-
A conceptual and numerical model for groundwater management:a
case study on a coastal aquifer in southern Tuscany, Italy
Piero Barazzuoli & Monica Nocchi & Roberto Rigati
&Massimo Salleolini
Abstract Ongoing hydrogeological research aims todevelop a
correct management model for the Plio-Pleistocene multi-aquifer
system of the Albegna Rivercoastal plain (southern Tuscany, Italy);
overexploitation ofthis aquifer for irrigation and tourism has
caused seawaterintrusion. The conceptual model is based on eld
andlaboratory data collected during the 19952003 period.Meteoric
inltration and ows from the adjoining carbon-ate aquifer recharge
the aquifer. Natural outow occursthrough a diffuse ow into the sea
and river; articialoutow occurs through intensive extraction of
groundwa-ter from wells. Water exchanges in the aquifer
occurnaturally (leakage, closing of aquitard) and
articially(multiscreened wells). The aquifer was represented by
athree-dimensional nite element model using theFEFLOW numerical
code. The model was calibrated forsteady-state and transient
conditions by matching com-puted and measured piezometric levels
(February 1995February 1996). The model helped establish that
seawaterintrusion is essentially due to withdrawals near the
coastduring the irrigation season and that it occurs above all
inthe Osa-Albegna sector, as well as along the river that attimes
feeds the aquifer. The effects of hypothetical aquiferexploitation
were assessed in terms of water budget andhydraulic head
evolution.
Keywords Groundwater modeling . Coastal aquifer .Salinization .
Water budget . Italy
Introduction
In many coastal areas the growth of human settlements,together
with the development of agricultural, industrialand tourist
activities, has led to the overexploitation ofaquifers. Such
overexploitation induces a rise in thefreshwatersaltwater interface
(seawater intrusion) andthus the degradation of the chemical
quality of ground-water; the problem will be aggravated by the
expected risein sea level associated with global warming (IPCC
2007).This situation occurs in several areas of the
Mediterraneanand will worsen due to the increase in the
residentpopulation and in coastal tourism (Lpez-Geta et al.2003).
The quality and quantity of groundwater resourcesalong the Italian
coasts has been degrading for some time;the impact of the growing
population is alarming,especially in the southern regions, where
45% of the totalresident population lives in coastal zones (Barrocu
2003).
The southern coast of Tuscany (central Italy) is largelyaffected
by seawater intrusion (Bencini and Pranzini 1992,1996; Barazzuoli
et al. 1999; Angelini et al. 2000; Benciniet al. 2001) and by the
consequent deterioration of thequality of groundwater and the local
anomalous accumu-lation of heavy metals (Grassi and Netti 2000;
Protano etal. 2000; Agati et al. 2001); this is due to intense
pumpingfor different purposes (above all irrigation and
domesticuse), especially during summer, when the water demandfor
agriculture and tourism is highest and the naturalavailability of
water is lowest. The problem has beenaggravated in the last few
decades by the progressivedecrease in the potential renewable water
resources ofsouthern Tuscany (Barazzuoli et al. 2002) due to
areduction in total annual precipitation. The coastal plainof the
Albegna River is currently experiencing seawaterintrusion owing to
an irrational exploitation of the aquiferthrough hundreds of wells
of different types and depthsand with different pumping rates. The
deterioration ofgroundwater quality is currently a limiting factor
for localeconomic growth; agriculture has either been
completelyabandoned or has been directed towards crops which
cantolerate saltwater but are of inferior quality.
The intrusion of seawater in coastal aquifers was
rstconceptualized independently by Badon-Ghijben (1889)and Herzberg
(1901) assuming hydrostatic equilibrium,immiscible uids and the
existence of a sharp interface
Received: 2 March 2007 /Accepted: 16 May 2008Published online: 1
July 2008
* Springer-Verlag 2008
P. Barazzuoli :M. Nocchi :R. Rigati :M. Salleolini ())Department
of Earth Sciences,University of Siena,Via Laterina 8, 53100 Siena,
Italye-mail: [email protected].: +39-577-233811Fax:
+39-577-233938
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
between fresh- and saltwater in a homogeneous uncon-ned aquifer.
They found that the depth of the freshwatersaltwater interface
below sea level (zs) is given by:
zs hf fs f
where f is the density of freshwater, s is the density
ofsaltwater, and hf is the elevation of the water table abovesea
level. When the equation is applied correctly, theestimated depth
closely approximates the real one (Chengand Ouazar 1999); it is
still widely used to simulatesaltwater intrusion (Essaid 1990;
Cheng and Chen 2001)and, especially for educational purposes, to
gain clearinsight into the behaviour of fresh and saline
groundwaterin coastal aquifer systems (Oude Essink 2003). Due
tomolecular diffusion and hydrodynamic dispersion, freshand salt
water are actually miscible liquids: the contactbetween the two
uids is therefore a transition zone ratherthan a sharp interface
(Gambolati et al. 1999; Cheng andChen 2001). The situation is
further complicated by thefact that the saltwater intrusion itself
changes the uiddensity, so that this parameter varies in space and
time as afunction of changes in concentration, temperature
andpressure in the uid. Furthermore, the porous mediumitself is
usually stochastically heterogeneous. In order toproperly reproduce
the mechanism of saltwater encroach-ment, a variable density ow and
transport modellingapproach is therefore currently adopted (Voss
and Souza1987; Koch and Zhang 1992; Diersch 1998b; Holzbecher1998;
Bear et al. 1999; Diersch and Kolditz 2002).
The medium- and long-term effects of land manage-ment policies
are difcult to foresee due to interactionamong numerous elements
and variables of differentnature, especially as far as seawater
intrusion is concernedbecause many aspects of this problem are not
completelyunderstood (Custodio and Bruggeman 1987; Custodio
andGalofr 1993; FAO 1997; Bear et al 1999; Cheng andOuazar 2004).
Groundwater management thus requires theuse of numerical models to
test present and alternativeexploitation scenarios taking into
account not onlytechnical aspects but also economic, legal, social
andpolitical ones (Wang and Anderson 1982; Bear andVerruijt 1987;
Emch and Yeh 1998; Custodio and Galofr1993; van Dam 1999; Maimone
et al. 2004; Bear 2004).
In this context, the authors developed a long-term
hydro-geological research program, the preliminary results of
whichwere published by Angelini et al. (2000) and Barazzuoli etal.
(2003, 2004). This work presents a conceptual andnumerical model
for simulating the hydrodynamics of themulti-aquifer system of the
lower Albegna River valley. Themodel will be used for hydrochemical
simulations andthe correct management of local water resources.
Thisaquifer system provides a good example of the situation
inrecent coastal plains (Custodio 2002; Morell 2003).
The study area
The Albegna River coastal plain, located in southernTuscany,
consists of aeolian and alluvial sediments covering
an area of about 100 km2 (Fig. 1). The river crosses the
plainfrom ENEWSW with an average annual discharge ofabout 5 m3/s at
the mouth. The area north of the Osa Riveris characterized by
gentle hills alternating with short atsectors where there are
sometimes both natural and articialditches resulting from land
reclamation works. The highestpeaks, no more than 354 metres above
sea level (m asl), arein the south-eastern sector (Poggio del
Leccio hill).
The average annual precipitation in the plain is about630 mm,
and the average annual temperature is about 16C.The effective
precipitation is rather low, varying from150 mm/year in the low
plain to 250 mm/year in thesurrounding hilly areas. As occurs along
the entire coastlineof southern Tuscany, most precipitation is
returnedto the atmosphere through actual evapotranspiration, withan
average of more than 70% (Barazzuoli et al. 1993).According to the
climatic classication proposed byThornthwaite (1948), the
investigated area can be consid-ered subarid C1 (moisture index
from 33.30).
This area is a tectonic depression made up of
continental,transition andmarine sediments (MioceneQuaternary).
Thesecover a pre-Neogene substratum composed of Liguride units(an
argillaceous-calcareous-ysch complex), cropping out onthe northern
and eastern borders of the plain, as well asTuscan units on the
southeastern and northern borders(Mancini 1960; Tozzi and Zanchi
1987; Bonazzi et al.1992; Bossio et al. 2004). Miocene (essentially
Messinian)sedimentary sequences consist of strongly eroded
con-glomerates that can be found only in the eastern
sector.Pliocene sediments prevalently consist of clay, togetherwith
regressive sands, gravel and conglomerates. In theeastern sector,
the limestones, lacustrine clays andtravertine deposits formed in
the Pleistocene are inter-ngered with the terraced Albegna River
deposits consist-ing of gravels, sands and conglomerates. The
aeoliansands and nest uvial deposits (clays and silts) date backto
the Holocene period.
The outcropping rocks can be divided into two maingroups with
different permeability through a qualitativeclassication according
to formation:
Quaternary and Neogene complexes. These depositshave weak or
non-existent cementation, and showpredominantly primary
permeability due to interstitialporosity. The degree of
permeability varies: it ismoderate-high in the Neogene
conglomerates, sand-stone and sand, travertine, terraced alluvial
deposits,shores, and in the horizons of coarse aeolian
sediments,but zero to low in the Pliocene clays,
transitiondeposits, actual and recent alluvial deposits, and inthe
horizons of ne aeolian sediments.
The pre-Neogene complex. Characterized by diageneticformations
showing predominantly secondary perme-ability due to ssuring or
ssuring and karst. Thedegree of permeability varies: it is
medium-high in theTriassic dolomitic limestones and
Cretaceous-Eocenecalcarenites, but zero to low in the
Palaeozoic-Triassicmetamorphic rocks and Cretaceous-Eocene
argilla-ceous calcareous and arenaceous ysch.
1558
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
Conceptualization
Geometry and structureThe main geometric-structural and
hydrogeological charac-teristics of the Plio-Pleistocene
multi-aquifer system of theAlbegna River coastal plain were
reconstructed on the basisof the general geologic reconstruction
and the 61 well/borehole and 31 geo-electric data points (Fig. 2).
The systemis made up of several gravely and sandy layers which
canbe combined into three main aquifer layers; these aquifersare
generally separated by aquitards composed of clayeydeposits with
silt or sand in variable proportions, but theysometimes combine to
form a single-layer aquifer. Another
aquitard consisting of sandy-clayey silt is present at the
top.The upper aquifer layer (SE), consisting mainly of
well-sortedsand with a hydraulic conductivity (K) of 104105
m/s,extends only up to 6 km inland from the coast. It overlies
theA2 aquitard consisting of clay and silty clay. The interme-diate
aquifer layer (SG) consists of gravely sediments with asandy-clayey
matrix that have a very heterogeneouslateral distribution due to
shifts in the course of theAlbegna River;K varies from 103 to 105
m/s, accord-ing to grain size. In almost the entire area covered
bythe model, the SG layer lies above the rather thick(about 2025
m), clayey A3 aquitard that separates itfrom the lower aquifer
layer (GL). The latter is made up
Fig. 1 Geological-hydrogeological sketch map of the Albegna
River coastal plain in southern Tuscany, Italy
1559
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
of sandy-clayey gravel of constant thickness (810 m),with K
ranging from 104105 m/s. The hydraulicconnection between SG and GL
is articially ensured bynumerous multiscreened wells, especially
along thecoast where they usually reach a depth of 4050 m.
This hydrogeological system overlies an imperviousclayey
basement, except at the borders of the plain, wherehydraulic
connections with the outside are possible throughpermeable rocks.
In the south-eastern sector, water exchangesbetween the cavernous
limestone (K of 103 to 104 m/s;Nocchi 2002; Nocchi and Salleolini
2007) and the aquiferlayers (especially outow from the limestone
towards thealluvial deposits) occur through faults, as conrmed
byhydrodynamic and hydrochemical evidence (Angelini et al.2000;
Nocchi 2002, 2004). A similar situation, but with areduced ow of
water, also occurs in the northern sector,where the fault along the
Osa River causes the yschformations and cavernous limestone to crop
out; further-more, at the Melosella locality, the GL aquifer layer
liesdirectly above arenaceous formations (K of about 105 m/s).
The aquifer system is mostly recharged by the directinltration
of precipitation falling in the modelled area; itis also recharged
by inltration into the aquifer layersoutside the modelled area
(lateral ows, in the northernand eastern sector) and by the ow of
water through thecontact alluvial deposits and the cavernous
limestone(lateral and vertical ows in the south-eastern sector).
Thenatural outow is discharged into the sea and the AlbegnaRiver;
the intensive withdrawal from the aquifer throughwells is the
articial outow. Water exchange within the
aquifer occurs due to both natural (leakage, close ofaquitards)
and articial causes (multiscreened wells).
Piezometric surface and hydraulic propertiesThe piezometric
level in 62 wells was measured monthlyfrom February 1995 to
February 1996 (Angelini et al. 2000).Note that, due to continuous
groundwater extraction fromthe various wells in the area, it was
only possible tomeasure the dynamic water level; hydraulic head
data thusdo not reect the natural equilibrium conditions,
especiallyin the summer months, when the demand for waterincreases
due to irrigation and tourism. Since both freshand salt water occur
in the area, a method similar to that ofPost et al. (2007) was used
to convert water levelmeasurements to fresh water heads; a maximum
correctionof about 0.15 m was obtained in the higher salinity
zones.
It was impossible to complete a detailed study of
thehydrodynamics of the Plio-Pleistocene aquifer due to thelack of
hydraulic head measurements at various depths.Nevertheless, the
partial stratigraphic continuity of theaquifer layers, the
widespread presence of semiperviousinterlayers, and the wide
variety of connections between thedifferent layers due to the
hundreds of multiscreened wellseffectively guarantee the hydraulic
continuity within theaquifer complex. This continuity was conrmed
by the goodlocal correlation among hydraulic head measurements.
Eachpiezometric measurement therefore corresponds to the
totalhydraulic head of all aquifer layers constituting a
singlehydrological system in hydraulic equilibrium, and piezo-
Fig. 2 Schematic cross-sections and conceptual model of the
multi-aquifer system in aeolian and alluvial sediments of the
Albegna Rivercoastal plain (redrawn from Nocchi 2004). The vertical
scale is exaggerated about 20 times
1560
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
metric surfaces were reconstructed using hydraulic
headmeasurements from all observation wells.
Collected data show that the aquifer is unconned (orsemi-conned)
in the north-eastern and southern sectors, butsemi-conned (or
conned) downstream due to the presenceof semipervious (or
impervious) covers and/or intercala-tions. The reconstructed
piezometric surfaces (Fig. 3) showthat withdrawals strongly affect
groundwaters, causingwide areas of negative hydraulic head in the
main welleld. Major depressions southeast of Fonteblanda and ENE of
Albinia are well below sea level from spring toautumn, with a
maximum depth of 7 m above sea level(asl) and maximum extension in
July. In June and July thewhole coastline between the Osa River and
Albinia, as farinland as the conuence Albegna-Magione Radicata
(about6 km from the coast), is marked by a negative hydraulichead;
this depression does not spread southeastward owingto the important
ow of water from the cavernouslimestone. Groundwater always has a
centripetal movementtowards the valley bottom; the various ows
originate inthe hills surrounding the plain and move towards
theAlbegna River (the main groundwater drainage axis),through which
the waters nally reach the sea. Hydraulichead measurements carried
out in September 2002 andFebruary 2003 yielded piezometric surfaces
very similar tothe previous ones (Nocchi 2002, 2004).
The hydraulic properties of the aquifer are hardly
known.Angelini et al. (2000) report on the only pumping tests
carriedout in the area (Fig. 4): tests A and B refer to wells
screened
in all aquifer layers, which have an average
hydraulictransmissivity (Ta) of 5.710
3 m2/s, average hydraulicconductivity (Ka) of 5.610
4 m/s and average storativity(Sa) of 310
3; test C refers to a well only screened in the SEand SG
Pleistocene layers (T=9104 m2/s,K=3.8105 m/s,S=3104). The authors
estimate the aquifer transmissivitydistribution by multiplyingKa
and the saturated thicknessmatrix; the relative map (see Fig. 4)
shows transmissivityvalues ranging from zero, where the pre-Neogene
basementcrops out, to 21103 m2/s in the Albinia area, with
anaverage value of 8.5103 m2/s. This reconstruction wasconrmed
indirectly and locally by means of the goodcorrelation between
transmissivity and the normalizedtransverse resistance obtained
through vertical resistivitysounding in the Osa-Albegna coastal
sector following theprocedure proposed by Ahmed et al. (1988), and
inagreement with the direct relationships derived by Urish(1981),
Ponzini et al. (1983), Gorman and Kelly (1990) andBarazzuoli et al.
(1999).
Groundwater qualityThe main physicalchemical characteristics of
groundwa-ter were dened through electrical conductivity
surveyscompleted at the time of the piezometric surveys andanalysis
of water samples from 38 wells to determine Cl
and NO3 concentrations (Angelini et al. 2000); further-
more, 12 samples of groundwater, seawater and thermo-mineral
spring water were taken to estimate major ion
Fig. 3 Groundwater level contour map for the Plio-Pleistocene
aquifer (redrawn from Angelini et al. 2000): a February 1995; b
July 1995
1561
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
concentrations. The distribution of chloride (Fig. 5)reveals how
the quality of groundwater has deterioratedsubstantially throughout
the coastline up to about 4 kminland from the shore, where local
values of 2,000 mg/Lare reached even in spring; lower values of 100
mg/Lwere observed in the southeastern sector (where lowerelectrical
conductivity values were also observed), con-rming the ow of
freshwater from the cavernouslimestone. The waters coming from
wells screened in thelower aquifer layer (Pliocene gravel and sand)
are thosewith the highest chloride concentrations; this
layertherefore seems to be the one most affected by
seawaterintrusion. These wells are located in the sector facing
thecoast, where the piezometric depression is greatest, andtheir
structure (with more screens) facilitates seawaterintrusion within
the aquifer system. Moreover, based onchemical and isotopic data,
Bencini and Pranzini (1996)suggested that the cation exchange
process and inltrationof sulphate river waters (originated by
dissolution of thegypsum formation and mixing with thermal waters)
may
be another cause of groundwater salinization in theAlbegna
plain.
Due to the high correlation observed between
electricalconductivity and chloride concentrations, in
September2002 only electrical conductivity was monitored in
surfacewater and groundwater (Fig. 6). Measurements show
thatseawater intrusion mostly affects natural and
articialwatercourses, even inland from the sea, whereas
theelectrical conductivity (and thus the salinity) of ground-water
is generally lower, even near the coast (Nocchi2002). At the Barca
del Grazi locality, the followingvalues were recorded for the
Albegna River: 12,565 S/cm(at 20C) at low tide and 29,135 S/cm at
high tide.
In March 2003, surface water was sampled in four riversections
in order to determine major ion concentrations(see Fig. 6).
Analytical data were processed according toPiper (1944) and
compared with the water quality of theSaturnia thermomineral spring
(Fig. 7), which dischargesinto the upper course of the Albegna
River, largelycontributing to its baseow (0.30.6 m3/s; Mancini
Fig. 4 Hydraulic transmissivity map of the Plio-Pleistocene
aquifer (modied after Angelini et al. 2000). The gure shows the
location ofpumping tests (star with upper-case letter)
1562
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
1960; Fanelli et al. 1982). Albegna River water in thisstretch
has the same earthy alkalinesulphate compositionof the
above-mentioned thermomineral water, but lowerabsolute
concentrations (Nocchi 2004); Patrignone water(sect. 3), with its
earthy bicarbonatealkaline composition,has very different
characteristics.
Role of surface waterThe 20022003 eld surveys were carried out
with theaim of better understanding the interaction betweensurface
water and groundwater. The investigated parame-ters therefore also
include (Nocchi 2002, 2004): the depthof the river bottom
(comparing data provided by the Osa-Albegna Land Reclamation
Consortium with that reportedon 1:5,000 scale regional technical
cartography), therelative water level, and the discharge rate (by
placing aspinner in the previously mentioned sections of theAlbegna
River, see Fig. 6).
The water level of the Albegna River in the investigatedarea was
everywhere lower than the piezometric level inwinter, conrming the
role of the river as the aquiferdrainage axis (the situation is
locally reversed in the summermonths by withdrawal from wells near
the river). Further-more, in its nal stretch (from the conuence
with theMagione-Radicata Stream to the mouth), the river bottom
ofthe Albegna River is below sea level; consequently, seawater
ascends the river and determines the free-surface water
level. Electrical conductivity measurements at the Barca
delGrazi locality corroborate this theory (see Fig. 6).
Discharge measurements in the river were completed inthe absence
of precipitation, so that watercourses receivedno runoff and only
the baseow was measured; results aresummarized in Fig. 7. Collected
data reveal a substantialincrease in discharge downstream, with
greater drainage inthe stretch between sections 4 and 1 (about 100
l/s/km),where the difference between the height of the
piezometriclevel and that of the water surface is greatest.
Numerical modelling
SoftwareThe groundwater numerical ow model was developedusing
FEFLOW (nite element subsurface ow system)working under both
steady-state and transient conditions.For theoretical and practical
information concerning soft-ware use and the solution of equations,
the reader can refer tothe respective manuals (Diersch 1998a, b).
The niteelement method was adopted for its exibility and capacityto
simulate complex geometric forms and to rene thenodal grid around
points and/or single lines (observationpoints, coastline, etc.).
Geolithological, hydrogeologicaland hydrochemical data processing
was carried out in aGIS environment (ArcView) that can be totally
interfacedwith FEFLOW; this was very useful in the development
of
Fig. 5 Chloride concentration contour map for the
Plio-Pleistocene aquifer (redrawn from Angelini et al. 2000): a
MarchApril 1995;b OctoberNovember 1995
1563
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
the conceptual model, in the creation of the numericalmodel and
in the analysis of simulation results.
Discretization and boundary conditionsThe aquifer system was
discretized using a grid of triangularelements made up of 6,794
nodes and 13,231 elementsranging from 80 to 170 m in size and
covering an area ofabout 90 km2. A higher degree of renement was
adoptedfor the control points (observation wells, which wereplaced
in the grid as xed nodes representing their trueposition) and along
the river and coast (important bound-aries for the correct
simulation of ow). The three-dimensional grid consists of six
layers corresponding tothe above-mentioned hydrogeological layers;
it is therefore
made up of nearly 80,000 linear prismatic elements with
atriangular base, for a total volume of about 4.3 km3.
Thereconstructed geometry matches that represented in geo-logical
sections constructed by Angelini et al. (2000). Theboundary
conditions assigned to the numerical modelderive directly from the
conceptual reconstruction of theaquifer system (Fig. 8).
A constant head value was assigned to nodes along thecoastline
(or lagoon), where groundwater is in contact withthe free surface
of the sea. A hydraulic head of 0 m asl wasassigned to grid nodes
coinciding with the topographicsurface; starting from the
underlying grids (along the verticalof the coast line), the
overpressures deriving from the higherdensity of sea water were
calculated considering a typicalfreshwatersaltwater interface
geometry (Diersch 1998c),
Fig. 6 Electrical conductivity measured in surface waters and
groundwaters during the September 2002 survey (redrawn from
Nocchi2004); surface water values at Barca del Grazi refer to low
tide (12,565 S/cm) and high tide (29,135 S/cm). The gure shows the
riversections in which discharge measurements and water sampling
for chemical analysis were completed in March 2003; only sections
1, 2, 3and 5 were sampled
1564
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
and results were assigned to the nodes as head increments.This
procedure was adopted because of the lack ofinformation on the
density of uids at various depths(obtainable only through specic
soundings and surveys)the greatest limitation of the model.
In the modelled area, the Albegna River is in contact withthe
aquifer; the specied head was therefore also assigned tothe nodes
representing the course of the river. In this case,constant
hydraulic heads were established on the basis of theheight of the
river bottom and the average height of the freesurface of water
measured in sections and interpolated alongthe river; head values
were considered constant throughoutthe year due to the rivers
regular and substantial baseowwith small hydrometric oscillations.
The test application ofvariable functions to the river nodes always
produced scant
Fig. 7 Results of the March 2003 survey for river sections
shownin Fig. 6 (redrawn from Nocchi 2004): a Piper diagramthe
crossrepresents the Saturnia thermomineral spring (values by
Fanelli etal. 1982), which discharges into the upper course of the
AlbegnaRiver (about 30 km from the coast)]; b discharge values
Fig. 8 Boundary conditions of the numerical model
1565
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
variations in the piezometric surface, in terms of both leveland
spatial distribution.
Based on head data, the following ux conditions wereassigned to
the boundaries crossed by water from outsidethe modelled area: uxes
of 4106 m/day from the northand 7106 m/day from the east, assigned
to layers GLand SG respectively, simulate meteoric inltration in
theirareas of recharge outside the model; a ux of 2106
m/day from the southeast, assigned to layer SG, representsthe
contribution of the local carbonate formation. East ofthe lagoon,
where the permeable alluvial sediments liedirectly above the
cavernous limestone, water exchangesbetween the two aquifers were
simulated through avertical inow at the base of the model (see Fig.
2) thatvaries according to piezometric relationships. Where
nocondition was specied, the model uses no-ow boundaryconditions
(impervious basement, no evidence of ow).
Flow conditions were also used to assign inows oroutows, which
vary in time and space; they were thus used
to simulate meteoric inltration in the model (arealrecharge).
Effective precipitation (Pe) was initially estimatedaccording to
Thornthwaite and Mather (1957) on an annualor monthly basis,
depending on simulation conditions(steady-state or transient);
inltration was then estimatedas the percentage of Pe according to
soil type and on thebasis of hydrologic classications proposed by
Favi andRossi (1991) and Civita et al. (1999). The
percentagecalculations were rened through calibration as
follows(Fig. 9): 96% for the sandy deposits cropping out in
thesouthern and western sectors (hydrologic group A: soilswith high
inltration rates even under conditions of fullimbibition and with
high transmission rates); 576474%for sandy sediments cropping out
in the northern sector(usually belonging to group B: moderate
inltration andtransmission rates); 21% for the river terrace
outcrops inthe eastern sector (group C: low inltration and
transmis-sion rates); 2% for the ner deposits of the Albegna
River(group D: very low inltration and transmission rates).
Fig. 9 Percent meteoric inltration zoning with respect to the
effective precipitation assigned to the model; letters (A, B1,
etc.) representsthe soil hydrologic group according to Favi and
Rossi (1991)
1566
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
The discharge condition was assigned to the wells whichexploit
the aquifer; water extractions at individual nodessimulate
withdrawal, which was estimated using statisticaldata and
distributed according to the density of wells. Theability of FEFLOW
to simulate hydraulic connectionbetween aquifer layers through
multiscreened wells wasused to determine discharge rates; a very
high hydraulicconductivity is assigned to the well node, and the
calculateddischarge is automatically distributed according to
thepermeability of the various layers. Water requirements inthis
area are determined chiey by agriculture needs, whiletourism needs
are met by the aqueduct, which exploits otheraquifers. Due to the
lack of reliable data on surface water andgroundwater withdrawals
for agriculture purposes, thesequantities were estimated on the
basis of informationprovided by the last censuses completed by the
Central
Institute of Statistics and processing the following
data(Bilardo et al. 1997): in the agricultural sector, the
extensionof cultivated lands according to crop type, water
require-ments and the irrigation system; in the zootechnical
sector,the type and number of livestock and water
requirements.Withdrawals for irrigation of back gardens and parks
andfor industrial purposes were not taken into consideration,since
they are not signicant in the investigated area.
CalibrationThe main objective of the calibration stage is to
obtainresults as much as possible in agreement with the elddata by
acting on the variables which characterize thesystem. The
parameters calibrated with the model were the
Fig. 10 Hydraulic conductivity zoning (K) within aquifer layers:
a upper SE (1) layer, b intermediate SG (2 and 3) layer, and c
lower GL(4 and 5) layer. The blank zone within the model boundary
represents a non-existent aquifer (aquifer layers are not
continuous throughoutthe modelled area). The table shows the
assigned value and the range of K values. For the aquitard,
Kvert=510
8 m/s
1567
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
hydraulic conductivity and specic storage of aquiferlayers and
the vertical hydraulic conductivity (Kvert) ofaquitards; aquifers
were further subdivided according totheir hydrogeological
characteristics.
The model was initially calibrated for steady-stateconditions,
taking into consideration the piezometric levelmeasured in February
1995 by Angelini et al. (2000); asthe latter is the seasonal
maximum, it was possible toexclude withdrawals from the initial
uncertainties. Thefew known values of hydraulic properties were
used asinput parameters; by varying the hydraulic
conductivity,simulations were carried out so as to achieve the best
tamong estimated and measured piezometric values for asufcient
number of control points. The PEST software(Doherty 1998), which is
able to optimize the parametersof any model, was used to achieve
optimum calibrations
for different starting conditions. The obtained results areshown
in Fig. 10, together with the adopted subdivision ofthe aquifer
layers.
The results of this calibration are plotted in Fig. 11,which
shows the good t between measurements andcalculations in the 60
observation wells inside themodelled area; the two sets of values
are highly correlatedand plot very close to the perfect
correspondence line. Theaverage absolute difference is about 0.50
m; this is ahighly satisfactory value, since the conceptual
modelcannot take into account local changes occurring in thereal
system, and hydraulic head values were determinedon the basis of
altimetry reported in the 1:5,000 scaleregional technical
cartography. The same considerationcan be made when comparing the
simulated piezometricsurface with the one measured in February 1995
(see also
Fig. 11 a Groundwater level contour map and b calibration curve
obtained for steady-state conditions (comparison is made with
February1995 values for 60 observation wells)
1568
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
Fig. 3a): the general framework is very similar, withpractically
the same gradients and heads.
The model was subsequently calibrated for transientconditions.
In this case, the simulation calculates piezo-metric surface
variations in time; it is therefore possible tocalculate the
parameters coming into play in the temporalequation (specic yield
or specic storage). The evolutionof inows (ux at boundary, meteoric
recharge) andoutows (withdrawals) at aquifer boundaries was
alsotaken into consideration. The aim was to simulatepiezometric
oscillations registered at the control pointsduring the observation
period as accurately as possible. Thesimulation reference period is
the one in which groundwa-ter level measurements were carried out;
the computedpiezometric surface in steady-state conditions was
assignedas the initial condition, and comparison was made on
thesame days in which measurements were completed. Thesimulation
started on 19 February 1995 and ended on 28February 1996, lasting a
total of 375 days.
The hydraulic heads assigned to the sea were constant(since the
small tidal oscillations of 0.150.20 m were not
taken into consideration), as were the ones assigned to
theriver, since no registration of its level is available.
Lateralows were also assumed to be constant in time, whereas
thevertical ow coming from the cavernous limestone
increasedprogressively and peaked in July, when the local wells
reachthe piezometric minimum (this simulates the recharge
effectthat occurs here in August, i.e. before meteoric
inltration).This simulation uses the wells which pump water mostly
forirrigation and livestock farming; since irrigation starts
inMarch and continues through to June or September, depend-ing on
the crop type, withdrawals vary in time and wereassigned
considering piezometric uctuations. Meteoricrecharge was assigned
by applying the mentioned inltrationpercentages to the effective
monthly precipitation; themonthly inltration value was distributed
evenly throughoutthe respective month.
The adopted specic yield of 0.12 was determined byestimating
porosity through the empirical relationshipproposed by Archie
(1942) and considering a specicretention of 0.10 (Angelini et al.
2000). Using this method,Gabbani and Gargini (1992) found a similar
value for
Fig. 12 Specic storage zoning. The gure indicates the location
of observation wells shown in Fig. 13 (dot with number)
1569
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
another alluvial coastal plain of southern Tuscany. Thisvalue
was assigned to all aquifer layers because of its minorinuence on
simulation results. Specic storage is a highlysignicant parameter,
because calculations were completedfor a conned aquifer. Figure 12
shows the calibrationresults for this parameter, which depend
essentially on thesystems degree of connement and thus on the
grain-sizedistribution of the cover sediments (see also Fig.
9).
The results of the calibration under transient conditionsare
shown in the plots of Fig. 13, where results arecompared with
measurements in selected control points.Note the good match between
the two sets of values (theaverage absolute difference is about
0.50 m in July 1995);in some cases, it is better than the match
between controlpoint measurements and the results of the
simulationunder steady-state conditions. Pumping causes a
substan-tial drawdown of the groundwater level in MarchJune;the end
of irrigation and autumn rains subsequentlydetermine a rise in the
water level. Comparison betweencomputed and measured piezometric
surfaces in the 19951996 period leads to the same considerations
made forcalibration under steady-state conditions.
Water budgetBy drawing up the aquifer water budget it is
possible tocalculate the volumes of water exchanged at the
modelboundaries (Fig. 14 and Table 1). In the period of 19February
199528 February 1996, the aquifer systemshows a negative balance of
over 6105 m3 (about 1,700m3/day). Note that this period was
characterized bybelow-average meteoric recharge ( 25%);
consideringan average inltration, there would be equilibriumbetween
inow and outow (also considering thatapproximations are unavoidable
in such calculations).The most interesting issue is the
considerable quantity ofwater crossing the coastline (over 142106
m3 inow and143106 m3 outow) through the typical ownet to
thefreshwatersaltwater contact (inow from the bottom,ascent along
the interface and outow from the top); thedifference (about 5.5105
m3) represents the net outowof groundwater towards the sea, which
occurs mostlysouth of the Albegna River. The sea water that
actuallyenters the aquifer system, causing the decay of thechemical
quality of groundwater (seawater intrusion), ismostly the balance
of ows crossing the coast in the Osa-
Fig. 13 Comparison between computed (red line) and measured
(blue line) hydraulic heads in different wells used for different
purposes(see also Fig. 12). The starting date of the simulation is
19 February 1995; comparison is made with monthly values
1570
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
Albegna sector (almost 3105 m3). The situation is madeworse by
sea water that ascends the Albegna River; duringcertain periods of
the year (when the piezometric level islower than the hydrometric
level) this water recharges theaquifer (about 6104 m3) for a total
of about 3.5105 m3
(over 900 m3/day).Figure 15 shows water ow variations in time
for the
various components of the water budget; note that
meteoricinltration occurs mostly in January and February, and
thatthe withdrawals are strongly inuenced by precipitation inthe
irrigation period. Figure 15b shows the coastal owdivided according
to zone and aquifer layer. An outow ofgroundwater (for a total of
about 8.5105 m3) alwaysoccurs in the southern sector of the Albegna
plain,whereas seawater intrusion affects the northern
sectorthroughout the investigated period, except in December
and January, when increased precipitation brings about anoutow
towards the sea. In accordance with eld data,seawater intrusion
mostly affects the lower aquifer layer(GL), whereas the upper one
(SE) is always characterizedby an outow towards the sea; this
difference is in partdue to the fact that during simulations, the
pumping wells,especially those withdrawing substantial amounts
ofwater, almost all extract water from the underlying layers.The
older wells located in the upper aquifer layer areexploited very
little. Figure 15c shows the variations inspace and time of
relationships between the aquifer andthe river within the modelled
area; the ow towards theaquifer occurs not only upstream (with a
balance of about2105 m3) but also downstream, where in June and
Julythe aquifer receives the above-mentioned 6104 m3 ofsaltwater
that ascends the river along the bottom.
Fig. 14 Computation grid for the three-dimensional numerical
model and water budget of the Plio-Pleistocene aquifer of the
AlbegnaRiver coastal plain in the 19 February 199528 February 1996
period obtained for transient conditions: balance of inows
(positive value)and outows (negative value) in the modelled area
expressed in m3/day
Table 1 Water budget of the Plio-Pleistocene aquifer of the
Albegna River coastal plain (19 February 199528 February 1996)
undertransient conditions
Item Inow (m3) Outow (m3) Balance (m3)
Meteoric inltration 1,646,577 1,646,577Lateral ows 747,347
747,347Flows from the cavernous limestone 785,432
785,432Withdrawals 2,932,010 2,932,010Albegna upstream 595,924
394,246 201,678Albegna downstream 59,784 578,413 518,629Total
Albegna 655,708 972,659 316,951Sea Osa-Albegna sector 73,425,320
73,138,660 286,660Sea opposite lagoon 69,015,580 69,864,140
848,560Total sea 142,440,900 143,002,800 561,900Aquifer system
146,275,964 146,907,469 631,505
Albegna: upstream from La Marsiliana to the conuence with the
Magione-Radicata stream; downstream from the conuence with
theMagione-Radicata stream to the mouth. Mare:Osa-Albegna sector
from the mouth of the Osa R. to the mouth of the Albegna
R.;oppositelagoon from the mouth of the Albegna R. to the southern
end of the model
1571
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
ValidationA calibrated model is validated when it can
closelyreproduce a different set of eld data, i.e. predicted
valuesapproximate measured ones. The model can be built evenwhen
time series data are lacking, since it is enough tohave data on a
single hydrological cycle (as in this case);however, the model can
only be effectively used after along period of experimentation,
during which it will likely
need to be rened, because the true relationships betweenthe
developed sequence and system trends in the mediumand long term are
not known (Anderson and Woessner2002; Oude Essink 2004).
Since the calibration was performed with reference to
the19951996 period, one characterized by below averagemeteoric
inltration, the model was validated against elddata acquired in
2003, a year in whichmeteoric inltrationwas
Fig. 15 Water ows in transient conditions for the 19 February
199528 February 1996 period (positive values are inows,
negativevalues are outows): a ow of the main components of the
water budget (values relative to the sea and to the Albegna River
are thedifference between inows and outows) with respect to local
precipitation (expressed in mm); b balance of ows crossing the
coastline,divided according to zone and aquifer layer; c balance of
ows exchanged with the Albegna River divided according to zone. For
thedenition of zones and symbols, see text and Table 1
1572
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
close to average. This validation aimed to assess the
modelsability to correctly reproduce the response of the
aquifersystem to different areal recharge conditions. Starting
from1996, monthly precipitation data were input into the model(the
rest remained unchanged), so that it was possible tocalculate the
piezometric surface for February 2003. Validationresults were
tested by comparing calculated values againstmeasurements in 51
observation wells; the two sets of valuesare highly correlated and
plot near the perfect correspondenceline (the absolute difference
is about 1.00 m). The sameconsiderations hold for the comparison
between computed andmeasured piezometric surfaces for February
2003.
Conclusions and future researchA conceptual and numerical model
was developed forsimulating the hydrodynamics of the
Plio-Pleistoceneaquifer of the Albegna River coastal plain. The
modelcan be used to determine the most important componentsof the
water budget and to identify river/aquifer/searelationships. At
present, inows and outows appear toensure the equilibrium of the
aquifer system; however,seawater intrusion along the coastline is
responsible for
the substantial deterioration of the chemical quality
ofgroundwater. The model helped establish that waterwithdrawal near
the coast during the irrigation season isthe main cause of seawater
intrusion, especially in theOsa-Albegna sector; sea water also
inows along the riverwhich at times inltrates the aquifer. The
deterioration ofgroundwater quality therefore appears to be due to
theway in which withdrawals are carried out (where andwhen) rather
than to the substantial quantity of waterbeing pumped (about 3106
m3/year). Groundwaterdegradation is currently a limiting factor for
localeconomic growth, above all for agriculture, and isproducing
conicts among the different users; there istherefore urgent need to
improve the state of the aquiferthrough a plan for its rational
exploitation based on acorrect conceptual representation of
physical conditions.
In this framework, the presented model represents thebasis for
future hydrochemical simulations that will help tobetter manage
local freshwater resources. This model is stillnot completely
reliable because of the incomplete knowl-edge of the system and the
few opportunities to fully checkresults; however, it can already
provide important informa-tion on the general evolution of the
system under differentstress conditions. In particular, the model
can be used toquantitatively assess the impact of variations in the
amountof water withdrawn and/or in the position of wells thatcould
help remediate saltwater intrusion.
The effects of hypothetical aquifer exploitations wereassessed
in terms of water budget and the evolutionhydraulic head starting
from February 2003; these fore-casts concern the next 10 years and
take into considerationthe local climate trends, i.e. a decrease in
effectiveprecipitation of about 2 mm/year (Barazzuoli et al.2002).
In particular, two exploitation scenarios wereexamined: the E
scenario considers the relocation of allcoastal wells to at least 3
km inland from the shore and themaintenance of the current overall
withdrawal; in contrast,the B scenario considers halving withdrawal
from all wellsin the plain but no relocation. The results of
thesesimulations are compared with those obtained
maintainingunchanged the current aquifer exploitation, i.e.
scenario A(Fig. 16 and Table 2).
Analysis of water budget variations reveals that halvingthe
discharge (B scenario) determines both an increase in
Fig. 16 Water budget in scenarios A,B and E for the
February2012February 2013 period (positive values are inows into
theaquifer system, negative values are outows). For the denition
ofcomponents and scenarios, see text and Table 2
Table 2 Difference between the water budget in scenarios B and A
and in scenarios E and A for the Plio-Pleistocene aquifer of
theAlbegna River coastal plain (February 2012February 2013)
Item Scenario Bscenario A Scenario Escenario A
Inow (m3) Outow (m3) Balance (m3) Inow (m3) Outow (m3) Balance
(m3)
Withdrawals 1,406,000 0Albegna upstream 203,700 284,700 488,400
81,400 167,300 248,700Albegna downstream 58,166 358,000 416,166
35,390 22,900 12,490Sea Osa-Albegna sector 90,000 90,000 180,000
80,000 80,000 160,000Sea opposite lagoon 40,000 40,000 80,000
50,000 60,000 110,000Balance aquifer system 222,860 33,200
All scenarios consider a 1.72 mm/year decrease in effective
precipitation (according to Barazzuoli et al. 2002). The different
scenarios areas follows: A exploitation of the aquifer continues
under the present conditions; B halving the discharge from all
wells in the plain, and norelocation of wells;E relocation of all
coastal wells up to 3 km inland from the shore and maintenance of
the current overall discharge. Forthe denition of components, see
Table 1
1573
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
the quantity of water drained by the river and a decrease inthe
quantity of water from the river, for an overall decreaseof about
9105 m3 in the ow of water from the AlbegnaRiver to the aquifer;
note that the quantity of water whichinltrates in the saline
portion of the river decreases bynearly 6104 m3. The opposite
effect is obtained byrelocating the coastal wells further inland
(scenario E):the Albegna River recharges the aquifer more (for a
total ofover 2.5105 m3) and increases by about 3.5104 m3
thequantity of seawater penetrating the aquifer as it ascendsthe
river. As for the coastal limit, in both scenarios there isan
increase in aquifer discharge to the sea, especially in
theOsa-Albegna sector, where extraction is greatest; therelocation
of coastal wells therefore determines a decreasein saltwater
intrusion, and can effectively contrast thedegradation of
groundwater quality.
Figure 17 shows the piezometric surface at the end ofthe
simulation (February 2013, when it has an averageheight of 5.34 m
asl), and its variations with respect tocalculated values in
scenarios B and E; note that the B
scenario produces a rather uniform increase in thepiezometric
level, with an average value of 0.56 m (thebalance of the aquifer
system increases by over 2105 m3,see Table 2), whereas the E
scenario leads to a decrease inthe piezometric level in the
northern sector and an increasein the southern sector near the
lagoon, for an averagereduction of 0.40 m (the amount of water in
the aquiferdecreases by about 3.5104 m3).
In conclusion, a considerable decrease in extractionfrom the
aquifer would lead to the reactivation of netgroundwater discharge
to the sea and especially to theAlbegna River. When wells are only
relocated, with nosignicant decrease in water extraction, there is
a decreasein seawater intrusion but a lowering of the
groundwaterlevel (a greater share of freshwater is extracted).
Future research will improve the geological andhydrogeological
reconstruction of the aquifer systemthrough the implementation of
new lithostratigraphic datafrom wells and boreholes, and through
further piezomet-ric, hydrometric (height of the river bottom and
of the free
Fig. 17 a Simulated groundwater level contour map for February
2013 (A scenario). b Piezometric difference between scenarios B and
Afor February 2013. c Piezometric difference between scenarios E
and A for February 2013. For the denition of scenarios, see text
andTable 2
1574
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
water surface), topographic (GPS measurement of groundelevation
in areas with a low hydraulic gradient) andphysicalchemical
(electrical conductivity, temperature,major ion and isotope
contents) surveys. Pumping andtracer tests will improve knowledge
of the hydraulic anddispersive properties of the aquifer. In
particular, perfo-rations planned in coastal areas will provide
data on thevertical distribution of aquifer salinity (and therefore
ofuid densities), thereby reducing the aforementioneduncertainties
in the local distribution of hydraulic heads,which currently
represent the greatest weakness of themodel. Lastly, public bodies
will monitor temperaturesand precipitation rates and will complete
a census of themost important wells and keep a record of water
dischargerates. These data will be used to signicantly improve
theconceptual and hydrodynamic models of the aquifer,thereby
allowing the construction of a robust numericaltransport model
essential for monitoring saltwater intru-sion and managing local
water resources.
Acknowledgements Research was supported by a grant from
theUniversity of Siena to M. Salleolini. The manuscript was
reviewedby P. Renard, G.H.P. Oude Essink, M. Antonelli and an
anonymousreviewer, who are gratefully acknowledged for their
constructivecriticism and suggestions that signicantly improved the
quality ofthe manuscript.
References
Agati L, Barbieri M, Bazzoli P, DallAglio M, Mantelli F,
Palmieri R,Venanzi G (2001) Drinking water quality in the
GrossetoProvince, Tuscany, Italy: an appraisal on the basis of the
studyof waterrock interaction. Proceedings of the
WaterRockInteraction (WRI-10) International Congress, Cagliari,
Italy,June 2001, pp 10591062
Ahmed S, De Marsily G, Talbot A (1988) Combined use ofhydraulic
and electrical properties of an aquifer in a geo-statistical
estimation of transmissivity. Ground Water 26(1):7886
Anderson MP, Woessner WW (2002) Applied groundwater model-ing:
simulation of ow and advective transport. Academic, SanDiego,
CA
Angelini A, Barazzuoli P, Cucini S, Mocenni B, Renai I, Rigati
R,Salleolini M (2000) Idrogeologia della bassa valle dei umi Osae
Albegna, Toscana meridionale [Hydrogeology of the lowerOsa-Albegna
River valley, southern Tuscany]. Quad Geol Appl7(3):521
Archie GE (1942) The electrical resistivity log as an aid
indetermining some reservoir characteristics. Trans Am Inst
MinMetal Eng 146:5462
Badon-Ghijben W (1889) Nota in Verband met de VoorgenomenPut
boring Nabij Amsterdam [Notes on the probable results ofwell
drilling near Amsterdam]. Tijdschrift van het KoninklijkInstituut
van Ingenieurs, The Haag, pp 822
Barazzuoli P, Guasparri G, Salleolini M (1993) Il clima
[Theclimate]. In: Giusti F (ed) La storia naturale della
Toscanameridionale [The natural history of southern Tuscany].
Pizzi,Milan, Italy, pp 140171
Barazzuoli P, Bouzelboudjen M, Cucini S, Kiraly L, Menicori
P,Salleolini M (1999) Holocene alluvial aquifer of the RiverCornia
coastal plain (southern Tuscany, Italy): database designfor
groundwater management. Environ Geol 39(2):123143
Barazzuoli P, Mocenni B, Rigati R, Salleolini M (2002)
Linuenzadella variabilit climatica sulle risorse idriche
rinnovabili dellaToscana meridionale [The inuence of climatic
variability on
the renewable water resources of southern Tuscany, Italy].
Attidel I Congresso Nazionale A.I.G.A. [Proceedings of the 1st
A.I.G.A. National Congress], Chieti, February 2003, Rendina,Rome,
pp 5568
Barazzuoli P, Mocenni B, Nocchi M, Rigati R, Salleolini M
(2003)Groundwater modelling of the Albegna River coastal
plain(southern Tuscany, Italy): preliminary results. Proceedings
ofthe 4th European Congress on Regional Geoscientic Cartogra-phy
and Information Systems, Bologna, June 2003, pp 591592
Barazzuoli P, Mocenni B, Nocchi M, Rigati R, Salleolini M
(2004)Numerical model of the Albegna aquifer system
(southernTuscany, Italy). Abstracts of the 32nd International
GeologicalCongress, Florence, August 2004
Barrocu G (2003) Seawater intrusion in coastal aquifers of
Italy. In:Lpez-Geta JA, Gmez J de D, de la Orden JA, Ramos
G,Rodrguez L (eds) Coastal aquifers intrusion
technology:Mediterranean countries, vol II, Alicante, March 2003,
InstitutoGeolgico y Minero de Espaa, Madrid, pp 207223
Bear J (2004) Management of a coastal aquifer. Ground Water
42(3):317
Bear J, Verruijt A (1987) Modelling groundwater ow andpollution.
Reidel, Dordrecht, The Netherlands
Bear J, Cheng AHD, Sorek S, Ouazar D, Herrera I (eds) (1999)
In:Seawater intrusion in coastal aquifers: concepts, methods
andpractices. Kluwer, Dordrecht, The Netherlands
Bencini A, Pranzini G (1992) The salinization of groundwater in
theGrosseto Plain (Tuscany, Italy). Proceedings of the 12th
SaltWater Intrusion Meeting, Barcellona, November 1992, pp
161175
Bencini A, Pranzini G (1996) Groundwater salinization in
SouthernTuscany (Italy). Proceedings of the 14th Salt Water
IntrusionMeeting, Malm, Sweden, June 1996, pp 261270
Bencini A, Gimenez Forcada E, Pranzini G (2001) Salinization
inthe coastal plain of Grosseto: hydrochemical study. Proceedingsof
the Water-Rock Interaction (WRI-10) International
Congress,Cagliari, Italy, June 2001, pp 517520
Bilardo U, Crestaz E, Dello Vicario C, Miccoli C, Pellegrini
M(1997) La valutazione dei consumi idrici nella Provincia diRavenna
nalizzata allimplementazione di un sistema per lagestione delle
acque sotterranee [Evaluation of water use in theprovince of
Ravenna with the aim of implementing a ground-water management
system]. Acque Sotterranee 53:2430
Bonazzi U, Fazzini P, Gasperi G (1992) Note alla carta
geologicadel bacino del Fiume Albegna [Notes on the geological map
ofthe Albegna River watershed]. Boll Soc Geol It 111:341354
Bossio A, Foresi LM, Mazzei R, Salvatorini G, Sandrelli F,
Bigotti M,Colli A, Rossetto R (2004) Geology and stratigraphy of
thesouthern sector of the Neogene Albegna River basin
(Grosseto,Tuscany, Italy). Geol Romana 37:165173
Cheng AHD, Ouazar D (1999) Analytical solutions. In: Bear
J,Cheng AHD, Sorek S, Ouazar D, Herrera I (eds) Seawaterintrusion
in coastal aquifers: concepts, methods and practices.Kluwer,
Dordrecht, The Netherlands, pp 163191
Cheng AHD, Ouazar D (eds) (2004) Coastal aquifer
management:monitoring, modeling, and case studies. Lewis, Boca
Raton, FL
Cheng J, Chen C (2001) Three-dimensional modeling of
density-dependent salt water intrusion in multi-layered coastal
aquifersin Jahe River Bassin, Shandong province, China. Ground
Water39(1):131139
Civita M, De Maio M, Vigna B (1999) Una metodologia GIS per
lavalutazione della ricarica attiva degli acquiferi [A GIS
method-ology for evaluating the active recharge of aquifers]. Atti
del 3Convegno Nazionale sulla Protezione e Gestione delle
AcqueSotterranee [Proceedings of the 3rd National Congress
onGroundwater Protection and Management], Parma, October1999,
Pitagora, Bologna, Italy, pp 12911302
Custodio E (2002) Coastal aquifers as important natural
hydro-geological structures. Proceedings of the XXXII IAH and
VIALHSUD Congress Groundwater and Human Development,Mar del Plata,
Italy, October 2002, pp 19051918
Custodio E, Bruggeman GA (1987) Groundwater problems in
coastalareas. Studies and Reports in Hydrology, 45, UNESCO,
Paris
1575
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
-
Custodio E, Galofr A (eds) (1993) Study and modelling
ofsaltwater intrusion into aquifers. Proceedings of the 12th
SaltWater Intrusion Meeting, Barcelona, November 1992
Diersch HJG (1998a) Reference manual FEFLOW. Institute forWater
Resources Planning and Systems Research, WASY,Berlin
Diersch HJG (1998b) FEFLOW: users manual. Institute for
WaterResources Planning and Systems, WASY, Berlin
Diersch HJG (1998c) About the formulation of hydraulic
headboundary (potential) conditions for uid
density-dependentgroundwater problems. FEFLOW white papers,
Institute forWater Resources Planning and Systems, WASY, Berlin
Diersch HJG, Kolditz O (2002) Variable-density ow and
transportin porous media: approaches and challenges. Adv Water
Resour25:899944
Doherty J (1998) PEST: Model-independent parameter
estimation.Watermark Numerical Computing, Brisbane, Australia
Emch PG, Yeh WWG (1998) Management model for conjunctiveuse of
coastal surface water and groundwater. J Water ResourPlan Manage
124(3):129139
Essaid HI (1990) A multilayered sharp interface model of
coupledfreshwater and saltwater ow in coastal systems: model
develop-ment and application. Water Resour Res 26(7):14311454
Fanelli M, Bellucci L, Nachira F (1982) Manifestazioni
idrotermaliitaliane [Thermomineral springs in Italy]. CNR-PFE,
SPEG, RF13, Rome
FAO (1997) Seawater intrusion in coastal aquifers: guidelines
forstudy, monitoring and control. Water reports, 11, FAO, Rome
Favi E, Rossi R (1991) Capacit duso agricolo, erodibilit
eraggruppamenti idrologici dei suoli [Agricultural
capability,erodibility and hydrological groups of soils]. In: La
valledellAlbegna: pedologia e valutazione del territorio
[TheAlbegna valley: pedology and land evaluation], Giunta
Region-ale Toscana, Marsilio, Venice, pp 137147
Gabbani G, Gargini A (1992) An application of the Archies law
tothe hydrogeological investigation of an alluvial coastal
plain.Proceedings of the 12th Salt Water Intrusion Meeting,
Barce-lona, November 1992, pp 617630
Gambolati G, Putti M, Paniconi C (1999) Three-dimensional
modelof coupled density-dependent ow and miscible salt
transport.In: Bear J, Cheng AHD, Sorek S, Ouazar D, Herrera I
(eds)Seawater intrusion in coastal aquifers: concepts, methods
andpractices. Kluwer, Dordrecht, The Netherlands, pp 315362
Gorman T, Kelly WE (1990) Electrical-hydraulic properties
ofunsaturated Ottawa sands. J Hydrol 118:118
Grassi S, Netti R (2000) Sea water intrusion and mercury
pollutionof some coastal aquifers in the province of Grosseto
(southernTuscany, Italy). J Hydrol 237:198211
Herzberg A (1901) Die Wasserversorgung einiger Nordseebaeder[The
water supply of selected North Sea towns]. ZF Gasbe-leucht
Wasserversorg 44:815819 842844
Holzbecher E (1998) Modeling density-driven ow in porousmedia:
principles, numerics, software. Springer, Berlin
IPCC (2007) Summary to policymakers: contribution of
WorkingGroup I to the Fourth Assessment Report of the
Intergovern-mental Panel on Climate Change. In: Climate change
2007: thephysical science basis, WMO, Geneva
Koch M, Zhang G (1992) Numerical simulation of the migration
ofdensity dependent contaminant plumes. Ground Water 5:731742
Lpez-Geta JA, Gmez J de D, de la Orden JA, Ramos G,Rodrguez L
(eds) (2003) Coastal aquifers intrusion technology:Mediterranean
countries, vol 2. Instituto Geolgico y Minero deEspaa, Madrid
Maimone M, Harley B, Fitzgerald R, Moe H, Hossain R, Heywood
B(2004) Coastal aquifer planning elements. In: Cheng AHD,Ouazar D
(eds) Coastal aquifer management: monitoring,modeling, and case
studies. Lewis, Boca Raton, FL, pp 12
Mancini F (1960) Sulla geologia della Piana dAlbegna,
Grosseto[The geology of the Albegna plain, Grosseto]. Boll Soc Geol
It79:190
Morell I (2003) Acuferos detrticos costeros. In: Lpez-Geta
JA,Gmez J de D, de la Orden JA, Ramos G, Rodrguez L (eds)Coastal
aquifers intrusion technology: Mediterranean countries,vol 2,
Alicante, March 2003, Instituto Geolgico y Minero deEspaa, Madrid,
pp 3143
Nocchi M (2002) Modlisation numrique de la plaine ctire de
larivire Albegna, Toscana meridionale, Italia [Numerical model-ing
of the Albegna River coastal plain, southern Tuscany, Italy].Diplme
de spcialisation en hydrogologie, Centre dHydro-gologie de
Neuchtel, France
Nocchi M (2004) Modellizzazione numerica del sistema
acquiferodella pianura del ume Albegna, Toscana meridionale
[Numer-ical modeling of the aquifer system of the Albegna River
plain,southern Tuscany]. Tesi di Dottorato di Ricerca in Difesa
eConservazione del Suolo, Vulnerabilit Ambientale e Prote-zione
Idrogeologica, Consorzio Universitario tra le Universitdegli Studi
di Cagliari e Siena, XVI ciclo [PhD Thesis on SoilConservation,
Environmental Vulnerability and Hydrogeolog-ical Protection,
University Consortium between Cagliari andSiena Universities,
Italy
Nocchi M, Salleolini M (2007) Ansedonia: simulazione degli
effettidellemungimento delle acque sotterranee (seconda
parte)[Ansedonia: simulating the effects of groundwater
withdrawal(second part)]. Period Trimestrale Ordine Geol Toscana
IlGeologo 67:27
Oude Essink GHP (2003) Mathematical models and their
applica-tion to salt water intrusion problems. In: Lpez-Geta JA,
GmezJ de D, de la Orden JA, Ramos G, Rodrguez L (eds)
Coastalaquifers intrusion technology: Mediterranean countries, vol
2,Alicante, March 2003, Instituto Geolgico y Minero de
Espaa,Madrid, pp 5777
Oude Essink GHP (2004) Modelling for geochemists: everythingyou
always wanted to know about modelling, but were afraid toask!
Proceedings of the 17th Salt Water Intrusion Meeting,Delft, The
Netherlands, May 2002, pp 1325
Piper AM (1944) A graphic procedure in the
geochemicalinterpretation of water analysis. Trans Am Geophys
Union25:914923
Ponzini G, Ostroman A, Molinari M (1983) Empirical
relationbetween electrical transverse resistance and hydraulic
transmis-sivity. Geoexploration 22:115
Post V, Kooi H, Simmons C (2007) Using hydraulic
headmeasurements in variable-density ground water ow
analyses.Ground Water 45(6):664671
Protano G, Riccobono F, Sabatini G (2000) Does salt
waterintrusion constitute a mercury contamination risk for
coastalfresh water aquifers? Environ Pollut 110(3):451458
Thornthwaite CW (1948) An approach toward a rational
classica-tion of climate. Geogr Rev 38(1):5594
Thornthwaite CW, Mather JR (1957) Instructions and tables
forcomputing potential evapotranspiration and the water balance,vol
10. Thornthwaite, Centerton, NJ, USA, pp 185311
Tozzi M, Zanchi A (1987) Evoluzione paleogeograca e
strutturalerecente del bacino del ume Albegna [Recent
paleogeograph-ical and structural evolution of the Albegna River
basin]. GeolRomana 26:305325
Urish DW (1981) Electrical resistivity-hydraulic
conductivityrelationship in glacial outwash aquifer. Water Resours
Res 17(5):14011408
Van Dam JC (1999) Exploitation, restoration and management.
In:Bear J, Cheng AHD, Sorek S, Ouazar D, Herrera I (eds)Seawater
intrusion in coastal aquifers: concepts, methods, andpractices.
Kluwer, Dordrecht, The Netherlands, pp 73125
Voss CI, Souza WR (1987) Variable density ow and solutetransport
simulation of regional aquifers containing a
narrowfreshwatersaltwater transition zone. Water Resour Res
23(10):18511866
Wang HF, Anderson MP (1982) Introduction to
groundwatermodelling: nite difference and nite element methods.
Free-man, New York
1576
Hydrogeology Journal (2008) 16: 15571576 DOI
10.1007/s10040-008-0324-z
A conceptual and numerical model for groundwater management: a
case study on a coastal aquifer in southern Tuscany,
ItalyAbstractIntroductionThe study areaConceptualizationGeometry
and structurePiezometric surface and hydraulic
propertiesGroundwater qualityRole of surface water
Numerical modellingSoftwareDiscretization and boundary
conditionsCalibrationWater budgetValidationConclusions and future
research
References
/ColorImageDict > /JPEG2000ColorACSImageDict >
/JPEG2000ColorImageDict > /AntiAliasGrayImages false
/DownsampleGrayImages true /GrayImageDownsampleType /Bicubic
/GrayImageResolution 150 /GrayImageDepth -1
/GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true
/GrayImageFilter /DCTEncode /AutoFilterGrayImages true
/GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict >
/GrayImageDict > /JPEG2000GrayACSImageDict >
/JPEG2000GrayImageDict > /AntiAliasMonoImages false
/DownsampleMonoImages true /MonoImageDownsampleType /Bicubic
/MonoImageResolution 600 /MonoImageDepth -1
/MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true
/MonoImageFilter /CCITTFaxEncode /MonoImageDict >
/AllowPSXObjects false /PDFX1aCheck false /PDFX3Check false
/PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true
/PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ]
/PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [
0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None)
/PDFXOutputCondition () /PDFXRegistryName (http://www.color.org?)
/PDFXTrapped /False
/SyntheticBoldness 1.000000 /Description >>>
setdistillerparams> setpagedevice