeable  that  within  the  Sarno  River  plain  there  is  a
gradient of post-AD 79 deposits from the slopes of Somma-
Vesuvius  in  the  northwest  to  the  actual  plain  in  the  south-
east.  Thicker  deposits  on  the  southwestern  slopes  of
Somma-Vesuvius can be explained as follows (see also Vo-
gel & Märker 2010):

(i) Originating from the vent of Somma-Vesuvius, AD 79

the  pumice  lapilli  fallout  was  dispersed  concentrically  to-
wards  the  southeast  resulting  in  thickest  deposits  in  the
perivolcanic areas, near the source of the eruption. This cor-
responds  to  isopachs  of  the  thickness  of  the  AD 79  pumice
fallout  as  reconstructed  by  Sigurdsson  &  Carey  (2002)  or
Pfeiffer et al. (2005).

(ii) The pumice fallout is overlain by pyroclastic surge and

flow  deposits  of  the  AD 79  eruption  that  are  especially  con-
centrated at the slopes of Somma-Vesuvius and gradually de-
crease  in  thickness  towards  the  plain  (Di  Vito  et  al.  1998;
Rossano  et  al.  1998;  Di  Maio  &  Pagano  2003).  However,
along  the  south-eastern  slope  of  Somma-Vesuvius  a  band  of
thinner deposits can be noticed (Fig. 4A) that extends concen-
trically  between  50  and  70 m  a.s.l.  (present-day  elevation).
This band may be caused by a downslope acceleration of the
pyroclastic surge due to the steepness of the slope temporarily
compensating  the  deceleration  of  the  surge  with  increasing
travel distance. Hence, the erosive power of the surge is tem-
porarily increased before deposition prevails towards the toe-
slope of Somma Vesuvius (see Rossano et al. 1998).

(iii) Additionally, the slopes of Somma-Vesuvius are cov-

ered by pyroclastic material and lava flows from more recent
eruptions such as AD 472, 1631 or 1944 (Geological map
1 : 10,000, Autorit

di Bacino del Sarno 2003).

Greater thickness of volcanic deposits in the coastal area on

the other hand can be attributed to mobilization of initially de-
posited material by the paleo-Sarno River and its tributaries,

transportation out of the plain and finally re-deposition in
shallow water along the Tyrrhenian coast. This coincides with
rather thin deposits in the central and southern parts of the
Sarno River plain. Steep slopes of 3 to 4 m in height near the
Tyrrhenian Sea give a first idea on the approximate position
of the coastline before AD 79 (Fig. 4A).

An accumulation of deposits on the foot slopes of the Sarno

and Lattari Mountains results from initially deposited volca-
nic material that was eroded by mountain streams or lahars,
for  example  after  intense  rainfall  events  (Lirer  et  al.  2001;
Fiorillo  &  Wilson  2004).  Accordingly,  the  mountain  slopes
from  where  this  material  was  eroded  show  relatively  thin
post-AD 79 deposits. Likewise, the volcanic material on the
Pompeian hill is rather thin on top and thicker along its foot
slopes  because  after  the  deposition  of  the  AD 79  volcanic
material it was eroded from the ridge and accumulated in the
adjacent plain.

It is noticeable that on the western slopes of the Lattari

Mountains  thicker  deposits  appear  within  the  deeply  incised
river valleys such as the ‘Fosso di Gragnano’ draining this part
of the mountains whereas the adjacent slopes show rather thin
deposits. This is supported by Rossano et al. (1998) who state
in the Vesuvius area that the movement and deposition of ash
flows,  lahars  or  debris  avalanches  accompanying  volcanic
eruptions  strongly  follow  topographic  irregularities  such  as
river  channels.  Furthermore,  due  to  the  redistribution  of  the
volcanic material subsequent to the AD 79 eruption paleo-riv-
er valleys were filled with volcanic material that was initially
deposited on the mountain slopes.

The improved digital elevation model (DEM) of the Sarno

River  plain  before  AD 79  was  calculated  by  subtracting  the
modelled  thickness  of  the  post-AD 79  deposits  from  the
present-day  surface.  The  pre-AD 79  paleo-river  network  was
deduced  by  deriving  the  linear  thalweg  features  of  the  pre-
AD 79  DEM  (Fig. 5).  Finally,  the  reconstruction  of  specific
paleo-environmental  features  before  AD 79  was  done  by  at-
tributing  the  classified  environmental  characteristics  of  the
pre-AD 79 layer taken from the drilling data to the pre-AD 79
topography  (Fig. 6A).  Hence,  the  approximate  course  of  the
coastline before AD 79 and the floodplain of the paleo-Sarno
River were delineated by means of the littoral and the fluvial-
palustrine  environmental  information  from  the  drilling  data.
Marine  and  littoral  deposits  that  were  found  inland  from  the
ancient  coastline  identified  the  approximate  location  of  the
protohistorical dune ridges of Messigno (5,600—4,500 yr BP)
and  Bottaro/Pioppaino  (3,600—2,500 yr  BP)  (Cinque  1991)
(Fig. 6B) (Vogel & Märker 2010).

As Vogel & Märker (2010) have already stated the pre-

AD 79 topography corresponds to the present-day topography
since  the  main  topographic  elements  reappear.  However,  the
pre-AD 79  paleo-surface  is  situated  approximately  5.8 m
(mean) below the modern surface resulting in a wide area of
the  Tyrrhenian  coast  lying  below  the  present-day  sea  level.
The course of the fluvial network before AD 79 and today also
have  a  lot  of  similarities  even  though  especially  the  lower
reaches of the Sarno River are regulated and canalized today
whereas  the  paleo-Sarno  River  shows  more  natural  fluvial
patterns  (Fig. 5).  South  of  Pompeii  the  paleo-Sarno  River
flows around the more elevated protohistorical dune ridge of

Table 2: Ranking of importance of the predictor variables used for
the model generation.

Variable Importance

[%]

Elevation above sea level

100.0

||||||||||||||||||||||||||||||||||||||||||

Aspect

93.1

|||||||||||||||||||||||||||||||||||||||

Channel base level

81.6

||||||||||||||||||||||||||||||||||

Altitude above Channel network

59.0

||||||||||||||||||||||||

Slope

52.8

||||||||||||||||||||||

Analytical hillshade

50.6

|||||||||||||||||||||

LS-factor

48.6

||||||||||||||||||||

Curvature classification

46.7

|||||||||||||||||||

Convergence index

45.9

|||||||||||||||||||

Profile curvature

44.3

||||||||||||||||||

Topographic wetness index

44.1

||||||||||||||||||

Stream power

43.8

||||||||||||||||||

Curvature

43.2

||||||||||||||||||

Plan curvature

41.9

|||||||||||||||||

Catchment area

38.5

||||||||||||||||

Channel network

9.8

|||

VOGEL, MÄRKER and SEILER

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 1, 5—16

The paleo-Gragnano River, the main tributary of the ‘conoide
di Muscariello’, before AD 79 flowed to the northwest and en-
tered  the  sea  in  the  north  of  the  alluvial  fan  (Fig. 5B).  The
present-day ‘Fosso di Gragnano’ instead flows to the west and
into the sea in the very southwest around 1,300 m away from
its paleo-mouth before AD 79 (Fig. 5A). The relocation of the
river  channel  might  have  been  caused  by  the  sudden  large
amount  of  sediments  during  the  AD 79  eruption  of  Somma-
Vesuvius or subsequent lahars. This may have led to the burial
of the paleo-river channel of the ‘Fosso di Gragnano’ and thus
to its relocation to the southwest.

Between the western slope of the Pompeian hill and the

southern slope of Somma-Vesuvius volcano another alluvial
fan was identified which is fed by the main discharge of that
area. Cinque (1991) referred to it as ‘conoide di Penniniello’.
It  is  striking  that  the  southwestern,  and  so  presumably
youngest part of the fan lies approximately 2 m higher than
its central or southeastern part. This accumulation may have
been  caused  by  a  high  sediment  load  after  an  eruption  of
Somma-Vesuvius that eventually forced the river to bend to
the south instead of flowing straight to the southwest and di-
rectly  entering  the  sea.  The  contour  lines  show  clearly  that
the stream has already cut into its own deposits. The north-
ern  branch  of  the  protohistorical  dune  ridge  of  Bottaro/
Pioppaino was most likely dissected by fluvial erosion leav-
ing  only  scattered  traces  of  dunal  deposits  (Fig. 7B)  due  to
the morphology of the ‘conoide di Penniniello’.

The pre-AD 79 coastline, prior to the eruption of Somma-

Vesuvius, was reconstructed according to the littoral deposits
from the drilling data and the pre-AD 79 coastal topography.
In its central part the ancient coast runs nearly parallel to the
modern coast in a distance of 1,100 to 1,300 m inland. On the
slopes of Somma-Vesuvius and the Lattari Mountains the pa-
leo-coast and the modern coast are much closer to each other
with distances between 50 and 150 m. Between the mountain
slopes and the plain the alluvial fans of Muscariello and Pen-
niniello result in a northwestwards and southwestwards propa-
gation  of  the  ancient  coastline,  respectively.  Regarding  the
pre-AD 79 topography, the pre-AD 79 coastline is mainly sit-
uated  between  the  —2 m  and  —4 m  contour  lines  in  the  north
near  Somma-Vesuvius  and  between  the  0  and  1 m  contour
line in the south of the plain. This elevation unconformity of
the pre-AD 79 littoral deposits can most likely be attributed to
some neotectonic activity in the study area. Sigurdsson et al.
(1985) and Livadie et al. (1990) also observed a deeper occur-
rence of the Roman littoral deposits in the north of the Sarno
River plain at approximately —4 m a.s.l. which Pescatore et al.
(2001) explained with a more rapid subsidence near Somma-
Vesuvius in comparison with more distal areas.

Converging contour lines in the coastal area of the Sarno

River plain may indicate abrasion terraces that can be related
to  marine  erosion  of  different  coastlines.  Abrasional  scarps
at  the  foot  of  the  Pompeian  hill  and  the  Lattari  Mountains
derive  from  protohistorical  coastlines  (Cinque  1991;  Vogel
&  Märker  2010).  Along  the  pre-AD 79  coastline  marine
scarps  can  be  found  on  the  slopes  of  Somma-Vesuvius  and
in the central part of the coast (Fig. 7B).

Additional marine scarps between the pre-AD 79 and the

present-day coastline refer to more recent processes

(Fig. 7B). Due to the modelling approach, post-AD 79 pro-
cesses have generated those features that are, by mistake, re-
flected in the pre-AD 79 topography. Consequently, they must
be regarded as modelling artefacts. Those terrace features may
be formed during post-AD 79 sea levels. As a result of Plinian
and sub-Plinian eruptions of Somma-Vesuvius, as in AD 472
or 1631 (Rosi et al. 1993; Rolandi et al. 1998) excessive loads
of volcanic sediments were produced that were subsequently
redistributed by the river network and redeposited in the
coastal area. Finally, this caused a further westward propaga-
tion of the coastline.

The floodplain of the paleo-Sarno River was reconstructed

by  combining  the  fluvial-palustrine  deposits  from  the  drill-
ing  data  with  the  pre-AD 79  topography  and  the  deduced
pre-AD 79 paleo-river network. As Vogel & Märker (2010)
already  stated  in  the  previous  model  the  fluvial-palustrine
deposits  fit  very  well  with  the  pre-AD 79  paleo-river  net-
work (Fig. 6A). They are characterized by the 1 to 2 m iso-
lines of the ‘vertical distance to channel network’ index from
the SAGA terrain analysis module (Olaya & Conrad 2008).

Conclusion and outlook

A revised reconstruction of the pre-AD 79 topography of

the Sarno River plain was carried out on the basis of an im-
proved  classification  and  regression  model  of  Vogel  &
Märker  (2010).  The  pre-AD 79  digital  elevation  model  and
the classified environmental characteristics of the pre-AD 79
layer derived from the drilling data was used to reconstruct
paleo-environmental  and  paleo-geomorphological  features
of the Sarno River plain such as the paleo-coastline, the pa-
leo-Sarno  River  and  its  floodplain,  alluvial  fans  near  the
Tyrrhenian Sea as well as abrasion terraces of historical and
protohistorical  coastlines.  However,  as  already  stated  by
Vogel & Märker (2010) this reconstruction must be consid-
ered  as  a  model  of  the  pre-AD 79  conditions  based  on  the
hypothesis  stated  above.  Especially  neotectonic  phenomena
that occurred between AD 79 and today are only in part con-
sidered in the modelling approach, for example, through the
elevation of the pre-AD 79 littoral deposits.

The collection of additional stratigraphical data especially

within areas of the Sarno River plain that are insufficiently
covered by previous core drillings as well as the combina-
tion of the pre-AD 79 DEM with archaeological findings can
enhance the paleo-environmental reconstruction of the Sarno
River plain.

In the next project phase geomorphologically interesting

landforms  such  as  the  alluvial  fan  of  ‘Muscariello’  and  the
floodplain of the paleo-Sarno River will be studied in more
detail  using  geophysical  prospections  to  further  verify  the
pre-AD 79  DEM  and  the  applied  methodology.  Moreover,
the pre-AD 79 DEM and paleo-environment will be utilized
to  further  reconstruct  the  ancient  cultural  landscape  of  the
Sarno  River  plain,  for  example,  the  phenomenon  of  the
Roman farms or the ancient road network.

Acknowledgments: This study is part of the geoarchaeolog-
ical research Project entitled “Reconstruction of the ancient

MODELLING OF VOLCANIC DEPOSITS TO RECONSTRUCT PRE-AD 79 TOPOGRAPHY OF SARNO RIVER PLAIN

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 1, 5—16

cultural  landscape  of  the  Sarno  River  plain”  undertaken  by
the German Archaeological Institute in cooperation with the
Heidelberg  Academy  of  Sciences  and  Humanities.  It  was
partly funded by the German Archaeological Institute, Berlin
Head  Office  (Cluster 3)  and  the  Deutsche  Forschungsge-
meinschaft  (German  Research  Foundation).  We  would  like
to  thank  our  local  project  partners  and  all  their  collaborators
for  their  cooperation,  particularly  the  Autorit

di Bacino del

Sarno, the Soprintendenza Speciale per i Beni Archaeologici
di Napoli e Pompei, the Soprintendenza per i Beni Archaeo-
logici di Salerno e Avellino. We also thank Giovanni Di Maio,
Giovanni Patricelli and Gaetana Saccone for various supports.
Finally, we would like to thank Angus Duncan, Jiří Šebesta,
and Jozef Minár for reviewing and substantially improving the
paper with constructive comments and suggestions.

References

Andronico D., Calderoni G., Cioni R., Sbrana A., Sulpizio R. &

Santacroce R. 1995: Geological map of Somma-Vesuvius vol-
cano. Per. Mineral. 64, 77—78.

Autorit

di Bacino del Sarno 2003: Geological map, scale 1 : 10,000.

CARG Project, Region Campania (in Italian).

Balducci S., Vasellini M. & Verdiani G. 1985: Exploration well in

the ‘Ottaviano’ Permit, Italy; ‘Trecase 1’. In: Strub A.S. & Un-
gemach P. (Eds.): European Geothermal Update. Proc. 3

Int.

Seminar on the Results of EC Geothermal Energy Research,
Reidel.

Beers T.W., Dress P.E. & Wensel L.C. 1966: Notes and observa-

tions. Aspect transformation in site productivity research. J.
Forestry 64, 691—692.

Bianco F., Castellano M., Milano G., Ventura G. & Vilardo G.

1997: The Somma-Vesuvius stress field induced by regional
tectonics: evidences from seismological and mesostructural
data. J. Volcanol. Geotherm. Res. 82, 199—218.

Breiman L., Friedman J., Olshen R. & Stone C. 1984: Classification

and regression trees. Chapman and Hall, New York.

Carey S. & Sigurdsson H. 1987: Temporal variations in column

height and magma discharge rate during the 79 A.D. eruption
of Vesuvius. Geol. Soc. Amer. Bull. 99, 303—314.

Cella F., Fedi M., Florio G., Grimaldi M. & Rapolla A. 2007: Shal-

low structure of the Somma-Vesuvius volcano from 3D inver-
sion of gravity data. J. Volcanol. Geotherm. Res. 161, 303—317.

Cioni R., Marianelli P. & Sbrana A. 1992: Dynamics of the A.D. 79

eruption: Stratigraphic, sedimentological and geochemical data
on the successions from the Somma-Vesuvius southern and
eastern sectors. Acta Vulcanol. 2, 109—123.

Cinque A. 1991: The versilian transgression in the Sarno River plain,

Campania. Geogr. Fis. Dinam. Quat. 14, 63—71 (in Italian).

Civetta L., Galati R. & Santacroce R. 1991: Magma mixing and

convective compositional layering within the Vesuvius magma
chamber. B. Volcanol. 53, 287—300.

Civetta L., Di Vito M.A., De Lucia M. & Isaia R. 1998: Eruptive

history of Somma-Vesuvius. In: Guzzo P. & Peroni R. (Eds.):
Archeologia e vulcanologia in Campania. Conference Proceed-
ings, Atti del. Convegno, Naples, 9—16 (in Italian).

Delibrias G., Di Paola G.M., Rosi M. & Santacroce R. 1979: The

eruptive history of Somma-Vesuvius volcanic complex recon-
structed from pyroclastic successions of Monte Somma. Rend.
Soc. Ital. Mineral. Petrol. 35, 411—438 (in Italian).

Del Pezzo E., Iannaccone G., Martini M. & Scarpa R. 1983: The 23

November 1980 southern Italy earthquake. Bull. Seismol. Soc.
Amer. 73, 187—200.

Del Pezzo E., Bianco F. & Saccorotti G. 2004: Seismic source dy-

namics at Vesuvius volcano, Italy. J. Volcanol. Geotherm. Res.
133, 23—39.

Dikau R. 1988: Entwurf einer geomorphographisch-analytischen Sys-

tematik von Reliefeinheiten. Heidelb. Geogr. Baust. 5, 1—45.

Di Maio G. & Pagano M. 2003: Considerations on the coast line and

the modality of burial of ancient Stabia after the Vesuvian
eruption AD 79. Riv. Studi Pomp. XIV, 197—245 (in Italian).

Di Vito M.A., Sulpizio R., Zanchetta G. & Calderoni G. 1998: The

geology of the South Western Slopes of Somma-Vesuvius, Ita-
ly, as inferred by borehole stratigraphies and cores. Acta
Vulcanol. 10, 2, 383—393.

Fiorillo F. & Wilson R.C. 2004: Rainfall induced debris flows in

pyroclastic deposits, Campania (southern Italy). Engng. Geol.
75, 263—289.

Irollo G. 2005: The Holocene evolution of the coastal area between

Naples and Stabia (Campania) based on geological and archae-
ological data. Doctoral Thesis. Faculty of Mathematical, Phys-
ical and Natural Science. University of Naples “Federico II”,
1—414 (in Italian).

Köthe R. & Lehmeier F. 1993: SARA – Ein Programmsystem zur

Automatischen Reliefanalyse. Z. Angew. Geogr. 4, 11—21.

Lirer L., Munno R., Petrosino P. & Vinci A. 1993: Tephrostratigra-

phy of the AD79 pyroclastic deposits in perivolcanic areas of
Mt. Vesuvio (Italy). J. Volcanol. Geotherm. Res. 58, 133—149.

Lirer L., Vinci A., Alberico I., Gifuni T., Bellucci F., Petrosino P.

& Tinterri R. 2001: Occurrence of inter-eruption debris flow
and hyperconcentrated flood-flow deposits on Vesuvio volca-
no, Italy. Sed. Geol. 139, 151—167.

Livadie C.A., Barra D., Bonaduce G., Brancaccio L., Cinque A.,

Ortolani  F.,  Pagliuca  S.  &  Russo  F.  1990:  Geomorphologic,
neotectonic  and  volcanic  evolution  of  the  coastal  plain  of  the
Sarno  River  (Campania)  related  to  the  settlements  before
AD 79 eruption. Pact 25—13, 237—256 (in Italian).

Luongo G., Perrotta A. & Scarpati C. 2003: Impact of the AD 79

explosive eruption on Pompeii. I. Relations amongst the depo-
sitional mechanisms of the pyroclastic products, the frame-
work of the buildings and the associated destructive events. J.
Volcanol. Geotherm. Res. 126, 201—223.

Marzocchi W., Scandone R. & Mulargia F. 1993: The tectonic set-

ting of Mount Vesuvius and the correlation between its erup-
tions and the earthquakes of the Southern Apennines. J.
Volcanol. Geotherm. Res. 58, 27—41.

Myles A., Feudale R., Liu Y., Woody N. & Brown S. 2004: An intro-

duction to decision tree modelling. J. Chemometr. 18, 275—285.

Ohlig C.P.J. 2001: De Aquis Pompeiorum. In: De Waele J.A.K.E.

& Moormann E.M. (Eds.): Das Castellum Aquae in Pompeji:
Herkunft, Zuleitung und Verteilung des Wassers. Circum-
vesuviana 4, 2001, 1—483.

Olaya V. & Conrad O. 2008: Geomorphometry in SAGA. In: Hengl

T. & Reuter H.I. (Eds.): Geomorphometry: concepts, software,
applications. Elsevier, 1—765.

Pescatore T., Senatore M.R., Capretto G. & Lerro G. 2001: Ho-

locene coastal environments near Pompeii before the A.D. 79
eruption of Mount Vesuvius, Italy. Quat. Res. 55, 77—85.

Pfeiffer T., Costa A. & Macedonio G. 2005: A model for the nu-

merical simulation of tephra fall deposits. J. Volcanol. Geo-
therm. Res. 140, 273—294.

Rolandi G., Maraffi S., Petrosino P. & Lirer L. 1993a: The Ottaviano

eruption of Somma-Vesuvio 8,000 y B.P. a magmatic alternat-
ing fall and flow-forming eruption. J. Volcanol. Geotherm. Res.
58, 43—65.

Rolandi G., Mastrolorenzo G., Barrella A.M. & Borrelli A. 1993b:

The Avellino plinian eruption of Somma-Vesuvio 3,760 y B.P
– the progressive evolution from magmatic to hydromagmatic
style. J. Volcanol. Geotherm. Res. 58, 67—88.

VOGEL, MÄRKER and SEILER

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 1, 5—16

Rolandi G., Petrosino P. & Geehin J. 1998: The interplinian activity

at Somma-Vesuvius in the last 3,500 years. J. Volcanol. Geo-
therm. Res. 82, 19—52.

Rosi M., Principe C. & Vecci R. 1993: The 1631 eruption of Vesu-

vius reconstructed from the review of chronicles and study of
deposits. J. Volcanol. Geotherm. Res. 58, 151—182.

Rossano S., Mastrolorenzo G. & De Natale G. 1998: Computer sim-

ulations of pyroclastic flows on Somma-Vesuvius volcano. J.
Volcanol. Geotherm. Res. 82, 113—137.

Santacroce R. 1987: Somma-Vesuvius. C.N.R., Quaderni de ‘La

Ricerca Scientifica’, Rome 114, 1—251.

Sigurdsson H., Carey S., Cornell W. & Pescatore T. 1985: The

eruption of Vesuvius in A.D. 79. Natl. Geogr. Res. 1, 332—387.

Sigurdsson H. 2002: Mount Vesuvius before the Disaster. In:

Jashemski W.F. (Ed.): The Natural History of Pompeii.
Cambridge University Press, 29—36.

Sigurdsson H. & Carey S. 2002: The Eruption of Vesuvius in

A.D. 79. In: Jashemski W.F. (Ed.): The Natural History of
Pompeii. Cambridge University Press, 37—64.

Stefani G. & Di Maio G. 2003: Considerations on the coast line of

AD 79 and to the ancient port of Pompeii. Riv. Stud. Pompei
XIV, 142—195 (in Italian).

Vogel S. & Märker M. 2010: Reconstruction the Roman topography

and environmental features of the Sarno River plain (Italy) be-
fore the AD 79 eruption of Somma-Vesuvius. Geomorphology
115, 67—77.

Zevenbergen L.W. & Thorne C.R. 1987: Quantitative analysis of

land surface topography. Earth Surf. Proc. Land. 12, 47—56.