arXiv subjects

D. Pierini

Publications and source records attributed to D. Pierini.

At least 55 records · Page 3Linked to original sources

arXiv2008

Diffuse stellar emission in X-ray luminous galaxy clusters at z~0.3 I. Is the diffuse optical light boosted and rejuvenated in merging clusters?

[Abridged] We searched for diffuse stellar emission around BCGs in three of the most X-ray luminous clusters found at z~0.3 in the REFLEX cluster survey and observed with XMM-Newton. These systems (RXCJ0014.3-3022, RXCJ0232.2-4420, and RXCJ2308.3-0211) are in different dynamical states, as witnessed by their X-ray morphology and optical appearence (e.g. multiplicity of BCGs). [Abridged] Diffuse stellar emission is robustly detected down to a surface brightness of 26 R-mag/arcsec^2 (observed frame) around a total of seven BCGs, extending up to galactocentric distances of ~100 kpc. In particular, it surrounds a pair of BCGs in RXCJ0232.2-4420, while it bridges two BCGs associated with the minor subcomponent of the merging cluster RXCJ0014.3-3022. The diffuse light detected at the greatest distances from the BCGs of the rather regular clusters RXCJ0232.2-4420 and RXCJ2308.3-0211 follows the ICM distribution. Its B-R colour is consistent with the colours measured within the BCG effective radii. The diffuse light around the two pairs of BCGs in RXCJ0014.3-3022 exhibits bluer colours than the BCG central regions by up to 0.5 mag. If the contribution of the intracluster light (ICL) to the detected diffuse light around BCGs is not negligible, ICL and BCGs have similar stellar populations in relatively relaxed clusters. Merging on a cluster scale eventually adds gravitational stresses to BCGs and other galaxies in subcluster cores. This event may affect the properties of the diffuse stellar emission around BCGs. Shredding of star-forming, low-metallicity dwarf galaxies is favoured as the cause of the bluer B-R colours of the diffuse stellar component around the two pairs of BCGs in the merging cluster RXCJ0014.3-3022.

astro-ph↗

arXiv2008

Unveiling the Important Role of Groups in the Evolution of Massive Galaxies: Insights from an Infrared Passive Sequence at Intermediate Redshift

The most massive galaxies in the Universe are also the oldest. To overturn this apparent contradiction with hierarchical growth models, we focus on the group scale haloes which host most of these galaxies. A stellar mass selected M_* >~ 2x10^10M_sol sample at z~0.4 is constructed within the CNOC2 redshift survey. A sensitive Mid InfraRed (MIR) IRAC colour is used to isolate passive galaxies. It produces a bimodal distribution, in which passive galaxies (highlighted by morphological early-types) define a tight MIR colour sequence (Infrared Passive Sequence, IPS). This is due to stellar atmospheric emission from old stellar populations. Significantly offset from the IPS are galaxies where reemission by dust boosts emission at 8microns (InfraRed-Excess or IRE galaxies). They include all known morphological late-types. Comparison with EW[OII] shows that MIR colour is highly sensitive to low levels of activity, and allows us to separate dusty-active from passive galaxies. The fraction of IRE galaxies, f(IRE) drops with M_*, such that f(IRE)=0.5 at a ``crossover mass'' of ~1.3x10^11M_sol. Within our optically-defined group sample there is a strong and consistent deficit in f(IRE) at all masses, and most clearly at M_* >~10^11M_sol. Using a mock galaxy catalogue derived from the Millenium Simulation we show that the observed trend of f(IRE) with M_* can be explained if suppression of star formation occurs primarily in the group environment, and particularly for M_*>~10^11M_sol galaxies. In this way, downsizing can be driven solely by structure growth in the Universe.

astro-ph↗

arXiv2007

Presence of dust with a UV bump in massive, star-forming galaxies at 1 < z < 2.5

Fundamental properties of the extinction curve, like the slope in the rest-frame ultraviolet (UV) and the presence/absence of a broad absorption excess centred at 2175 A (the UV bump), are investigated for a sample of 108 massive, star-forming galaxies at 1 < z < 2.5, selected from the FDF Spectroscopic Survey, the K20 survey, and the GDDS. These characteristics are constrained from a parametric description of the UV spectral energy distribution (SED) of a galaxy, as enforced by combined stellar population and radiative transfer models for different geometries, dust/stars configurations and dust properties. In at least one third of the sample, there is a robust evidence for extinction curves with at least a moderate UV bump. The presence of the carriers of the UV bump is more evident in galaxies with UV SEDs suffering from heavy reddening. We interpret these results as follows. The sample objects possess different mixtures of dust grains and molecules producing extinction curves in between the average ones of the Small and Large Magellanic Cloud, where the UV bump is absent or modest, respectively. Most of the dust embeds the UV-emitting stellar populations or is distributed out of the galaxy mid-plane. Alternatively, even dust with a pronounced UV bump, as for the average Milky-Way extinction curve, can be present and distributed in the galaxy mid-plane. In this case, variations of the continuum scattering albedo with wavelength or an age-dependent extinction are not sufficient to explain the previous trend with reddening. Hence, additional extraplanar dust has to be invoked. The data suggest that the carriers of the UV bump are associated with intermediate-age stellar populations, while they survive in the harshest UV-radiation fields owing to dust self-shielding. [abridged]

astro-ph↗

arXiv2007

The Representative XMM-Newton Cluster Structure Survey (REXCESS) of an X-ray Luminosity Selected Galaxy Cluster Sample

The largest uncertainty for cosmological studies using clusters of galaxies is introduced by our limited knowledge of the statistics of galaxy cluster structure, and of the scaling relations between observables and cluster mass. To improve on this situation we have started an XMM-Newton Large Programme for the in-depth study of a representative sample of 33 galaxy clusters, selected in the redshift range z=0.055 to 0.183 from the REFLEX Cluster Survey, having X-ray luminosities above 0.4 X 10^44 h_70^-2 erg s^-1 in the 0.1 - 2.4 keV band. This paper introduces the sample, compiles properties of the clusters, and provides detailed information on the sample selection function. We describe the selection of a nearby galaxy cluster sample that makes optimal use of the XMM-Newton field-of-view, and provides nearly homogeneous X-ray luminosity coverage for the full range from poor clusters to the most massive objects in the Universe. For the clusters in the sample, X-ray fluxes are derived and compared to the previously obtained fluxes from the ROSAT All-Sky Survey. We find that the fluxes and the flux errors have been reliably determined in the ROSAT All-Sky Survey analysis used for the REFLEX Survey. We use the sample selection function documented in detail in this paper to determine the X-ray luminosity function, and compare it with the luminosity function of the entire REFLEX sample. We also discuss morphological peculiarities of some of the sample members. The sample and some of the background data given in this introductory paper will be important for the application of these data in the detailed studies of cluster structure, to appear in forthcoming publications.

astro-ph↗

arXiv2005

Dust properties of UV bright galaxies at z ~ 2

We investigate the properties of the extinction curve in the rest-frame UV for a sample of 34 UV-luminous galaxies at 2 < z < 2.5, selected from the FORS Deep Field (FDF) spectroscopic survey. A new parametric description of the rest-frame UV spectral energy distribution is adopted; its sensitivity to properties of the stellar populations or of dust attenuation is established with the use of combined stellar population and radiative transfer models. The distribution of the z ~ 2 UV-luminous FDF galaxies in several diagnostic diagrams shows that their extinction curves range between those typical of the Small and Large Magellanic Clouds (SMC and LMC, respectively). For the majority of strongly reddened objects having a UV continuum slope beta > -0.4 a significant 2175 A absorption feature is inferred, indicating an LMC-like extinction curve. On the other hand, the UV continua of the least reddened objects are mostly consistent with SMC-like extinction curves, lacking a significant 2175 A bump, as for the Calzetti et al. sample of local starbursts. Furthermore, the most opaque (beta ~ 0) UV-luminous galaxies tend to be among the most metal rich, most massive, and largest systems at z ~ 2. The presence of the UV bump does not seem to depend on the total metallicity, as given by the equivalent width (EW) of the C IV doublet. Conversely, it seems to be associated with large EWs of prominent interstellar low-ionisation absorption lines, suggesting a link between the strength of the UV bump and the topology of the interstellar medium of the most evolved UV-luminous, massive galaxies at z ~ 2.

astro-ph↗

arXiv2005

The nature of the red disk-like galaxies at high redshift: dust attenuation and intrinsically red stellar populations

We investigate which conditions of dust attenuation and stellar populations allow models of dusty, continuously star-forming, bulge-less disk galaxies at 0.8 5.3, Ic-K>4, J-K>2.3). As a main novelty, we use stellar population models that include the thermally pulsating Asymptotic Giant Branch (TP-AGB) phase of stellar evolution. The star formation rate of the models declines exponentially as a function of time, the e-folding time being longer than 3 Gyr. In addition, we use calculations of radiative transfer of the stellar and scattered radiation through different dusty interstellar media in order to explore the wide parameter space of dust attenuation. We find that synthetic disks can exhibit red optical/near-infrared colours because of reddening by dust, but only if they have been forming stars for at least about 1 Gyr. Extremely few models barely exhibit Rc-K>5.3, if the inclination i=90 deg and if the opacity 2*tauV>6. Hence, Rc-K-selected galaxies at 1 0.5. This explains the large fraction of observed, edge-on disk-like galaxies with Ks<19.5 and F814W-Ks>4. Finally, models with 2 1 for i about 70 deg). In conclusion, red disk-like galaxies at 0.8<z<3.2 may not necessarily be dustier than nearby disk galaxies (with 0.5<2*tauV<2) and/or much older than about 1 Gyr. This result is due both to a realistic description of dust attenuation and to the emission contribution by TP-AGB stars... (Abridged)

astro-ph↗

arXiv2004

Dust Attenuation in Late-Type Galaxies. I. Effects on Bulge and Disk Components

We present results of new Monte Carlo calculations made with the DIRTY code of radiative transfer of stellar and scattered radiation for a dusty giant late-type galaxy like the Milky Way, which illustrate the effect of the attenuation of stellar light by internal dust on the integrated photometry of the individual bulge and disk components. Here we focus on the behavior of the attenuation function, the color excess, and the fraction of light scattered or directly transmitted towards the outside observer as a function of the total amount of dust and the inclination of the galaxy, and the structure of the dusty interstellar medium (ISM) of the disk. We confirm that dust attenuation produces qualitatively and quantitatively different effects on the integrated photometry of bulge and disk, whatever the wavelength. In addition, we find that the structure of the dusty ISM affects more sensitively the observed magnitudes than the observed colors of both bulge and disk. Finally, we show that the contribution of the scattered radiation to the total monochromatic light received by the outside observer is significant, particularly at UV wavelengths, even for a two-phase, clumpy, dusty ISM. Thus understanding dust scattering properties is fundamental for the interpretation of extragalactic observations in the rest-frame UV.

astro-ph↗

arXiv2003

Near-infrared observations of galaxies in Pisces-Perseus. V. On the origin of bulges

We investigate the scaling relations of bulge and disk structural parameters for a sample of 108 disk galaxies.Structural parameters are obtained from 2D bulge/disk decomposition of H-band images.Bulges are modelled with a generalized exponential (Sersic) with variable shape index n. We find that bulge effective scalelength and luminosity increase with increasing n, but disk properties are independent of bulge shape. As Hubble type T increases, bulges become less luminous and their mean effective surface brightness gets fainter; disks shows a similar, but much weaker, trend. When bulge parameters are compared with disk ones, they are tightly correlated for n=1 bulges. The correlations gradually worsen with increasing n such that n=4 bulges appear virtually independent of their disks. The Kormendy relation, surf. brightness vs. effect. radius, is shown to depend on bulge shape; the two parameters are tightly correlated in n=4 bulges, and increasingly less so as n decreases; disks are well correlated .Bulge-to-disk size ratios are independent of Hubble type, but smaller for exponential bulges. Strongly barred SB galaxies with exponential bulges are more luminous than their unbarred counterparts. Exponential bulges appear to be closely related to their underlying disks, while bulges with higher n values are less so. We interpret our results as being most consistent with a secular evolutionary scenario, in which dissipative processes in the disk are responsible for building up the bulges in most spirals.

astro-ph↗

arXiv2003

Extremely red galaxies: dust attenuation and classification

We re-address the classification criterion for extremely red galaxies (ERGs) of Pozzetti and Mannucci (2000 -- PM00), which aims to separate, in the Ic-K (or Rc-K) vs. J-K colour--colour diagram, passively evolving, old (> 1 Gyr) stellar populations in a dust-free environment, associated with ellipticals (Es), from dusty starburst galaxies (DSGs), both at 1 < z < 2. We explore a category of objects not considered previously, i.e., galaxies forming in this redshift range on short (0.1 Gyr) timescales and observed also in their early, dusty post-starburst phase. We also investigate the impact of structure of the dusty medium and dust amount on the observed optical/near-IR colours of high-z DSGs/DPSGs, through multiple-scattering radiative transfer calculations for a dust/stars configuration and an extinction function calibrated with nearby dusty starbursts. As a main result, we find that dusty post-starburst galaxies (DPSGs), with ages between 0.2 and 1 Gyr, at 1.3 < z < 2 mix with Es at 1 < z < 2 for a large range in dust amount. This ``intrusion'' is a source of concern for the present two-colour classification of ERGs. On the other hand, we confirm, in agreement with PM00, that DSGs are well separated from Es, both at 1 < z < 2, in the Ic-K vs. J-K colour--colour diagram, whatever the structure (two-phase clumpy or homogeneous) of their dusty medium and their dust amount are. This result holds under the new hypothesis of high-z Es being as dusty as nearby ones. Thus the interpretation of the optical/near-IR colours of high-z Es may suffer from a multiple degeneracy among age, metallicity, dust and redshift. We also find that DPSGs at z around 1 mix with DSGs at 1 < z < 2, as a function of dust amount and structure of the dusty medium. All these results help explaining the complexity of the ERG classification... (Abridged)

astro-ph↗

arXiv2003

Dust emission in the far-IR as a star formation tracer at z=0: systematic trends with luminosity

We investigate if dust emission in the far-IR continuum provides a robust estimate of star formation rate (SFR) for a nearby, normal late-type galaxy. We focus on the ratio of the 40--1000 micron luminosity (L_dust) to the far-UV (0.165 micron luminosity, which is connected to recent episodes of star formation. Available total photometry at 0.165, 60, 100 and 170 micron limits the statistics to 30 galaxies, which, however, span a large range in observed (and, thus, attenuated by dust) K-band (2.2 micron) luminosity, morphology and inclination (i). This sample shows that the ratio of L_dust to the observed far-UV luminosity depends not only on i, as expected, but also on morphology and, in a tighter way, on observed K-band luminosity. We find that L_dust/L_FUV is proportional to e^(-tau_K) (alpha+0.62) (L_K)^(0.62), where L_FUV and L_K are the unattenuated stellar luminosities in far-UV and K, respectively, and alpha is the ratio of the attenuation optical depths at 0.165 micron (tau_FUV) and 2.2 micron (tau_K). This relation is to zeroth order independent of i and morphology. It may be further expressed as L_dust/L_FUV proportional to (L_K)^delta, where delta = 0.61 - 0.02 alpha, under the observationally-motivated assumption that, for an average inclination, e^(-tau_K) is proportional to (L_K)^(-0.02). We adopt calculations of two different models of attenuation of stellar light by internal dust to derive solid-angle averaged values of alpha. We find that delta is positive and decreases towards 0 from the more luminous to the less luminous galaxies. This means that there is no universal ratio of far-IR luminosity to unattenuated far-UV luminosity for nearby, normal late-type galaxies. (Abridged)

astro-ph↗

arXiv2003

The far-IR/radio correlation in the ISO era

We present the correlation between the far-infrared (FIR) and radio emissions from a composite sample of 72 nearby normal galaxies observed with the ISOPHOT instrument on board the Infrared Space Observatory. The galaxies in the sample have measurements at three FIR wavelengths (60, 100 and 170 micron), which allowed a direct determination of the warm and cold FIR emission components. This is the first time that the correlation has been established for the total FIR luminosity, of which most is carried by the cold dust component predominantly emitting longwards of the spectral coverage of IRAS. The slope of this correlation is slightly non-linear (1.10+/-0.03). Separate correlations between the warm and cold FIR emission components and the radio emission have also been derived. The slope of the warm FIR/radio correlation was found to be linear (1.03+/-0.03). For the cold FIR/radio correlation we found a slightly non-linear (1.13+/-0.04) slope. We qualitatively interpret the correlations in terms of star formation rate and find that both the FIR and radio emissions may be consistent with a non-linear dependence on star formation rate for galaxies not undergoing starburst activity.

astro-ph↗

arXiv2002

[C II] emission and star formation in late-type galaxies. II A model

We study the relationship between gas cooling via the [C II] (158 micron) line emission and dust cooling via the far-IR continuum emission on the global scale of a galaxy in normal (i.e. non-AGN dominated and non-starburst) late-type systems. It is known that the luminosity ratio of total gas and dust cooling, L(C II)/L(FIR), shows a non-linear behaviour with the equivalent width of the Halpha line emission, the ratio decreasing in galaxies of lower massive star-formation activity. This result holds despite the fact that known individual Galactic and extragalactic sources of the [C II] line emission show different [C II] line-to-far-IR continuum emission ratios. This non-linear behaviour is reproduced by a simple quantitative model of gas and dust heating from different stellar populations, assuming that the photoelectric effect on dust, induced by far-UV photons, is the dominant mechanism of gas heating in the general diffuse interstellar medium of the galaxies under investigation. According to the model, the global L(C II)/L(FIR) provides a direct measure of the fractional amount of non-ionizing UV light in the interstellar radiation field and not of the efficiency of the photoelectric heating. The model also defines a method to constrain the stellar initial mass function from measurements of L(C II) and L(FIR). A sample of 20 Virgo cluster galaxies observed in the [C II] line with the LWS on board ISO is used to illustrate the model. The limited statistics and the necessary assumptions behind the determination of the global [C II] luminosities from the spatially limited data do not allow us to establish definitive conclusions but data-sets available in the future will allow tests of both the reliability of the assumptions of our model and the statistical significance of our results.

astro-ph↗

arXiv2002

1.65 micron (H-band) surface photometry of galaxies. IX: photometric and structural properties of galaxies

As a result of a systematic NIR H-band (1.65 micron) imaging survey of normal galaxies in the local universe, we are able to measure in a highly homogeneous way photometric and structural properties for a sample of 1143 galaxies. We base our analysis on a quantitative structural parameter, the concentration index C_31, instead of relying on the galaxies' morphological classification. The C_31 parameter provides a model independent, quantitative and continuous characterization of the light distribution within galaxies, and it is thus to be preferred to either the Hubble type or a parameter like the bulge-to-disk or bulge-to-total light ratio. We find that galaxies can be completely described by three parameters, namely: a scale parameter (the H-band luminosity), a shape parameter (the concentration index C_31) and a colour parameter (e.g. the B-H colour). At low luminosity dEs and dIs, having similar C_31, are colour-discriminated, while at very high luminosity different C_31 discriminate S0s from Scs, otherwise undistinguishable on the basis of their colour. A single, monotonic relation exists between luminosity and mu_e in the H-band, as opposed to the two separate regimes that are generally observed in the B-band. As NIR luminosity traces quite accurately the galaxy mass distribution, this relation re-enforces the indication in favour of a scale-dependent mass collapse mechanism which produces higher surface-brightness and more centrally peaked galaxies with increasing mass. However, the presence of high-luminosity low-C_31 galaxies hints at other machanisms and physical properties (such as angular momentum) playing an important role in galaxy formation. (abridged)

astro-ph↗

arXiv2002

[CII] at 158 mic as a star formation tracer in late-type galaxies

We present a calibration of the massive star formation rate vs. [CII] luminosity relation based on a sample of nearby, late-type galaxies observed with ISO-LWS and imaged in the Halpha+[NII] line. The relation holds for far-IR luminosities 10^8 < L_{FIR} < 10^{10.5} L (solar units). The derived star formation rates have an uncertainty of about a factor of 10. Part of this uncertainty is due to the different mix of contributions to the [CII] emission from the different components of the interstellar medium in individual galaxies, as discussed in an appendix.

astro-ph↗

arXiv2002

1.65 micron (H-band) surface photometry of galaxies. VIII: the near-IR k-space at z=0

We present the distribution of a statistical sample of nearby galaxies in the k-space (k1 ~ log M, k2 ~ log Ie, k3 ~ log M/L). Our study is based on near-IR (H-band: lambda = 1.65 micron) observations, for the first time comprising early- and late-type systems. Our data confirm that the mean effective dynamical mass-to-light ratio M/L of the E+S0+S0a galaxies increases with increasing effective dynamical mass M, as expected from the existence of the Fundamental Plane relation. Conversely, spiral and Im/BCD galaxies show a broad distribution in M/L with no detected trend of M/L with M, the former galaxies having M/L values about twice larger than the latter, on average. For all the late-type galaxies, the M/L increases with decreasing effective surface intensity Ie, consistent with the existence of the Tully--Fisher relation. These results are discussed on the basis of the assumptions behind the construction of the k-space and their limitations. Our study is complementary to a previous investigation in the optical (B-band: lambda = 0.44 micron) and allows us to study wavelength-dependences of the galaxy distribution in the k-space. As a first result, we find that the galaxy distribution in the k1--k2 plane reproduces the transition from bulge-less to bulge-dominated systems in galaxies of increasing dynamical mass. Conversely, it appears that the M/L of late-types is higher (lower) than that of early-types with the same M in the near-IR (optical). The origins of this behaviour are discussed in terms of dust attenuation and star formation history.

astro-ph↗

arXiv2001

Extended Red Emission in the "Evil Eye" Galaxy (NGC4826)

We obtained low-resolution, long-slit 5300--9100 A spectroscopy of NGC4826 (a nearby galaxy with an absorbing dust lane (DL) asymmetrically placed across its bulge, associated with several HII regions) with a slit encompassing its bulge, positioned across its nucleus. The wavelength-dependent effects of absorption and scattering by dust in the lane are evident when comparing the observed stellar SEDs of pairs of positions symmetrically located with respect to the nucleus, one on the DL side and one on the opposite side of the bulge, by assuming that the intrinsic ISRF is axi-symmetric. We analyzed these SED ratios through the multiple-scattering radiative transfer model of Witt and Gordon and we discovered strong residual Extended Red Emission (ERE) from a region of the DL within a distance of 13 arcsec from the nucleus, adjacent to a broad, bright HII region. ERE is an established phenomenon in the literature interpreted as originating from photoluminescence by nanometer-sized clusters, illuminated by UV/optical photons. The complex radial variation of the ERE band-integrated intensity and of the ERE-to-scattered light band-integrated intensity ratio with the optical depth of the model derived for the DL and with the strength and hardness of the illuminating ISRF is reproduced consistently through the theoretical interpretation of the photophysics of the ERE carrier by Smith and Witt. When examined within the context of ERE observations in a variety of Galactic dusty environments (e.g. the diffuse ISM, reflection nebulae, planetary nebulae and the Orion Nebula), we conclude that the ERE photon conversion efficiency in NGC4826 is as high as found elsewhere, but that the size of the actively luminescing nanoparticles there is about twice as large as those thought to exist in the Galactic diffuse ISM.

astro-ph↗

arXiv2001

New clues to the evolution of dwarf early-type galaxies

Surface photometry of 18 Virgo cluster dwarf elliptical (dE) and dwarf lenticular (dS0) galaxies, made by Gavazzi et al. (2001) in the H-band (1.65 micron) and in the B-band (0.44 micron), shows that the ratio of the effective radii of these stellar systems in the B- and H-band, r_{e B}/r_{e H}, ranges between 0.7 and 2.2. In particular, dwarf ellipticals and lenticulars with a red total color index B-H (i.e. with 3.2 < B-H < 4) have equal effective radii in these two pass-bands. By contrast, blue (i.e. with 2.5 < B-H < 3.1) dEs and dS0s have B-band effective radii about 50% longer than the H-band ones, on average. Consistently, strong negative gradients in B-H along the galactocentric radius are found to be associated with blue total colors. This trend is not found in a sample of 29 giant E and S0 galaxies of the Coma cluster with analogous data available in the literature. These early-type giants span a broad range in r_{e B}/r_{e H} (0.2--2.2), with a mean r_{e B}/r_{e H} of about 1.1, but a narrow range in (red) color (3.3 < B-H < 4.2). In these stellar systems, color gradients are usually interpreted as due either to age/metallicity gradients along the radial coordinate or to dust attenuation, whatever the total color of the system is. Assuming each of these three distinct interpretations of the origin of color gradients, we discuss the origin of the association of strong negative color gradients with blue colors found in the early-type dwarfs under study, in relation with current scenarios of formation and evolution of dE and dS0 galaxies.

astro-ph↗

arXiv2001

Gas cooling within the diffuse ISM of late-type galaxies

We combine observations of spiral galaxies in the [CII] line at 158 micron, made with the Long Wavelength Spectrometer aboard ISO, with previous data from the Kuiper Airborne Observatory to study the origin of this line, which is the main coolant of the interstellar medium at relatively low temperatures. We also use HI and CO(1-0) observations of these galaxies and estimate the respective line fluxes in the same beam as the [CII] observations. We confirm the existence of a linear relation between the [CII] line intensity and the CO(1-0) line intensity, that we extend to intrinsically fainter galaxies. The dispersion around this relation is significant and due to variations in the far-UV flux, thus in the star formation rate. We find that for the least active galaxies of our sample, in terms of star formation, the rate of [CII] line emission per interstellar hydrogen atom is similar to that in the Solar neighbourhood. For those galaxies, most of the [CII] line emission comes probably from the diffuse cold atomic medium. In more active galaxies, considered globally, the average [CII] line emission is dominated by dense photodissociation regions and to some extent by the warm ionized diffuse medium. This is true in the central region of many spiral galaxies, and probably even in the interarm regions of the most actively star-forming ones.

astro-ph↗