JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 110, B08204, doi:10.

1029/2004JB003542, 2005

Carbon dioxide diffuse degassing and estimation of heat

release from volcanic and hydrothermal systems
G. Chiodini, D. Granieri, R. Avino, S. Caliro, and A. Costa
Osservatorio Vesuviano, Istituto Nazionale di Geofisica e Vulcanologia, Naples, Italy

C. Werner
Institute of Geological and Nuclear Sciences, Taupo, New Zealand
Received 16 November 2004; revised 19 April 2005; accepted 3 May 2005; published 16 August 2005.

[1] We present a reliable methodology to estimate the energy associated with the subaerial
diffuse degassing of volcanic-hydrothermal fluids. The fumaroles of 15 diffuse degassing
structures (DDSs) located in eight volcanic systems in the world were sampled and
analyzed. Furthermore, each area was measured for soil temperature gradients and for soil
CO2 fluxes. The results show that each hydrothermal or volcanic system is characterized
by a typical source fluid which feeds both the fumaroles and diffuse degassing through
the soil. Experimental data and the results of physical numerical modeling of the process
demonstrate that the heat released by condensation of steam at depth is almost totally
transferred by conduction in the uppermost part of the soil. A linear relationship is
observed between the log of the steam/gas ratio measured in the fumaroles and the log of
the ratio between soil thermal gradient and soil-gas flux. The main parameter controlling
this relation is the thermal conductivity of the soil (Kc). For each area, we computed
the values of Kc which range from 0.4 to 2.3 W m
1
C
1. Using the CO2 soil fluxes as a
tracer of the deep fluids, we estimated that the total heat released by steam condensation in
the systems considered varies from 1 to 100 MW.
Citation: Chiodini, G., D. Granieri, R. Avino, S. Caliro, A. Costa, and C. Werner (2005), Carbon dioxide diffuse degassing and
estimation of heat release from volcanic and hydrothermal systems, J. Geophys. Res., 110, B08204, doi:10.1029/2004JB003542.

1. Introduction caldera (90 km2). These data suggest that the monitoring
of energy fluxes associated with diffuse degassing may be
[2] The energy released through hydrothermal and vol-
an important tool in the surveillance of volcanic activity,
canic gaseous emissions is an important component of the
and continued development of quick and reliable methods
energy balance of quiescent volcanoes. Typically energy
to measure the thermal energy released by the hydrother-
balances of volcanoes are derived from mass balance
mal activity will encourage the systematic use of this
calculations at crater lakes. In previous studies at 14 crater
technique.
lakes, energy outputs ranged from 0.54  108 to 3.85  108 W
[3] In this work we focus our attention on the energy
[Brown et al., 1989; Sheperd and Sigurdsson, 1978; Hurst et
released by subaerial diffuse degassing, i.e., where the gas is
al., 1991; Pasternack and Varekamp, 1997]. In four volcanic
not directly emitted by fumaroles. A recent study has shown
systems without crater lakes (Vulcano, Solfatara di Pozzuoli,
that diffuse degassing does not occur across entire volca-
Nisyros, Ischia), the thermal energy released was estimated to
noes, but rather in restricted areas named diffuse degassing
range from 0.23  108 to 1.39  108 W [Caliro et al., 2004;
structures (DDSs [Chiodini et al., 2001a]) commonly asso-
Chiodini et al., 2004, 2001a, 1996] based on measurements of
ciated with regions of high permeability (faults, fractures).
volcanic-hydrothermal CO2 released through soil diffusion
[4] Thermal Infrared images of DDSs present an intrigu-
emission. For instance, the energy dissipated daily by the
ing visualization of the heat flux associated with these
diffuse degassing structure of Solfatara di Pozzuoli (0.5 km2)
structures (Figure 1). It is worth noting both the large extent
is the main source of energy in the entire Campi Flegrei
of the hot soil areas and the generally low magnitude of the
caldera in the current period of quiescence. According to
thermal anomaly, i.e., during the survey periods, maximum
Chiodini et al. [2001a], the thermal energy released daily is
soil temperatures rarely exceeded 20– 30
C. At these tem-
one order of magnitude greater than the elastic energy
peratures steam condenses below the surface and most of
released during 1983 – 1984 seismic crisis, greater than the
the heat should be transmitted by conduction.
energy associated with ground deformation, and is about
[5] In a purely conductive regime, the heat flux is a linear
10 times grater than the conductive heat flux over the
function of the soil thermal gradient with a slope propor-
tional to the thermal conductivity of the medium. On the
Copyright 2005 by the American Geophysical Union. basis of this consideration, previous studies have estimated
0148-0227/05/2004JB003542$09.00 the heat released from hot soils measuring soil thermal

B08204 1 of 17
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING
gradients and assuming thermal conductivities of the soils
[Chiodini et al., 2001a; Brombach et al., 2001; Lardy and
Tabbagh, 1999; Severne and Hochstein, 1994]. Chiodini et
al. [2001a] and Brombach et al. [2001] assumed the thermal
conductivity to be 1 W m
1
C
1 following Clauser
and Huenges [1995], while Severne and Hochstein [1994,
p. 212] used ‘‘appropriate thermal conductivity values’’ to
assess the surface heat transfer. Lardy and Tabbagh [1999]
directly measured the ground thermal conductivity using
active methods in superficial glassy ash, and obtained a
mean of 0.51 ± 0.16 W m
1
C
1 with a large distribution
of the data (0.30– 1.06 W m
1
C
1) depending on the
variability of the soil composition. However, the long time
required for the direct measurements of the thermal con-
ductivity limits the use of this technique.
[6] An alternative way to estimate the heat flux from
DDSs is based on CO2 flux measurements. The degassing
of CO2 from a DDS can be computed and mapped with
adequate precision applying the sequential Gaussian simu-
lation (sGs) method to the measurement points [Cardellini
et al., 2003]. Because CO2 is the second most abundant gas
species (after H2O) in volcanic and hydrothermal systems
considered here and because it is a noncondensable gas at
T-P conditions of the DDSs, it serves as a tracer of the
diffuse degassing process. In particular, CO2 fluxes were
used to estimate the heat released by DDSs [Brombach et
al., 2001; Chiodini et al., 2001a, 2004; Frondini et al.,
2004]. These studies calculated the total amount of steam
released at depth based on the assumption that the fluids
which supply the diffuse degassing process have, at depth,
the same composition as those emitted by the fumarolic
vents of highest temperature and flow rate. The thermal
energy was then computed by multiplying the steam flux by
the enthalpy of the steam minus the enthalpy of the liquid at
ambient temperature.
[7] In this paper, we discuss the interrelation of CO2
fluxes, thermal gradients, and fumarolic compositions at
15 DDSs from eight different volcanic systems. The aim
is to test the reliability of the above method for heat flux
estimation by using the measured soil thermal gradients,
CO2 fluxes and fumarole compositions. We will describe
also a new method for the estimation of the thermal
conductivity of the DDSs’ soil based on the combination
of the different measured parameters.
[8] In addition, we present the results of a long period of
continuous monitoring of the soil temperature at different
depths in the crater of Solfatara di Pozzuoli. This data set is
used to determine the influence of diurnal and seasonal
temperature variations on the thermal heat released from the
soil.
2. Study Sites
[9] The DDSs investigated in this study are located in
8 different volcanic systems in the world: Stefanos, Kami-
nakia, Polybotes Micros, and SU area in the Nisyros
volcanic system (Greece), Solfatara in the Campi Flegrei
district (Italy), Donna Rachele in Ischia Island (Italy), NE
rim and Bottom Crater in the Vesuvio system (Italy),
Comalito cinder cone in the Masaya Caldera Complex
(Nicaragua), Cerro Negro in the Cerro Negro volcanic
system (Nicaragua), Vulcano Porto di Levante Beach and
2 of 17
B08204
Vulcano Crater in Vulcano island (Italy), Favara Grande and
Favara Piccola in Pantelleria island (Italy). Typically, the
DDSs are not exclusively contained within crater areas but
rather are often related to structural features such as faults/
fracture zones.
[10] Nisyros volcanic island, built up during the last
100 ka, lies at the eastern end of the Aegean active volcanic
arc. Hydrothermal eruptions were frequent in historical
times. The oldest hydrothermal craters are located at Kami-
nakia, while the last eruption, which occurred in September
1887, originated in Polybotes Micros crater. At present,
Stefanos hydrothermal crater is the most active fumarolic
site. According to the geochemical model by Chiodini et al.
[1993a], the hydrothermal system of Nisyros is made up of a
deep reservoir (>1000 m) with a temperature of 300
C,
and of a shallower system, with temperatures of 150 – 250
C.
[11] Ischia is the westernmost volcanic complex of the
Campania area and the last eruption took place in 1301 A.D.
[Vezzoli, 1988]. Our study is focused on the so-called Donna
Rachele fumarolic area, which is located on the western
flank of Mount Epomeo. It is characterized by hydrother-
mally altered terrain, steaming ground and focused vents.
Chiodini et al. [2004] proposed two distinct hydrothermal
reservoirs feeding the fumarolic area: a shallower reservoir
characterized by a temperature of 250
C and a pressure of
40 bars and a deeper reservoir characterized by a temper-
ature of 300
C and a pressure of 90 bars.
[12] Solfatara is a tuff cone affected by intense diffuse
degassing. It is located inside Campi Flegrei caldera com-
plex. A conceptual geochemical model of the hydrothermal
system suggests one or multiple aquifers are located over
the magma chamber [Cioni et al., 1984; Chiodini and
Marini, 1998]. Gas equilibrates in a ‘‘superheated’’ vapor
zone at temperature of 210– 220
C in the shallower part of
the hydrothermal system [Chiodini et al., 2001a].
[13] Somma-Vesuvio is a central composite volcano lo-
cated in the southern sector of the Campanian Plain. It was
formed by an ancient stratovolcano, Mount Somma, and by
a more recent cone, Vesuvio [Santacroce, 1987]. At present,
Vesuvio volcano is characterized by widespread fumarolic
activity on the inner slopes and bottom of the crater, as well
as by diffuse soil CO2 degassing. Geochemical studies of
crater fumarolic fluids reveal the presence of a hydrothermal
system located inside the Somma caldera [Chiodini et al.,
2001b]. On the basis of C-H-O gas equilibria of crater
bottom fumarolic fluids, the temperature of this system was
estimated to be 400 – 500
C. Hydrostatic models suggest a
hydrothermal reservoir exists at depths of 2– 4 km within
the carbonate sequence, which is present at depths >2.5 km
underneath the volcano [Chiodini et al., 2001b].
[14] Masaya is an active basaltic volcano on the Central
American volcanic front, situated about 25 km southeast of
Managua, Nicaragua. It is an unusual subduction zone
volcano because of its shield-like form, consistent tholeiitic
basaltic composition (Masaya is one of the few volcanoes
known to have hosted basaltic Plinian activity [Williams,
1983]), frequent eruptive activity, and a 25-year cycle of
major noneruptive degassing crises [Stoiber et al., 1986].
The flux of SO2 and HCl from Masaya represents the largest
reported noneruptive volcanic release of these species in the
world [Stoiber et al., 1986]. Comalito cinder cone is located
approximately 3.5 km NE of the Masaya crater, along a
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING B08204

Figure 1

3 of 17
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING
northeast trending lineament. The site investigated was
located adjacent to the Comalito cinder cone, and was
characterized by steam and diffuse CO2 emissions, minimal
vegetation, and soil temperatures up to 80
C [Lewicki et al.,
2003].
[15] Cerro Negro has been interpreted to be an old cinder
cone [Mooser et al., 1956; Walker and Carr, 1986; Simkin
and Siebert, 1994] or a young composite volcano [Wood,
1978; McKnight and Williams, 1997] and is located in
northwestern Nicaragua on the flank of the El Hoyo – Las
Pilas volcanic complex. It erupted 22 times since its
formation in 1850, with the last eruption in May– August
1995. Thermal features are limited to superheated fumaroles
in the crater (350
C), and low-temperature (<100
C)
fumaroles located in arcuate fractures on the crater rim
and near phreatic pit craters between Cerro Negro and the
cinder cone Cerro La Mula [Connor et al., 1996]. At the
summit of Cerro Negro volcanic edifice, where we inves-
tigated a small area (300 m2), the soil was affected by
intense CO2 diffuse degassing (up to 35,000 g m
2 d
1) and
high temperatures [Salazar et al., 2001].
[16] Vulcano is one of the currently active volcanoes of
the Aeolian island arc, north of Sicily, Italy, and is charac-
terized mainly by explosive eruptions. The last eruption in
the La Fossa crater occurred in 1888 – 1890 [Mercalli and
Silvestri, 1891]. Since that time the crater has hosted high-
temperature fumaroles (up to 700
C in 1993) with chemical
composition typical of volcanic emissions [Chiodini et al.,
1995; Nuccio et al., 1999], i.e., high content of acidic gases
like HCl, HF, and SO2. Fumarolic activity is also present at
Porto di Levante Beach (PL Beach). The beach fumaroles
have temperatures of 100
C and chemical compositions
typical of the hydrothermal environment [Chiodini et al.,
1995; Tedesco et al., 1995], i.e., absence of acidic gases and
presence of reduced gases species, like CH4 and H2S.
[17] Pantelleria island is a quiescent volcano located in
the Sicily Channel, 110 km south of Sicily and 70 km north
of Tunisia. Most of the exposed volcanics on the island are
alkaline or peralkaline products with high silica contents
(pantellerites). The most recent volcanic activity in the area
took place during the 19th century with two submarine
eruptions (1831 and 1891) in the vicinity of the island
[Civetta et al., 1988]. At present, the volcanic activity at
Pantelleria is represented by steam emissions in the areas of
Favare (Favara Grande and Favara Piccola), Passo del
Vento, and Cuddia di Mida and gas manifestations in the
Lake of Venus and the Gadir gulf. A thermal aquifer is
present at depth, with a maximum estimated temperature of
260
C [D’Alessandro et al., 1994; Squarci et al., 1994].
3. Methods
3.1. Field and Laboratory Methods
[18] The following procedures were used in each DDS:
[19] 1. CO2 soil flux measurements were made within the
diffuse degassing zones using the accumulation chamber
method. This is a direct, passive method based on the
measurement of the initial CO2 concentration increase
Figure 1. Thermal infrared images of (a) Vulcano crater, (b) Vulcano PL Beach, (c) Vesuvio crater, (d) Solfatara
crater, (e) Lakki plain of Nisyros and (f) Stefanos crater. Images were taken in the IR wavelengths 8 –14 mm using a
Thermo Tracer Nec TH7102MV.
4 of 17
B08204
inside a chamber of known volume that is placed on the
soil. The increase in CO2 concentration is directly propor-
tional to the CO2 flux [Parkinson, 1981; Tonani and Miele,
1991; Chiodini et al., 1996]. The method has been tested by
several authors under controlled laboratory conditions.
Chiodini et al. [1998] measured values within 15% of the
imposed fluxes, whereas Evans et al. [2001] recognized a
systematic underestimation of the flux (average of

12.5%), over an imposed flux range of 200 –12,000 g
m
2 d
1. In a field reproducibility test of the method,
carried out at two points with high and low CO2 flux,
Carapezza and Granieri [2004] found an uncertainty of
12% for high fluxes (mean of 997 g m
2 d
1) and 24%
for low fluxes (mean of 5.7 g m
2 d
1).
[20] 2. Temperatures were measured along vertical pro-
files at 0.01 to 0.05 m spacing between 0.02 and 0.40 m
depth in the soil by means of a digital thermocouple type K
(the measurement accuracy is ±0.05% of reading). Soil
temperature profiles were measured in the sites corresponding
with the CO2 flux locations.
[21] 3. Fumaroles were sampled within the degassing
areas. We selected fumaroles at each site with the highest
flow rate and temperature to obtain samples that were
most representative of the fluids at depth. Complete
chemical analyses were generally conducted at the Osser-
vatorio Vesuviano laboratories following standard meth-
ods [Giggenbach, 1975, 1991; Giggenbach and Goguel,
1989; Chiodini et al., 1993b]. Fumaroles from Cerro
Negro and Masaya (Nicaragua) were analyzed at the
INGV-Palermo Laboratories.
[22] In addition, soil temperatures were continuously
measured during two years (October 2001 to October
2003) within the crater of Solfatara using an automatic
station. The station was equipped with a Lastem BABUC
ABC data logger that acquires, at 1-hour time intervals, the
soil temperature at 0.1, 0.2, and 0.3 m depth. CO2 soil flux
and various parameters that can potentially influence the
CO2 soil flux, i.e., barometric pressure, air and soil tem-
perature, air and soil humidity, wind speed, and rainfall
were also measured at the same location [Granieri et al.,
2003]. The station was used as a ‘‘base station’’ to inves-
tigate the thermal gradient and its temporal variability in the
shallower part of a typical volcaniclastic soil.
3.2. Modeling Procedures
[23] Physical modeling was applied to simulate heat and
fluid flow through the uppermost few meters of each
hydrothermal system, including the surface, where the
thermal gradients and CO2 fluxes were measured. The used
geothermal simulator (TOUGH2 [Pruess, 1991; Pruess et
al., 1999]) accounts for the coupled transport of heat and a
multiphase (gas and liquid) and a multicomponent (water
and carbon dioxide) fluid, and was recently used to model
large-scale diffuse degassing at Solfatara crater [Chiodini et
al., 2003; Todesco et al., 2003]. These studies were based
on a two-dimensional, axisymmetric domain that was 2500 m
wide and 1500 m deep. The model successfully reproduced
measured changes in both the gas composition and discharge
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING B08204

Table 1. Composition of Sampled Fumarolesa

DDS Date T,
C H2O CO2 Stot Ar O2 N2 CH4 H2 He HCl HF
Campi Flegrei Volcanic System
Solfatara Mar 1999 161 84.36 15.4 0.1394 4.89E-05b 1.86E-05 5.07E-02 3.18E-03 4.23E-02 1.29E-04 nd nd
Apr 1999 162 84.85 14.9 0.1454 5.67E-05 bdl 5.41E-02 3.51E-03 4.36E-02 1.23E-04 nd nd
May 1999 162 84.84 14.92 0.1451 5.06E-05 7.55E-06 5.24E-02 3.41E-03 4.21E-02 1.21E-04 nd nd
Jun 1999 162 84.83 14.93 0.1448 4.44E-05 1.51E-05 5.06E-02 3.31E-03 4.05E-02 1.18E-04 nd nd
Jul 2000 161 84.54 15.24 0.1366 6.09E-05 bdl 4.53E-02 1.94E-03 3.51E-02 1.35E-04 nd nd
Mar 2002 159 83.68 16.1 0.1389 4.21E-05 bdl 4.24E-02 2.00E-03 3.72E-02 1.25E-04 nd nd

Nisyros Volcanic System

Stefanos Sep 1999 99.5 98.3 1.36 0.361 2.96E-05 bdl 2.99E-03 4.77E-03 1.27E-02 5.67E-05 nd nd
May 2000 99.5 98.3 1.34 0.33 3.22E-05 bdl 2.81E-03 4.41E-03 1.05E-02 4.21E-05 nd nd
Sep 2000 99.5 98.43 1.18 0.3393 5.84E-05 bdl 3.77E-03 3.30E-03 9.31E-03 3.74E-05 nd nd
Feb 2001 99.6 98.5 1.14 0.3456 1.29E-05 1.94E-06 1.62E-03 3.47E-03 9.97E-03 3.83E-05 nd nd
May 2001 97.8 98.49 1.14 0.3524 2.29E-05 bdl 2.18E-03 3.76E-03 1.05E-02 3.86E-05 nd nd
Sep 2001 97.6 98.55 1.08 0.3489 1.16E-05 bdl 1.54E-03 3.65E-03 9.44E-03 3.56E-05 nd nd
Feb 2002 98.9 98.61 1.03 0.3458 9.81E-06 bdl 1.42E-03 3.32E-03 9.29E-03 3.40E-05 nd nd
Jul 2002 98.3 98.6 1.04 0.3355 1.05E-05 bdl 1.43E-03 3.57E-03 9.94E-03 3.46E-05 nd nd
May 2003 96.7 98.63 1.17 0.1849 1.76E-05 1.68E-06 1.78E-03 4.73E-03 8.25E-03 2.89E-05 nd nd
SU area May 2003 98.1 97.48 2.18 0.2668 4.61E-05 9.25E-07 4.29E-03 3.05E-02 3.03E-02 5.14E-05 nd nd
Polybotes Micros May 2001 97.5 99 0.75 0.232 2.38E-05 bdl 1.77E-03 1.02E-03 1.66E-02 2.70E-05 nd nd
Kaminakia Sep 2000 98 94.9 4.45 0.444 8.48E-04 bdl 6.63E-02 4.70E-02 5.02E-02 8.99E-05 nd nd

Masaya Volcanic System

Comalito Mar 2003 79 42 10.76 nd nd 9.94E+00 3.74E+01 3.28E-04 4.41E-04 3.68E-04 nd nd

Cerro Negro Volcanic System

Cerro Negro Mar 2003 197 95.66 2.41 0.327 nd 1.95E-01 1.33E+00 bdl 4.10E-02 8.28E-06 3.60E-02 1.86E-05

Pantelleria Volcanic System

Favara Grande Jul 2004 98.9 97.35 2.32 0.0006 2.16E-04 2.12E-03 1.27E-02 1.26E-01 1.83E-01 5.88E-05 nd nd
Favara Piccola Jul 2004 98 96.32 3.01 0.0003 4.86E-03 8.45E-02 3.76E-01 1.15E-01 9.09E-02 9.65E-05 nd nd

Ischia Volcanic System

Donna Rachele Oct 2000 99.2 99.63 0.36 0.0037 1.86E-05 1.01E-06 2.02E-03 2.41E-05 1.95E-03 9.57E-06 nd nd
May 2001 97.4 99.76 0.24 0.0043 7.00E-06 4.47E-07 8.62E-04 1.80E-05 1.02E-03 3.13E-06 nd nd

Vulcano Volcanic System

Crater Mar 2002 255 94.7 4.85 0.2188 7.21E-05 4.46E-07 3.85E-02 bdl 1.26E-02 2.48E-05 1.47E-01 2.83E-02
PL Beach Mar 2002 98.1 89.88 9.79 0.2182 4.04E-04 bdl 7.57E-02 1.31E-02 1.73E-02 2.44E-05 nd nd
PL Beach Mar 2002 99 94.01 5.72 0.1505 5.00E-05 bdl 3.83E-02 8.30E-03 7.56E-02 7.95E-06 nd nd

Vesuvio Volcanic System

NE Rim Feb 2002 68.5 33.45 1.56 nd 6.06E-01 1.36E+01 5.08E+01 1.09E-04 nd nd nd nd
Bottom Crater Mar 2002 94 88.86 10.98 0.0435 3.91E-05 4.25E-06 1.79E-02 5.17E-03 9.38E-02 7.07E-05 nd nd
a
Diffuse degassing structures (DDSs); analytical data are expressed in vol %; nd, not determined; bdl, below detection limit; PL, Porto di Levante.
b
Read 4.89E-05 as 4.89  10
5.

rate with time, and modeling results were correlated to recent
bradyseismic crises at Campi Flegrei. In the present paper we
used the same physical approach to model the heat and the
fluid transfer in the uppermost part of the Solfatara hydro-
thermal system and test the applicability of using shallow soil
thermal gradients to estimate total energy transfer by soil
diffuse degassing. In order to investigate the effect of the input
of fumarolic fluids into the soil, a two-dimensional (xz),
axisymmetric domain, 10 m wide (x) and 2 m deep (z) was
considered. The domain was discretized into 1000 elements
(25  40 cells) of variable size, with finer elements located
near the axis of symmetry and at the top and bottom
boundaries. The medium is assumed homogeneous, with
constant permeability (10
12 m2), density (2000 kg m
3),
thermal conductivity (0.5 W m
1
C
1) and specific heat
(1000 J kg
1
C
1). At the initial condition the uniform
porous (n = 0.4) medium is filled with gas at atmospheric
pressure (i.e., CO2 at 1 bar) and a temperature of 30
C. A fluid
source is located along the bottom boundary, from the
symmetry axis to a maximum radial distance of 1.2 m
and it supplies a gas phase composed of water vapor and
CO2 at a temperature of 100
C, with a H2O/CO2 ratio
consistent with the composition of fumarolic gas of the
modeled system (e.g., H2O/CO2 = 2.2 by weight at
Solfatara). At Solfatara the rate of the gas injection was
assumed 0.0034 kg s
1. This value was chosen to
reproduce the highest measured thermal gradients in the
area (see section 4). The bottom boundary was imperme-
able to the fluid and heat fluxes while the side and the
upper boundaries are open with respect to heat and fluid
flux, with temperature and pressure values fixed at the
initial conditions (30
C and atmospheric pressure). After
200– 400 days of injection the simulated temperature and
the pressure values reach constant values.
4. Results and Discussion
4.1. Chemical Composition of the Fumaroles
[24] The main component of the fumaroles within the
different DDSs is generally H2O followed by CO2 and

5 of 17
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING B08204

Table 2. Summary of Measured and Estimated Parameters in 15 DDSs in the Worlda

Number of rT, jCO2, jGAS, jH2O, Kc, HF,
DDS Date Measurements
C m
1 mol m
2 s
1 xH2O/xGAS mol m
2 s
1 mol m
2 s
1 W m
1
C
1 W m
2
Campi Flegrei Volcanic System
Solfatara Mar 1999 30 107 4.79E-04b 5.392 4.87E-04 2.63E-03 1 107
Apr 1999 30 130 5.99E-04 5.601 6.09E-04 3.41E-03 1.07 139
May 1999 30 141 5.73E-04 5.597 5.83E-04 3.26E-03 0.94 132
May 1999 200 160 6.58E-04 5.597 6.68E-04 3.74E-03 0.95 152
Jun 1999 30 121 4.80E-04 5.593 4.88E-04 2.73E-03 0.92 111
Jul 2000 94 170 4.80E-04 5.568 4.88E-04 2.72E-03 0.65 108
Mar 2002 30 171 1.08E-03 5.126 1.10E-03 5.63E-03 1.34 229

Nisyros Volcanic System

Stefanos Sep 1999 84 208 7.28E-05 57.824 9.13E-05 5.28E-03 1.03 214
May 2000 87 183 5.45E-05 57.824 6.91E-05 3.99E-03 0.89 162
Sep 2000 86 294 7.89E-05 62.83 1.05E-04 6.60E-03 0.91 268
Feb 2001 73 241 7.87E-05 65.604 1.04E-04 6.79E-03 1.14 276
May 2001 102 256 5.13E-05 65.334 6.79E-05 4.44E-03 0.7 180
Sep 2001 88 223 6.37E-05 68.181 8.51E-05 5.80E-03 1.06 236
Feb 2002 90 180 6.30E-05 71.181 8.51E-05 6.06E-03 1.36 246
Jul 2002 84 232 3.97E-05 70.669 5.31E-05 3.75E-03 0.66 152
May 2003 87 223 6.38E-05 72.112 7.47E-05 5.39E-03 0.98 219
SU Area May 2003 37 286 2.71E-04 38.758 3.12E-04 1.21E-02 1.72 492
Polybotes Micros May 2001 29 405 6.26E-05 99.357 8.37E-05 8.32E-03 0.83 338
Kaminakia Sep 2000 28 153 1.18E-04 18.608 1.35E-04 2.52E-03 0.67 102

Masaya Volcanic Systemc

Comalito Mar 2003 36 64 5.76E-04 0.724 3.10E-03 2.25E-03 1.43 91

Cerro Negro Volcanic Systemc

Cerro Negro Mar 2003 22 101 9.86E-05 22.05 1.78E-04 3.92E-03 1.58 159

Pantelleria Volcanic System

Favara Grande Jul 2004 188 34 2.35E-05 36.773 2.68E-05 9.86E-04 1.19 40
Favara Piccola Jul 2004 44 178 1.15E-04 26.197 1.40E-04 3.67E-03 0.84 149

Ischia Volcanic System

Donna Rachele Oct 2000 61 201 2.50E-05 269.409 2.55E-05 6.87E-03 1.39 279
May 2001 19 199 2.16E-05 410.769 2.21E-05 9.09E-03 1.86 369

Vulcano Volcanic System

Crater Mar 2002 54 60 7.94E-05 17.874 8.67E-05 1.55E-03 1.05 63
PL Beach Mar 2002 39 100 3.53E-04 8.877 3.65E-04 3.24E-03 1.31 131
PL Beach Mar 2002 66 65 2.23E-04 15.687 2.34E-04 3.67E-03 2.3 149

Vesuvio Volcanic System

NE Rim Feb 2002 21 239 4.87E-04 0.503 2.08E-02 1.04E-02 1.77 424
Bottom Mar 2002 10 168 2.76E-04 7.976 2.80E-04 2.24E-03 0.54 91
a
Measured parameters are the number of measurements, thermal gradient rT, soil CO2 flux jCO2, and fumarolic molar ratio xH2O/xGAS and the estimated
parameters are soil gas flux jGAS, steam flux jH2O, soil thermal conductivity Kc, and heat flux (HF).
b
Read 4.79E-04 as 4.79  10
4.
c
Data collected during the 8th Field Workshop on Volcanic Gases (IAVCEI-CCVG), Nicaragua and Costa Rica, March 2003.

certain other gases. The absence of acidic gases (SO2, HCl,
and HF) and relatively high CH4 contents indicate hydro-
thermal origin of fluids at all DDSs, apart from Cerro Negro
and Vulcano crater DDSs which are characterized by high-
temperature fumaroles rich in acidic gases indicating a more
typical volcanic environment (Table 1).
4.2. CO2 Fluxes and Temperatures in the Soils
[25] The data of CO2 soil fluxes and thermal gradients in
the soil were collected in 15 DDSs during the last 5 years.
The number of measurements varies from 10 to 200,
depending on the different extensions of the studied DDSs.
Campaigns comprising a different number of measure-
ments, distributed in the same area, implicate different
sampling densities (e.g., Solfatara, Stefanos and Donna
Rachele DDSs). A summary of measured and estimated
parameters is shown in Table 2.
[26] The measured parameters are thermal gradients
(rT), soil CO2 fluxes (jCO2) and the ratio between H2O
and the content of noncondensable gases of the fumaroles
(xH2O/xGAS where xGAS is the sum of noncondensable gases
molar fractions). Thermal gradients range from 34
C m
1 at
Favara Grande DDS to 405
C m
1 at Polybotes Micros.
Soil CO2 fluxes range from 2.16  10
5 mol m
2 s
1 to
1.08  10
3 mol m
2 s
1. The molar ratio xH2O/xGAS varies
in a large range from 0.5, at Vesuvio NE Rim, to 410 at
Donna Rachele, Ischia.
[27] The estimated parameters are the flux of gas (jGAS),
the flux of the steam (jH2O), the thermal conductivity of the
soil (Kc) and the soil heat flux (HF). The flux of gas from
soil is computed by multiplying soil CO2 flux (jCO2) times
the ratio xGAS/xCO2. The computed jGAS values are close to
the original jCO2 values in all the DDSs with the exception
of Vesuvio NE Rim and Comalito DDSs (Table 2), where

6 of 17
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING B08204

Figure 2. Vertical soil temperature profiles measured in Solfatara crater. Profiles are grouped into four
classes based on their different shapes and ranges of values. Therm. Grad. and jCO2 are the mean
thermal gradient and the mean CO2 soil flux for all the n samples of the class. Selected profiles of classes
a, b, and c are reported. The grey (with triangles) profile in the class d is relative to the 84 (see text and
Figure 7).

the main component of the fumaroles is air (essentially N2
and O2) followed by H2O and CO2 (Table 1). Similarly the
flux of the steam (jH2O) is computed by multiplying the
measured jCO2 values by the fumarolic ratio xH2O/xCO2.
The derivation of Kc and thermal heat release is described in
detail in subsection 4.4.
[28] The data of the campaign performed at Solfatara
crater in July 2000 (Table 2) were used as a ‘‘base case’’ to
investigate the possible shapes of the thermal gradients.
This campaign consisted of the measurement of 94 vertical
temperature profiles (0.02– 0.40 m depth) and soil jCO2
(Table 2). The temperature profiles were subdivided into
four classes (Figure 2) according to different shapes and
ranges of thermal gradients: class a (comprising 18% of
the profiles) was characterized by low thermal gradients
and, in this case, the effect of the diurnal solar radiation was
relevant, resulting in a decrease in soil temperature with
depth in the uppermost part of the soil (0.05 –0.06 m); class
b (18% of the profiles) includes the profiles with a break
in the slope of the soil temperatures with increased depth,
due to presence of two thermally different layers; class c
(53% of the profiles) was characterized by medium
thermal gradients, constant slope with depth and temper-
atures at 0.25– 0.30 m depth up to 70– 80
C; and class d
(10% of the profiles) included profiles with the highest
thermal gradient and included curves which asymptotically
tend to a common isotherm of 95
C.
[29] We use soil temperatures measured continuously at
Solfatara at 0.1, 0.2, and 0.3 m depths to investigate the
influence of the daily and seasonal thermal variations on
the thermal gradient of the upper part of the soil. Daily and
seasonal thermal variations could potentially modify the
heat flux measured at the surface arising from the input of
the deep hydrothermal system. The soil temperatures at 0.2
and 0.3 m depths were measured from October 2001 to
October 2003, while the 0.1 m depth was measured from
October 2001 to July 2002. Soil temperatures increase with
depth due to the presence of the deep thermal source
(Figure 3). The low mean value of 0.2– 0.3 m thermal
gradient (59
C m
1) indicate that this point belongs at the
class a, in which the effects of daily and seasonal thermal
excursions are most evident. The high frequency variations
in soil temperature are due to the rainfall episodes which
also produce positive peaks in the soil humidity (grey line
in Figure 3). The amplitude of these temperature fluctua-
tions is damped with depth. This suggests that measure-
ment of the vertical temperature gradient should be avoided
in the rainy (transient) periods, the effects of which

7 of 17
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING B08204

Figure 3. Soil temperature at three depths (0.1, 0.2, and 0.3 m) and relative humidity of the soil at
0.15 m depth in the same site (grey line). The uppermost sensor (0.1 m) was in operation until the end
of July 2002. The positive peaks in the soil humidity coincide with rainfall episodes.

generally tend to last 4 – 5 days. The seasonal (or diurnal)
variations of the temperature can be described by a
sinusoidal law:
T ¼ Ta þ Ao sinðwt þ fÞ ð1Þ
where A0 and Ta are the amplitude of the seasonal (or
diurnal) temperature wave and the average annual (or daily)
soil temperature at z = 0, w is 2p/t, where t is the period of
the cycle (1 year or 1 day), t is the time and f is the initial
phase of the temperature wave. Using relation (1) we fitted
the seasonal temperature variations of the Figure 3 which
are evident at both investigated depths (the shallower sensor
was in operation until July 2002). The amplitude of the
seasonal temperature fluctuations (Ao = 9.6
C) does not
change significantly between 0.2 and 0.3 m depth (Figure 4).
This means that temperatures at 0.2 m and 0.3 m depth
reflect the periodic surface variations in the same way and
that the thermal gradient, at least in the 0.2– 0.3 m depth,
has a comparable mean value for different seasons.
Moreover, using continuous data from Solfatara, it is
evident that the daily temperature excursions of the soil
(SDTE) are minimal (few percentage points) compared to
the average daily temperature (SDAT) at those same depths
(Table 3 and Figure 4 insert). In other words, at depths
higher than 0.2 m in the hot soil of volcanic-hydrothermal
areas, the ‘‘stationary’’ component of temperature, reflect-
ing the endogenous effects, is dominant relative to the
external temperature changes (due to solar heating, wind,
etc.).
4.3. Modeling Results and Comparison to
Measured Gradients
[30] The results of the TOUGH2 modeling at the steady
state (9000 days) are presented in Figures 5a – 5c and 6a –
6c. Figures 5a, 5b, and 5c, which represent the mass flux of
steam, liquid and CO2, respectively, describe a convective
plume of hot carbon dioxide-steam mixture extending from
the source toward the surface. The composition of the

Figure 4. Soil temperature at 0.2 and 0.3 m depths and relative sinusoidal fitting. The correlation
coefficient between measured and estimated values is 0.939 for both recordings. Insert shows daily
fluctuations in 1-week period.

8 of 17
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING B08204

Table 3. Summary of Results of the Continuous Soil Temperature phase advection is therefore the primary mode of heat
Monitoring at Solfatara Crater transfer in this region. Some energy is transported laterally
Depth, m SDATa SDTEb by the liquid phase produced by steam condensation
0.1 29.27
C c
4.13
C (14.1%)c
(Figure 6b). Near the surface, where the vapor phase is
0.2 35.18
C 2.32
C (6.6%) negligible, the heat is primarily transferred by conduction
0.3 40.73
C 0.98
C (2.4%) (Figure 6c). Numerical simulations show that the original
a
SDAT, soil daily average temperature. heat of the deep source (6348 W) is almost totally trans-
b
SDTE, soil daily thermal excursion (average value). The values in ferred at the surface (5940 W corresponding at 93.6% of the
parentheses are the percentage value of the ratio SDTE/SDAT. original heat). Of this amount about 98.4% is released by
c
Data from October 2001 to July 2002.
conduction and causes linear thermal gradients in the upper
part of the domain. A larger fraction of the heat transported
gaseous mixture changes during its ascent: the upflowing by conduction is computed when the amount of energy
steam (Figure 5a) condenses and produces a liquid phase injected in the system is reduced. This finding suggests
that flows from the border of the plume toward the bottom that measurements of vertical thermal gradients in the upper
of the domain (Figure 5b), whereas CO2, which is noncon- part of the soil (0.02– 0.40 m), provide a good estimate of
densable gas, is discharged almost entirely at the soil the total energy transferred by soil diffuse degassing at
surface, extending to a maximum radial distance of 4 m Solfatara.
(Figure 5c). [32] The injection rate of the steam-gas mixture for the
[31] Figures 6a –6c show the modes of the heat transfer simulation (Figures 5 and 6) was chosen to reproduce, in
associated with the rising fluid. The heat is transferred at the correspondence with the symmetrical axis of the domain,
surface by advective and conductive processes (Figure 6a). one of the highest thermal gradients measured at Solfatara
At depth the distribution of temperature is almost uniform crater (i.e., sample n.84 rT = 562
C m
1, see Figure 2).
and the conductive heat transfer rate can be neglected. Two- The correlation between simulated and measured temper-

Figure 5. Simulation results showing (a) steam, (b) liquid and (c) CO2 fluxes. The initial composition
of the carbon dioxide-steam mixture changes during its ascent: (a) most of the steam condenses in the soil
and (b) produces a liquid phase that flows toward the bottom of the domain, (c) whereas carbon dioxide is
discharged almost entirely at the soil surface.

9 of 17
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING B08204

Figure 6. Simulation results showing (a) total, (b) advective, and (c) conductive heat fluxes. At depths,
where two-phase advection dominates heat transport, the distribution of temperatures is almost uniform.
In Figure 6a, heat associated with the rising fluid is transferred by both advection and conduction toward
the surface. In Figure 6b, some heat is transported laterally by advection by the liquid phase. In Figure 6c,
near the surface, the heat is mainly transferred by conduction.

atures at depth is very good (Figure 7) and shows that
temperatures asymptotically reach T  95
C at a depth of
0.15 m. We reproduced the thermal gradients measured in
other areas, for example at Stefanos crater and Comalito
(Figure 7), by changing xH2O/xGAS ratio of the injected
mixture (Table 2), rock properties (i.e., thermal conductivity
computed in the subsection 4.4 and reported in Table 2), and
the injection rate of the fluid (0.0040 kg s
1 for Stefanos
crater and 0.0109 kg s
1 for Comalito). Here we find the
shape of the vertical thermal profile strongly depends on the
initial content of noncondensable gases. In the case of
Stefanos crater, where the noncondensable gas content is
smaller (i.e., xH2O/xGAS = 65.7) than in the fumarolic fluid
of Solfatara (i.e., xH2O/xGAS = 5.48), the asymptote of the
curve is at 99
C and the curve shows a clear change from
advective to conductive regime. The conductive component
of heat transfer can be observed by the linear change of
temperature with depth, whereas the advective component is
represented by constant temperature with depth. At Coma-
lito, where fluid emissions are richer in noncondensable
gases (i.e., xH2O/xGAS = 0.72), the thermal profile is
nonlinear in the entire range of the considered depths and
shows an asymptote at 79
C. In this case, the advective
heat transfer occurs until the surface. However, even at
Comalito, where the contribution of advection to total heat
transport near the surface is higher relative to the Solfatara
and Stefanos cases, 84% of the total heat is still transferred
by conduction. This finding suggests that in the uppermost
part of the soil, the heat released through the diffuse
degassing process is mainly by conduction in all the
modeled cases. Similar conclusions were suggested by the
results of other simulations performed where the controlling
parameters (i.e., depth of the injection, properties of the
rocks) were varied.
[33] Our finding contrasts with the results of a recent
work in an area of steaming ground at Karapiti, New
Zealand [Hochstein and Bromley, 2004]. Hochstein and
Bromley [2004, p. 131] concluded that the total thermal
flux at the surface of ‘‘steaming ground’’ consists ‘‘of a
convective and a conductive component, even in the ab-

10 of 17
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING
Figure 7. Simulated and measured soil temperatures at Solfatara, Stefanos, and Comalito.
sence of any visible steam discharge at the surface.’’ At
Karapiti the measurements were conducted using a new
type of calorimeter which consists of a cylindrical vessel
containing up to 1.5 L water and a very thin and flat
stainless steel bottom. The heat flux measurements are
made by placing the calorimeter over the hot soils and
measuring the temperature increase of the water in the
vessel. Heat fluxes between 30 and 2000 W m
2 were
measured. At sites where heat fluxes were greater than
200 W m
2, Hochstein and Bromley [2004] found the
ground was usually slightly moist at the surface and traces
of liquid were visible on the underside of the calorimeter
when it was lifted. This was interpreted as the contribution
of diffuse steam which condenses at the cool (stainless steel)
bottom of the calorimeter. A first observation is that the heat
fluxes measured in our study, except in a few cases, were
typically lower than 200 W m
2 (Table 2) and, conse-
quently, no diffuse steam emission should occur according
to the results of Hochstein and Bromley [2004]. However, in
order to understand the origin of the moisture at the soil
surface at Karapiti, we modeled a similar case. We selected
high input rate of steam (0.046 kg s
1) in order to obtain
at the surface, in the axial zone, a very high heat flux
(1000 W m
2). Figure 8a shows the simulated liquid
saturation of the soil (Sl) in the axial zone as a function of
depth. A high concentration of liquid (condensate) forms at
a depth of only 0.1 m, corresponding to the maximum
production of condensates. It is worth noting that measured
Sl profiles at Karapiti [Hochstein and Bromley, 2004] are
very similar to that shown in Figure 8a. Simulation results
show a significant flux of liquid water from the depth at
which condensation occurs toward the surface (Figure 8b),
while only a small amount of steam is emitted (Figure 8c).
In conclusion, the simulation suggests that the liquid on the
underside of the calorimeter of the Karapiti measurements
was the liquid condensed at a shallow depth and possibly
transferring to the surface by capillarity and evaporating.
This result suggests that even at the Karapiti area, the heat is
released mainly by conduction in the upper part of the soil.
11 of 17
B08204
4.4. Heat Transfer in Diffuse Degassing Structures:
Theoretical Considerations and Field Observations
[34] Assuming conduction is the main mode of heat
transfer in the uppermost portion of a soil affected by
diffuse degassing, a simple heat balance between the source
zone and the conductive zone gives
Kc * rT ffi QTj ; ð2Þ
where the terms on the left represent the heat conducted in
the upper part of the soil (Kc is thermal conductivity of the
soil, rT is average thermal gradient), while QTj is the total
energy injected at temperature Tj. The assumption expressed
by the equation (2) is pivotal for the following applications
where we assume that the thermal gradients measured in the
upper part of the hot soils are directly proportional to the
total heat transported by diffuse degassing. To validate this
assumption, we first derive a theoretical relationship
between the ratio of the thermal gradient to jGAS at the
surface and the initial H2O/gas ratio (xH2O/xGAS)Tj of the
source of the hydrothermal or volcanic gas emissions, and
then demonstrate that the data collected at different volcanic
systems validate the theoretical relationship.
[35] In the injection zone, the total heat flux (QTj) is equal
to the heat transported by the vapor phase which is
approximately given by the heat content of the steam
HH2O(Tj,g) at the injection temperature (Tj) times the steam
mass flux (jH2O(Tj)). Contribution of CO2 is neglected, as
verified in the numerical simulation. Expressing jH2O(Tj) as
a function of jGAS (in mol m
2 s
1) and of the molar ratio
(xH2O/xGAS)Tj of the injected gas mixture, QTj is given by:
QTj ffi jGAS * ðxH2O =xGAS ÞTj * HH2OðTj;gÞ ð3Þ
Now we examine the process of energy transfer in a one-
dimensional domain, i.e., in which the heat and mass flows
are constrained to the z direction. Because of the process of
continuous condensation, we can assume the presence of
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING B08204

Figure 8. (a) Liquid saturation (Sl) versus depth in the axial zone of the simulated domain. A
condensate-rich level at 0.1 m depth is formed. (c) Upward movement of the vapor phase and (b) the
downward movement of the liquid phase produced by the condensation of the steam in the uppermost
part of the soil are evident. Only a small amount of steam is emitted from the surface (Figure 8c).

liquid water at depth and compute the pressures of the
gaseous species at any temperature Tn of the soil. The
partial pressure of H2O (PH2O) is a function of the temperature
(Steam Tables [Keenan et al., 1969]) and the pressure of the
noncondensable gases (PGAS), primarily CO2. Assuming that
the total pressure (Ptot) of the fluids
1 bar, which is in the cases of Stefanos, Kaminakia, and Solfatara) the
consistent with the shallow depth of the investigated soil
layer, this relationship is given as PGAS = Ptot
PH2O ffi 1

PH2O. If we assume the jGAS remains constant with depth, the
heat transported by the vapor phase (QH2O(Tn,g)) at any
temperature Tn is given by
QH2 OðTn;gÞ ¼ jGAS * ðxH2O =xGAS ÞTn * HH2 OðTn;gÞ
ffi jGAS * ðPH2O =PGAS ÞTn * HH2 OðTn;gÞ ð4Þ
where the molar fraction ratio xH2O/xGAS is equal to PH2O/
PGAS assuming fugacity coefficients equal to 1, and the heat
content of the steam HH2O(Tn,g) is known at any temperature
(Steam Table [Keenan et al., 1969]).
[36] Dividing equation (4) by equation (3) we obtain the
relation
h i
QH2 OðTn;gÞ =QTj ffi ðPH2O =PGAS ÞTn * HH2 OðTn;gÞ
h i
= ðxH2O =xGAS ÞTj * HH2OðTj;gÞ ð5Þ
[ 37 ] The percentage ratios Q H 2 O (Tn,g) /Q Tj computed
through equation (5) for temperatures between Tj and
Tmin (where Tmin is the minimum temperature measured
in the vertical profiles corresponding to the shallower depth)
are reported versus temperature in Figure 9. Figure 9
considers five different volcanic systems (Solfatara,
Stefanos, Kaminakia, Vesuvio, and Comalito), which are
characterized by different (xH2O/xGAS)Tj ratios measured at
the respective fumaroles. In general, the shape of the
QH2O(Tn,g)/QTj versus T curve largely depends on the ratio
(xH2O/xGAS)Tj. For relatively high (xH2O/xGAS)Tj ratios (i.e.,
contribution of the heat transported by the steam to the total
energy quickly decreases with the temperature becoming
negligible (QH2O(Tn,g)/QTj < 0.10) at T < 70
C. Below this
temperature, the heat is transported almost exclusively by
conduction and linear trends in the soil thermal profiles are
expected. In fact, all measured thermal gradients were linear
at the DDSs characterized by fumarolic fluids with xH2O/
xGAS ratios greater than 5.4 (i.e., Solfatara, Kaminakia and
Stefanos) at temperatures <70
C. At Comalito and Vesuvio
NE Rim, characterized by fumarolic xH2O/xGAS molar ratios
of 0.724 and 0.503, respectively, our calculations suggest
that significant amount of heat is transported by advection
at relatively low temperatures (Figure 9). In agreement with
these calculations, linear thermal profiles at NE Vesuvio
Rim and at Comalito were measured only at temperatures
below 35– 40
C.
[38] Results of the two-dimensional simulations and the
one-dimensional theoretical model reported above suggest
that almost all of the energy transferred by advection in the
deeper part of the soil is transferred primarily by conduction
in the upper part of the soil, with heat transfer capacity
depending on the initial content of noncondensable gases.
Combining equations (2) and (3) in a logarithmic form, the
following relation is obtained:
 
logðxH2O =xGAS ÞTj ¼ log Kc =HH2OðgÞ þ logðrT=jGAS Þ ð6Þ

12 of 17
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING B08204

Figure 9. Contribution of the heat transported advectively by the steam to the total energy in the range
of measured temperature. In the systems with xH2O/xGAS ratios above 5.4 (Solfatara, Kaminakia, and
Stefanos) the contribution of the heat transported by the steam to the total energy becomes negligible
(QH2O(Tn,g)/QTj < 0.10) for T < 70
C. At Comalito and Vesuvio, characterized by lower fumarolic
xH2O/xGAS ratios, a significant amount of heat is transported by advection process until a few
centimeters from the surface.

Equation (6) highlights a linear relation between the molar
ratios (xH2O/xGAS)Tj of the fluids before condensation
and the ratios rT/jGAS with a proportional factor equal to
Kc/HH2O(g). Assuming that HH2O(g) does not significantly
vary in the temperature range of interest (HH2O(g) ffi
40626 J mol
1), equation (6) describes a family of
rectilinear curves with slope 1 and intercepts propor-
tional to Kc on a (xH2O/xGAS)Tj versus rT/jGAS log
diagram (Figure 10). Figure 10 demonstrates that the
fumarolic xH2O/xGAS ratios are proportional to the average
values of the measured rT/jGAS for 30 campaigns in 15
different DDSs. The measured data plot along the curve
Kc = 1 and are well correlated (R2 = 0.954). The Kc
values, computed for each set of data using the measured
values (equation (6)), range from 0.5 to 2.3 W m
1
C
1
(Table 2) with an average value of 1.14 ± 0.4 W m
1
C
1,
which are plausible values for hot soils of hydrothermal areas
[Brombach et al., 2001; Sorey and Colvard, 1994]. Moreover,
thermal conductivities calculated based on data collected
during different periods of the year in the same DDS (e.g.,
Solfatara and Stefanos), are higher in the wet season than the
dry season. These results are expected as Kc is a function of
the soil water content. In order to compare the calculated
thermal conductivities shown in Figure 10 with those
calculated using an independent method, we applied the
relation of Woodside and Messmer [1961] that allows direct
derivation of the soil thermal conductivity. They suggest a
geometric average law between the conductivities of the soil
(Kc), and that of the solid (Ks), water (Kw) and air (Ka)
fractions of the soil:
Kc ¼ Ksxs * Kwxw * Kaxa ð7Þ
where xs, xw, and xa are the volumetric content of solid, water
and air, respectively. Of the three constituent phases, the
solid phase has the highest conductivity. Therefore
conductivity is increased as bulk density increases. Air
is a poor conductor and in soil it reduces the value of
total conductivity. Water content has a marked effect
because when water replaces air it provides bridges
between particles that greatly increase the conductivity of
the soil. Assuming the thermal properties of soil constituents
(Ks = 2.9 W m
1
C
1, Kw = 0.6 W m
1
C
1 e Ka =
0.025 W m
1
C
1) reported by de Vries [1963], varying
the porosity n from 0.20 to 0.50, a suitable range for the
studied soils (e.g., at Solfatara, n was measured to be
0.46 [Nunziata et al., 1999]), and considering that in
temperate climates commonly xw > 0.15, we obtain thermal
conductivities Kc ranging from 0.4 to 2.1 W m
1
C
1
which practically overlap the range of the values obtained
in the Figure 10 (from 0.5 to 2.3 W m
1
C
1) for the
30 campaigns.
[39] An interesting further consideration can be drawn by
the previous results. In previous studies [Chiodini et al.,
2001a; Brombach et al., 2001] the soil thermal conductivity
was assumed equal to 1 W m
1
C
1 following Clauser and
Huenges [1995] or was chosen as ‘‘appropriate values’’
[Severne and Hochstein, 1994] to assess the surface heat
transfer. In this work, the thermal conductivities of the soil,
derived from the measured parameters (i.e., soil thermal
gradients and diffuse CO2 flux) and from the composition of
the fumarolic fluids in the same area, are directly estimated
by the adopted procedure.
[40] The proposed method for the estimation of the soil
thermal conductivity is affected by the uncertainties of the
measured parameters, i.e., of the molar ratio (xH2O/xGAS)Tj

13 of 17
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING B08204

Figure 10. Linear relation among the log ratios xH2O/xGAS of the fumarolic fluids and the measured
log ratios rT/jGAS for 30 campaigns in 15 different DDSs. Error bars refer to the effects of 50%
condensation of the original steam and to an uncertainty of ±30% in the estimation of rT/jGAS (see text
for explanation).

of the fluids before condensation, which is assumed equal to
that measured in the fumarolic effluents, and of rT/jGAS,
i.e., the ratio between the average thermal gradients and the
average soil gas flux. The error of the ratio rT/jGAS is
mainly due to the uncertainty in the estimation of the mean
values from a finite number of measurements. This matter
has been discussed in detail in previous studies regarding
CO 2 gas flux measurements [Chiodini et al., 1998;
Cardellini et al., 2003]. The horizontal error bar reported
in Figure 10 refers to the uncertainty of the rT/jGAS ratio
of ±30%, which is considered here as a reasonable maxi-
mum for the expected error.
[41] The xH2O/xGAS ratios of the fumarolic fluids are
minimally affected by analytical errors, yet can be not fully
representative of the fluids at depth mainly because the
occurrence of condensation before the discharge. This
problem would most likely affect the fumaroles at boiling
temperature while it should not affect the fumaroles with
temperatures higher than the boiling point (such as at
Solfatara, Vulcano crater, and Cerro Negro). The uncer-
tainties due to condensation generally cannot be directly
quantified because the compositions of the fluids in the
subsoil are not available. A comparison between deep
fluids and surface discharges is possible only for the
fumaroles located in correspondence with drilled geother-
mal systems. This is the case of Larderello (Italy) where
the composition of three fumaroles at boiling temperature
(Mr1, Mr2, and Mr3 of Chiodini and Marini [1998]) can
be compared with the composition of the pure vapor phase
discharged by two geothermal wells of the area (wells M21
and P2 of Chiodini et al. [1991]). The xH2O/xGAS ratio of
the fumaroles ranges from 78 to 130 and it is close to that
of the geothermal wells (67 and 102) suggesting that the
composition of the fumarolic effluents is representative of
the vapor at depth. In general, we believe that the conden-
sation effects are minor in high flow rate fumaroles, such
as most of those sampled in this work. Some indirect
confirmation of the minor or nil effect of condensation in
high flow rate fumarole is given for example by the
comparison between the xH2O/xGAS ratios of three fumar-
oles at boiling point (95–97
C) and of two fumaroles at
high temperatures (140 – 160
C) sampled at Solfatara
(Campi Flegrei) in March, April and June 1999 (unpub-
lished data of Osservatorio Vesuviano, INGV). The aver-
age xH2O/xGAS ratios in the boiling point fumaroles (5.4,
6.0, and 6.3 in March, April and June, respectively) were
close to those in the high-temperature fumaroles (5.3, 5.7,
and 5.9). The effects of condensation can be important in
low flow rate fumaroles, and can be roughly evaluated by
studying the stable isotope composition of steam. For
example, a previous study on the geochemistry of Nisyros
fumaroles [Brombach et al., 2003] ruled out the occurrence
of significant condensation at Stefanos and Polibotes
micros, while condensation affected the fumaroles at
Kaminakia. In this study we thus considered the xH2O/
xGAS of the Kaminakia fumarole richest in water and least
affected by condensation. For reference the effect of 50%
condensation, here considered as a maximum, is shown by
the vertical error bar in Figure 10.
[42] In conclusion the good correlation between the two
independent data sets of 15 DDSs from 8 different volcanic
systems (Figure 10) suggests that these uncertainties are
minor, and that the estimation of the thermal energy and of
the thermal soil conductivities, based on soil temperature

14 of 17
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING B08204

Table 4. Total CO2 Output and Energy Release from DDSs

DDS Area,b H2O/CO2 Steam, Energy Release,
Volcano, DDS a
Date m2 CO2, t d
1 Fumaroles t d
1 MW
Campi Flegrei, Solfatara Jul 2000 620,000 1480 2.27 3359.6 100.8
Ischia, Donna Rachele Apr 2001 57,600 9.1 147 1337.7 40.1
Nisyros, all DDSs 2000 – 2002 903,000 68 6.1 – 36 1420 42.6
Vesuvio, cone Apr 2000 331,000 151 3.66 552.66 16.6
Vulcano, crater Jul 1998 415,000 158 4.42 698.36 21
Vulcano, PL Beach Mar 2002 17,700 19 5.24 99.56 3
Pantelleria, Favara Grande Jul 2004 57,700 7 17.2 120.4 3.6
Masaya, Comalito Mar 2003 10,300 19 1.6 30.4 0.9
Yellowstone, Mud volcanoes Aug 1993 400,000 290 3.49 1021.1 30.4
a
The data of Campi Flegrei, Ischia, Nisyros, Vesuvio, Masaya, and Yellowstone are from the literature [Cardellini et al., 2003; Chiodini et al., 2004;
Caliro et al., 2004; Lewicki et al., 2003; Werner and Brantley, 2003]. The data for Vulcano crater, Vulcano PL beach, and Pantelleria are unpublished data
of Osservatorio Vesuviano, INGV.
b
The area of DDS is computed as the area where CO2 fluxes are the double of local background values.

gradient and soil CO2 flux measurements, is a reliable
methodology.
4.5. Total Heat Released by DDS
[43] Figure 10 shows that each hydrothermal or volcanic
system is characterized by a typical source of fluid which
feeds both the fumarolic system and the diffuse degassing
through the soil. In the upper part of the soil, where the
thermal gradient is measured, the heat associated with the
condensation of the steam is almost totally transferred by
conduction. The heat released by diffuse degassing pro-
cesses can be estimated by (1) measuring the total CO2
output, (2) analyzing the chemical composition of fumar-
oles of the area that can be considered representative of the
composition of the fluids before the steam condensation,
and (3) computing the total steam output and the heat
released during the condensation process and cooling of the
condensates to ambient temperature. Table 4 shows the
results obtained for the investigated systems. For Solfatara,
Vesuvio, Ischia, Nisyros, Comalito and Yellowstone we
referred to published data of total CO2 output [Cardellini et
al., 2003; Chiodini et al., 2004; Caliro et al., 2004; Lewicki
et al., 2003; Werner and Brantley, 2003], while for Vulcano
crater, Vulcano beach, and Pantelleria, we used unpub-
lished data. These energy releases constitute a significant
term in the energy balance of these quiescent volcanoes
and can be used as baselines for energy release against
which changes due to possible volcanic unrest periods can
be compared.
5. Conclusions
[44] Quiescent volcanoes can discharge significant quan-
tities of energy through soil diffuse degassing (e.g., from 1
to 100 MW for the cases in Table 4). We presented a
reliable methodology that permits estimation of the amount
of energy released from volcanic/hydrothermal systems.
Measurements of vertical soil thermal profiles and soil CO2
fluxes carried out in 15 diffuse degassing structures, which
are characterized by different sources of fluid, were used.
We assumed that the fluids which feed the diffuse degass-
ing process and the fumarolic vents within each thermal
area have the same composition at depth. Using the CO2 as
tracer of the degassing process, we found a linear relation-
ship between the fumarolic ratios xH2O/xGAS of the fluids
before condensation and the ratios rT/jGAS in the shal-
lower part of the soil, where most of the steam has already
condensed and the heat is principally transferred by con-
duction. The two independent data sets show a significant
positive correlation (R2 = 0.954) indicating an average soil
thermal conductivity of 1.14 ± 0.4 W m
1
C
1. Different
volumetric content of the soil components (water, solid
matrix, and air) and different soil temperatures can explain
the variation in the thermal conductivity values from 0.5 to
2.3 W m
1
C
1. These estimated values are plausible for
the thermal conductivities of hot soils and consequently
allow us to directly confirm the procedure to derive the
thermal heat flux from the soil in the DDSs.
[45] Fluid and heat transport were simulated through the
uppermost part of the Solfatara hydrothermal system. This
model highlighted a simultaneous upward movement of the
vapor phase and a downward movement of the liquid phase
produced by the condensation of the steam in the uppermost
part of the soil. The heat associated with the rising fluid is
transferred at the surface by advective and conductive
processes. Near the surface, where the vapor phase is
negligible, the heat is primarily transferred by conduction
and the soil temperature profile linearly changes with depth.
By changing the composition and the rate of the injected
mixture and choosing appropriate rock properties, we were
able to reproduce the thermal gradients measured in other
areas (e.g., at Stefanos crater and Comalito) and show that
heat is almost completely transported by conduction in the
uppermost part of the soil thereby generating linear thermal
gradients.
[46] These endogenous heat fluxes constitute a ‘‘station-
ary’’ component in the hot soils of DDSs which is dominant
relative to the heat fluxes generated by external temperature
changes (diurnal and seasonal variations), as verified by the
continuous monitoring performed at Solfatara.
[47] Acknowledgments. We wish to thank the staff of the Istituto di
Geochimica dei Fluidi, INGV-Palermo, who sampled and analyzed the
fumaroles of the Nicaraguan volcanoes. This work was financially sup-
ported by the National Volcanic Group of Italy (GNV-INGV) and by
European Community (Geowarn Project).
References
Brombach, T., J. C. Hunziker, G. Chiodini, C. Cardellini, and L. Marini
(2001), Soil diffuse degassing and thermal energy fluxes from the south-
ern Lakki plain, Nisyros (Greece), Geophys. Res. Lett., 28, 69 – 72.
Brombach, T., S. Caliro, G. Chiodini, J. Fiebig, J. C. Hunziker, and B. Raco
(2003), Geochemical evidences for mixing of magmatic fluids with sea-

15 of 17
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING
water, Nisyros hydrothermal system, Greece, Bull. Volcanol., 65, 505 –
516, doi:10.107/s00445-003-0278-x.
Brown, G., H. Rymer, J. Dowden, P. Kapadia, D. Stevenson, J. Baquero,
and L. D. Morales (1989), Energy budget analysis for Poàs crater lake:
Implications for predicting volcanic activity, Nature, 339, 370 – 373.
Caliro, S., G. Chiodini, G. Galluzzo, D. Granieri, M. La Rocca, G. Saccorotti,
and G. Ventura (2004), Recent activity of Nisyros volcano (Greece)
inferred from structural, geochemical and seismological data, Bull. Volca-
nol., 67, 358 – 369, doi:10.1007/s00445-004-0381-7.
Carapezza, M. L., and D. Granieri (2004), CO2 soil flux at Vulcano (Italy):
Comparison of active and passive methods and application to the identi-
fication of actively degassing structure, Appl. Geochem., 19, 73 – 88.
Cardellini, C., G. Chiodini, and F. Frondini (2003), Application of stochas-
tic simulation to CO2 flux from soil: Mapping and quantification of gas
release, J. Geophys. Res., 108(B9), 2425, doi:10.1029/2002JB002165.
Chiodini, G., and L. Marini (1998), Hydrothermal gas equilibria: The H2O-
H2-CO2-CO-CH4 system, Geochim. Cosmochim. Acta, 62, 2673 – 2687.
Chiodini, G., R. Cioni, B. Raco, and G. Scandiffio (1991), Carbonyl sul-
phide (COS) in geothermal fluids: An example from the Larderello Field
(Italy), Geothermics, 20(5/6), 319 – 327.
Chiodini, G., R. Cioni, C. Leonis, L. Marini, and B. Raco (1993a), Fluid
geochemistry of Nisyros island, Dodecanese, Greece, J. Volcanol.
Geotherm. Res., 56, 95 – 112.
Chiodini, G., R. Cioni, and L. Marini (1993b), Reactions governing the
chemistry of crater fumaroles from Vulcano Island, Italy, and implications
for volcanic surveillance, Appl. Geochem., 8, 357 – 371.
Chiodini, G., R. Cioni, L. Marini, and C. Panichi (1995), Origin of the
fumarolic fluids of Vulcano Island, Italy, and implications for the volca-
nic surveillance, Bull. Volcanol., 57, 99 – 110.
Chiodini, G., F. Frondini, and B. Raco (1996), Diffuse emission of CO2
from the Fossa crater, Vulcano Island (Italy), Bull. Volcanol., 58, 41 –
50.
Chiodini, G., R. Cioni, M. Guidi, B. Raco, and L. Marini (1998), Soil CO2
flux measurements in volcanic and geothermal areas, Appl. Geochem.,
13, 543 – 552.
Chiodini, G., F. Frondini, C. Cardellini, D. Granieri, L. Marini, and
G. Ventura (2001a), CO2 degassing and energy release at Solfatara
Volcano, Campi Flegrei, Italy, J. Geophys. Res., 106, 16,213 – 16,221.
Chiodini, G., L. Marini, and M. Russo (2001b), Geochemical evidences of
high temperature hydrothermal brines at Vesuvio volcano (Italy), Geo-
chim. Cosmochim. Acta, 65(13), 2129 – 2147.
Chiodini, G., M. Todesco, S. Caliro, C. Del Gaudio, G. Macedonio, and
M. Russo (2003), Magma degassing as a trigger of bradyseismic events:
the case of Phlegrean Fields (Italy), Geophys. Res. Lett., 30(8), 1434,
doi:10.1029/2002GL016790.
Chiodini, G., R. Avino, T. Brombach, S. Caliro, C. Cardellini, S. De Vita,
F. Frondini, D. Granieri, E. Marotta, and G. Ventura (2004), Fumarolic
and diffuse soil degassing west Mount Epomeo, Ischia (Italy), J. Volca-
nol. Geotherm. Res., 133, 291 – 309.
Cioni, R., E. Corazza, and L. Marini (1984), The gas/steam ratio as indi-
cator of heat transfer at the Solfatara fumaroles, Phlegrean Fields (Italy),
Bull. Volcanol., 47, 295 – 302.
Civetta, L., Y. Cornette, P. Y. Gillot, and G. Orsi (1988), The eruptive
history of Pantelleria (Sicily Channel) in the last 50 ka, Bull. Volcanol.,
50, 47 – 57.
Clauser, C., and E. Huenges (1995), Thermal conductivity of rocks and
minerals, in Rock Physics and Phase Relations: A Handbook of Physical
Constant, AGU Ref. Shelf. Ser., vol. 3, edited by T. J. Ahrens, pp. 105 –
126, AGU, Washington, D. C.
Connor, C., B. Hill, P. LaFemina, M. Navarro, and M. Conway (1996),
Soil 222Rn pulse during the initial phase of the June – August 1995
eruption of Cerro Negro, Nicaragua, J. Volcanol. Geotherm. Res., 73,
119 – 127.
D’Alessandro, W., G. Dongarra, S. Gurrieri, F. Parello, and M. Valenza
(1994), Geochemical characterization of naturally occurring fluids on the
island of Pantelleria, Miner. Petrogr. Acta, 27, 91 – 102.
de Vries, D. A. (1963), Thermal properties of soils, in Physics of Plant
Environment, edited by W. R. van Wijk, pp. 210 – 235, Elsevier, New
York.
Evans, W. C., M. L. Sorey, B. M. Kennedy, D. A. Stonestrom, J. D. Rogie,
and D. L. Shuster (2001), High CO2 emissions through porous media:
transport mechanisms and implications for flux measurement and frac-
tionation, Chem. Geol., 177, 15 – 29.
Frondini, F., G. Chiodini, S. Caliro, C. Cardellini, D. Granieri, and
G. Ventura (2004), Diffuse CO2 degassing at Vesuvio, Italy, Bull. Volca-
nol., 66, 642 – 651.
Giggenbach, W. F. (1975), A simple method for the collection and analysis
of volcanic gas samples, Bull. Volcanol., 39, 132 – 145.
Giggenbach, W. F. (1991), Chemical techniques in geothermal exploration,
in Application of Geochemistry in Geothermal Reservoir Development,
16 of 17
B08204
coordinated by F. D’Amore, pp. 119 – 144, United Nations Inst. for Train.
and Res., New York.
Giggenbach, W. F., and R. L. Goguel (1989), Collection and analysis of
geothermal and volcanic water and gas discharges, Rep. CD.2401, Dep.
of Sci. and Ind. Res., Chem. Div., Petone, New Zealand.
Granieri, D., G. Chiodini, W. Marzocchi, and R. Avino (2003), Continuous
monitoring of CO2 soil diffuse degassing at Phlegraean Fields (Italy):
Influence of environmental and volcanic parameters, Earth Planet. Sci.
Lett., 212, 167 – 179.
Hochstein, M. P., and C. J. Bromley (2004), Measurement of heat flux from
steaming ground, Geothermics, 34, 131 – 158.
Hurst, A. W., H. M. Bibby, B. J. Scott, and M. J. McGuinness (1991), The
heat source of Ruapehu Crater Lake: Deductions from the energy and
mass balances, J. Volcanol. Geotherm. Res., 46, 1 – 20.
Keenan, J. H., F. G. Keyes, P. G. Hill, and J. G. Moore (1969), Steam
Tables: Thermodynamic Properties of Water Including Vapor, Liquid, and
Solid Phases, 162 pp., John Wiley, Hoboken, N. J.
Lardy, M., and A. Tabbagh (1999), Measuring and interpreting heat fluxes
from shallow volcanic bodies using vertical temperature profiles: A pre-
liminary test, Bull. Volcanol., 60, 441 – 447.
Lewicki, J. L., C. Connor, K. St-Amand, J. Stix, and W. Spinner (2003),
Self-potential, soil CO2 flux, and temperature on Masaya volcano, Ni-
caragua, Geophys. Res. Lett., 30(15), 1817, doi:10.1029/2003GL017731.
McKnight, S. B., and S. N. Williams (1997), Old cinder cone or young
composite volcano? The nature of Cerro Negro, Nicaragua, Geology, 25,
339 – 342.
Mercalli, G., and O. Silvestri (1891), Le eruzioni dell’isola di Vulcano
incominciate il 3 agosto 1988 e terminate il 22 marzo 1890. Relazione
scientifica, Ann. Uff. Cent. Meteor. Geodin., 10(4), 213 pp.
Mooser, F., H. Meyer-Abich, and A. R. McBirney (1956), Catalog of Active
Volcanoes of the World, part VI, Naples, 146 pp., Int. Assoc. of Volcanol.
and Chem. of the Earth’s Inter., Rome, Italy.
Nuccio, P. M., A. Paonita, and F. Sortino (1999), Geochemical modeling of
mixing between magmatic and hydrothermal gases: The case of Vulcano
Island, Italy, Earth Planet. Sci. Lett., 167, 321 – 333.
Nunziata, C., R. Mele, and M. Natale (1999), Shear wave velocities and
primary influencing factors of Campi Flegrei – Neapolitan deposits, Eng.
Geol., 54, 299 – 312.
Parkinson, K. J. (1981), An improved method for measuring soil respiration
in the field, J. Appl. Ecol., 18, 221 – 228.
Pasternack, G. B., and J. C. Varekamp (1997), Volcanic lake systematics:
I. Physical constraints, Bull. Volcanol., 58, 528 – 538.
Pruess, K. (1991), TOUGH2: A general purpose numerical simulator for
multiphase fluid and heat flow, Rep. LBL 29400, Lawrence Berkeley
Natl. Lab., Berkeley, Calif.
Pruess, K., C. Oldenburg, and G. Moridis (1999), TOUGH2: User’s guide,
version 2.0. Rep. LBNL-43134, Lawrence Berkeley Natl. Lab., Berkeley,
Calif.
Salazar, J. M. L., P. A. Hernandez, N. M. Perez, G. Melian, J. Alvarez,
F. Segura, and K. Notsu (2001), Diffuse emission of carbon dioxide from
Cerro Negro volcano, Nicaragua, Central America, Geophys. Res. Lett.,
28, 4275 – 4278.
Santacroce, R. (1987), Somma-Vesuvius, Quad. Ric. Sci., 114(8), Cons.
Naz. delle Ric., Rome, Italy.
Severne, C. M., and M. P. Hochstein (1994), Heat and mass transfer of the
Hipaua thermal area (Tokaanu-Waihi geothermal field) Lake Taupo, New
Zealand, Proceedings of the 16th New Zealand Geothermal Workshop,
edited by S. Soengkono and K. C. Lee, pp. 209 – 214, Univ. of Auckland,
Auckland, New Zealand.
Sheperd, J. B., and H. Sigurdsson (1978), The Soufriere Crater Lake as a
calorimeter, Nature, 271, 344 – 345.
Simkin, T., and L. Siebert (1994), Volcanoes of the World, 2nd ed., 349 pp.,
Geosci. Press, Tucson, Ariz.
Sorey, M. L., and E. M. Colvard (1994), Measurements of heat and mass
flow from thermal areas in Lassen volcanic national park. California,
1984 – 93, U.S. Geol. Surv. Water Resour. Invest. Rep., 94-4180-A.
Squarci, P., G. Giannelli, S. Grassi, M. Mussi, and F. D’Amore (1994),
Preliminary results of geothermal prospecting on the Island of Pantelleria,
Acta Vulcanol., 5, 117 – 123.
Stoiber, R. E., S. N. Williams, and B. J. Huebert (1986), Sulfur and halogen
gases at Masaya caldera complex, Nicaragua: Total flux and variations
with time, J. Geophys. Res., 91, 2215 – 2231.
Tedesco, D., G. Miele, Y. Sano, and J. P. Toutain (1995), Helium isotopic
ratio in Vulcano island fumaroles: Temporal variations in shallow level
mixing and deep magmatic supply, J. Volcanol. Geotherm. Res., 64,
117 – 128.
Todesco, M., G. Chiodini, and G. Macedonio (2003), Monitoring and mod-
elling hydrothermal fluid emission at La Solfatara (Phlegrean Fields,
Italy): An interdisciplinary approach to the study of diffuse degassing,
J. Volcanol. Geotherm. Res., 125, 57 – 79.
B08204 CHIODINI ET AL.: CARBON DIOXIDE DIFFUSE DEGASSING B08204

Tonani, F., and G. Miele (1991), Methods for measuring flow of carbon
dioxide through soils in the volcanic setting, paper presented at Interna-
tional Conference on Active Volcanoes and Risk Mitigation, Int. Assoc.
of Volcanol. and Chem. of the Earth’s Inter., Naples, Italy, 27 Aug. to
1 Sept.
Vezzoli, L. (1988), Island of Ischia, Quad. Ric. Sci., 114(10), Cons. Naz.
delle Ric., Rome, Italy.
Walker, J. A., and M. J. Carr (1986), Compositional variations caused by
phenocryst sorting at Cerro Negro volcano, Nicaragua, Geol. Soc. Am.
Bull., 97, 1156 – 1162.
Werner, C., and S. Brantley (2003), CO2 emissions from the Yellowstone
volcanic system, Geochem. Geophys. Geosyst., 4(7), 1061, doi:10.1029/
2002GC000473.
Williams, S. N. (1983), Plinian airfall deposits of basaltic composition,
Geology, 11, 211 – 214.
Wood, C. A. (1978), Morphometric evolution of composite volcanoes,
Geophys. Res. Lett., 5, 437 – 439.
Woodside, W., and J. Messmer (1961), Thermal conductivity of porous
media. I: Unconsolidated sand, J. Appl. Phys., 32, 1688 – 1699.























R. Avino, S. Caliro, G. Chiodini, A. Costa, and D. Granieri, Osservatorio
Vesuviano, Istituto Nazionale di Geofisica e Vulcanologia, via Diocleziano
328, 80124 Naples, Italy. (chiod@ov.ingv.it)
C. Werner, Institute of Geological and Nuclear Sciences, Private Bag
2000, Taupo, New Zealand.

17 of 17