arXiv:0904.3257v1 [astro-ph.CO] 21 Apr 2009
Young stellar populations and star clusters in NGC 1705
1
F. Annibali2,3 , M. Tosi4 , M. Monelli5 , M. Sirianni3,6 , P. Montegriffo4 , A. Aloisi3,6 , L.
Greggio2
2
INAF - Osservatorio Astronomico di Padova, Vicolo dell’Osservatorio 5, I-35122 Padova, Italy
3
Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218
4
INAF - Osservatorio Astronomico di Bologna, Via Ranzani 1, I-40127 Bologna, Italy
5
Instituto de Astrofisica de Canarias, Calle Via Lactea, E38200 La Laguna, Espana
6
On Assignment from the Space Telescope Division of the European Space Agency
ABSTRACT
We present HST photometry of the late-type dwarf galaxy NGC 1705 observed with the
WFPC2 in the F380W and F439W bands and with the ACS/HRC in the F330W, F555W, and
F814W broad-band filters. We cross-correlate these data with previous ones acquired with the
WFPC2 in the F555W, F814W bands, and derive multiband Color-Magnitude diagrams (CMDs)
of the cross-identified individual stars and candidate star clusters. For the central regions of the
galaxy, where HST-NICMOS F110W and F160W photometry is also available, we present U, B,
V, I, J, H CMDs of the 256 objects with magnitudes measured in all bands.
While our previous study based on F555W, F814W, F110W and F160W data allowed us to
trace the star formation history of NGC 1705 back to a Hubble time, the new data provide a
better insight on its recent evolution. With the method of the synthetic CMDs, we confirm the
presence of two strong bursts of star formation (SF). The older of the two bursts (B1) occurred
between ∼ 10 and 15 Myr ago, coeval to the age of the central SSC. The younger burst (B2)
started ∼ 3 Myr ago, and it is still active. The stellar mass produced by B2 amounts to ∼ 106
M⊙ , and it is a factor of ∼ 3 lower for B1. The interburst phase was likely characterized by a
much lower level of star formation rather than by its complete cessation.
The two bursts show distinct spatial distributions: while B1 is centrally concentrated, B2 is
more diffused, and presents ring and arc-like structures that remind of an expanding shell. This
suggests a feedback mechanism, in which the expanding superbubble observed in NGC 1705,
likely generated by the (10–15) Myr burst, triggered the current strong star formation activity.
The excellent spatial resolution of the HRC allowed us to reliably identify 12 star clusters
(plus the SSC) in the central ∼ 26” × 29” region of NGC 1705, 10 of which have photometry in
all the UBVIJH bands. The comparison of the cluster photometry with the GALEV populations
synthesis models provides ages from ≈ 10 Myr to ≈ 1 Gyr, and masses between ≈ 104 and
105 M⊙ . The conspicuous cluster population in the central regions, with one super star cluster,
one populous cluster and several regular ones, confirm the strong star forming activity of NGC
1705.
Subject headings: galaxies: evolution — galaxies: individual: NGC 1705 — galaxies: irregular — galaxies:
dwarf — galaxies: stellar content
1.
1 Based
on observations with the NASA/ESA Hubble
Space Telescope, obtained at the Space Telescope Science
Institute, which is operated by AURA for NASA under
contract NAS5-26555
Introduction
NGC 1705 is a blue compact dwarf (BCD)
galaxy, with two characteristics which make it
extremely interesting: a) it shows the best doc1
umented proof of an ongoing galactic outflow
(Meurer et al. 1992, hereafter MFDC; Heckman et
al. 2001), and b) its nucleus hosts a luminous super star cluster (SSC) with estimated mass ∼ 105
M⊙ , probably a proto-globular cluster only 10
Myr old (Melnick et al. 1985; O’Connell et al.
1994; Ho & Filippenko 1996). NGC 1705 is almost unaffected by intervening obscuring clouds
and is not excessively crowded, in spite of a distance of 5.1 Mpc, as determined from the Tip of
the Red Giant Branch (Tosi et al 2001, hereafter
T01). This allows HST to resolve its individual
stars down to fairly faint magnitudes, and makes
it an ideal benchmark to study the evolution of
BCDs, an important class of galaxies absent in
the Local Group.
NGC 1705 (α2000 = 04 54 15.2, δ2000 =
−53 21 40, l = 261.08 and b = –38.74) has been
classified by MFDC as a BCD with a fairly continuous SF regime and an approximate oxygen abundance of 12+log(O/H) ≃ 8.46. An oxygen abundance of 8.36 was derived by Storchi-Bergmann,
Calzetti & Kinney (1994) from UV, optical and
near infrared spectra, while more recently Lee &
Skillman (2004) have inferred more precise abundances from 16 HII regions, 5 of which had the
electron temperature directly estimated from the
[OIII]λ 4363 line. The resulting oxygen abundance is 8.21 ± 0.05, as in the Small Magellanic
Cloud. Lee & Skillman also measured the nitrogen abundance and found a (N/O)HII =-1.75 ±
0.06, in agreement with the (N/O)HI =-1.70 ± 0.4
found by Heckman et al. (2001) and Aloisi et al.
(2005) from FUSE observations of the neutral gas.
This N/O abundance ratio is among the lowest
values derived in late-type dwarfs and is difficult
to explain in terms of galactic chemical evolution,
unless strong galactic winds with enhanced nitrogen loss are assumed (Romano, Tosi & Matteucci
2006).
UV spectra acquired with HST (Heckman &
Leitherer 1997) showed that nearly half of the optical/ultraviolet light is contributed by the young
stellar population, and possibly by the central SSC
which may also power the observed bipolar outflow. From the lack of spectral stellar wind features, Heckman & Leitherer suggested that the
stars in the luminous SSC of NGC 1705 are not
more massive than 10–30 M⊙ . Ground-based
work (e.g. Quillen et al. 1995) revealed the pres-
ence of a composite stellar population, with the
older (∼ 1–10 Gyr) field population defining the
galaxy morphology. Our HST data (T01) confirmed this result showing that the stellar population in the inner central region is dominated by
stars 10-15 Myr old (i.e. roughly coeval to the
SSC), which are not present in intermediate and
outer regions. Intermediate-age stars populate all
the galaxy within about 700–800 pc from the SSC,
while stars up to several Gyrs old are detected
wherever crowding is not too severe (i.e. outwards
of 200–300 pc from the SSC) and become increasingly dominating towards the outer regions.
In Annibali et al. (2003, hereafter A03) we
applied the synthetic CMD method to the T01
CMDs to infer the star formation history (SFH)
of 8 concentric regions of NGC 1705 and found
that the data are consistent with a rather continuous Star Formation (SF) activity, started at
least 5 Gyr ago, and recently overcome by a strong,
centrally concentrated burst with an age between
15 and 10 Myr ago. This recent SF episode appears then coeval to the birth of the SSC and to
the onset of the galactic wind. No (or very little) SF seems to have occurred in NGC 1705 between 10 and 3 Myr ago, possibly as a consequence
of gas sweeping/heating by the strong wind, but
then a new, even stronger burst appears to have
started everywhere in the galaxy about 3 Myr ago.
The latter result suggests that the gas must have
been able to cool and fall down on a very short
timescale.
To better study the young stellar populations in
NGC 1705, we have observed the galaxy at shorter
wavelengths, in the F380W and F439W filters, and
derived the CMDs of the resolved stars. In this
respect, this work is complementary to that presented by T01, where the F555W, F814W, F110W
and F160W images allowed us to map more accurately the intermediate age and old populations.
Observations of the central region were also performed in F330W, F555W and F814W with the
ACS/HRC. These new data allow us to check
the SFH inferred from the previous data by A03
and to study the properties of the star clusters in
NGC1705.
The data are described in Sect.2, the properties of the candidate star clusters are presented in
Sect.3, while the CMDs of the resolved individual
stars are presented and analysed in Sect.4. The
2
SFH of the resolved stars is discussed in Sect.5,
and the overall results are discussed and summarized in Sect.6.
2.
2.1.
StarFinder, the code developed for astrometry and
photometry in crowded stellar fields described in
detail by Diolaiti et al. (2000). With this code
a numerical PSF template can be either directly
extracted by modelling the stars observed in the
frame or it can be simulated with Tiny Tim (Krist
& Hook 1999). In this case we have inferred the
PSF from the images by computing the median
average of 10 suitable stars after correction for the
local background and surrounding sources.
A provisional list of candidate single stars is
created including all brightness peaks at 5σ above
the background, and is then correlated with the
template PSF in order of decreasing flux. We considered a candidate as a real star only if the correlation coefficient (i.e. a measure of the similarity with the PSF) was greater than 0.7. From the
analysis of many simulations, this value was in fact
derived as the best representation of the boundary
between real stars and spurious detections.
During the PSF fitting, the code progressively
creates a virtual copy of the observed field as a
smooth background emission with superimposed
stellar sources. The stars are modeled as weighted
shifted replicas of the PSF template and added
one by one, in order of decreasing intensity. Photometry is performed on each individual star by
using a sub-image of size comparable to the diameter of the first diffraction ring of the PSF. The
local background is approximated by a bilinear
surface, the underlying halos of brighter stars outside the fitting region are given by the synthetic
image, and the brighter stars inside this region
are represented as weighted shifted replicas of the
PSF and re-fitted together with the analyzed star.
The fainter candidate stars inside the fitting region are instead neglected at this point and are
analyzed with the same strategy only in a further
step. If the photometric fit is acceptable, the catalog and the synthetic image are updated with the
new entry. For a better astrometric and photometric accuracy, all known sources are fitted again
after examining and fitting all the candidate objects. The stars are then subtracted to search for
possible lost objects, which are examined and fitted with the same procedure in the original frame.
This step was iterated three times for each filter.
All the sources with extended profile or detected in the immediate neighborhood of the SSC
were listed separately. The instrumental magni-
Observations and Data Reduction
WFPC2 data
The F380W and F439W observations of NGC
1705 were part of a WFPC2 four-band program
(GO-7506). The F555W and F814W images were
successfully obtained in March 1999 (and have already been presented by T01), while the F380W
and F439W ones could be successfully acquired
in 2000 November 10–11. Contrary to the original
planning, technical failures has prevented us to obtain the F380W and F439W images with the same
telescope orientation as the F555W and F814W
ones. The two orientations turned out to differ
by 111 degrees from each other, implying that the
WF fields overlap only partially, as shown in Fig.1.
The pointing was organized to follow a dither
pattern with CR-split of 0.5. For the F439W images we have 4 different pointing positions, while
for the F380W we have only two different positions. Two exposures at each dither position were
also requested for an easier cosmic-ray removal.
We thus have 8 images of 2,900 s each and one of
140 s in F439W, and 4 images of 500 s, 2 of 600
s and 1 of 140 s in F380W. The adoption of the
dithering and CR-split techniques allows us to improve the background estimate, identify hot pixels
and smooth local pixel to pixel variations from images taken at different dither points, and improve
the image sampling.
For each filter the dithered frames, calibrated
through the standard STScI pipeline procedure,
were combined in a single, fully sampled image
(with total exposure time 23,200 s in F439W and
3,200 s in F380W), using the software package
Drizzle (Fruchter & Hook 1998). The effective
pixel size in the resampled images corresponds to
0.′′ 023 and 0.′′ 05 for the PC and WFCs respectively.
The PSF in the resulting drizzled images has a
FWHM of ≈ 3 pixels in F380W and F439W for
both the PC and the WFs. Fig.2 displays the final images, with superimposed both the WFPC2
fields of the previous F555W and F814W data and
the isophotal line contours used in T01 to define
the 8 concentric regions.
The drizzled frames have been analyzed with
3
tude mi of each object in each filter and detector
was estimated via the PSF-fitting technique and
then calibrated into the HST-VEGAMAG system following the standard procedure described by
Holtzman et al. (1995a, 1995b) and the updated
coefficients provided by Biretta et al. (2000):
m = mi + Cap + C∞ + ZPV + CCT E
where, the aperture correction C∞ is an offset of
−0.10 (irrespective of filter and detector) to convert the magnitude from the 0.′′ 5 radius into a
nominal infinite aperture; ZPV are the zero points
taken from Baggett et al. (1997); CCT E is the correction for the charge transfer efficiency given by
Whitmore, Heyer, & Casertano (1999). Cap is the
correction to convert the photometry performed
in the adopted aperture to that in the conventional 0.′′ 5 aperture. StarFinder actually provides
the stellar flux already in an infinite aperture, but
for sake of homogeneity with the standard calibration prescriptions, we have preferred to translate its measured fluxes into the classical 0.′′ 5 aperture. In this case the correcting coefficients Cap
were derived directly from the same PSF templates adopted for the photometric reduction and
turned out to be negligible. The F380W and the
F439W catalogs, derived independently with the
above procedure, were then cross-correlated and
only the objects with a spatial offset smaller than
1 pixel were retained. This led to 2083 stars detected in both filters.
The central region of NGC 1705 was also observed in F110W and F160W with the HST-NIC2
camera (see T01 for details). There are 256 objects cross-identified in all the bands from the ultraviolet through the near-infrared, 10 of which
candidate star clusters (see Sect. 2.3). The CMDs
of these 256 objects are shown in Fig.3, where individual stars are represented by dots and candidate clusters by filled circles. The field of view of
the NIC2 camera covers region 7, most of region 6
and a small fraction of region 5 (see Fig.2 in T01).
Naturally, only the brightest stars are visible from
the F380W through the F160W bands. By comparing the CMDs of Fig.3 with the stellar models,
we see that the 256 stars are mostly on the MS,
red super-giant phases, and a few are on the AGB.
Hereinafter, we will name U, B, V, I, J and
H the mF 380W , mF 439W , mF 555W , mF 814W ,
mF 110W , and mF 160W magnitudes calibrated in
the HST-Vegamag system.
2.2.
Errors and incompleteness
To evaluate the degree of incompleteness and
blending of our photometry at each mag level, and
to derive a reliable estimate of the photometric errors, we have performed a series of artificial star
experiments, following the procedure created by
P.M. at the Bologna Observatory and described by
T01. In summary, we recursively added artificial
stars in random locations of the frames, by dividing the fields of view in grids of non-overlapping
cells and positioning at most one star per cell in
each run. Then, we repeated exactly the same
procedure of PSF fitting and calibration applied
to the actual data. The artificial stars were added
with a magnitude distribution similar to the observed luminosity function (LF), but with more
objects at the faint end to take into account that
the empirical LF at the fainter mags is more likely
to be affected by incompleteness. We added a
few objects at a time, in order not to alter the
field crowding conditions, repeating the process as
many times as needed to total about 200,000 artificial stars. The same threshold and correlation
selection criteria applied to the real stars, and described above, were also applied to to the output
catalogue of the artificial stars.
The completeness of our photometry at each
magnitude level was computed as the ratio of the
number of recovered artificial stars over the number of added ones (considering as recovered objects only those found within 0.5 pix of the given
coordinates, satisfying the adopted selection criteria and with magnitudes differing from the input
ones less than ± 0.75 mag). The resulting completeness levels in the B band are listed in Table 1
for the 8 galactic regions.
The difference between input and output magnitudes of the artificial stars represents a robust
estimate of the actual photometric error to be associated to each magnitude bin. These errors are
usually larger than those provided by the data reduction packages, especially towards the fainter
magnitudes. They have the further advantage of
providing a good characterization of the effects of
blending affecting crowded field photometries. In
Fig. 4 we show the results of our procedure for the
F439W filter for each of the 8 concentric regions
(regions with the same behavior are grouped together). The error distribution is skewed towards
4
rived from the residuals of the fit to the selected
stars. The additive corrections include only firstorder derivatives of the PSF with respect to the
X and Y positions in the image. This procedure
allows us to follow the first order spatial variation
of the PSF in the ACS/HRC field of view.
A source catalog was created from the detections above 4 σ in the sum of the U, V and I
images. Aperture photometry with PHOT, and
then PSF-fitting photometry with the ALLSTAR
package, were performed at the position of the detected sources in each band. The U,V and I catalogs were then cross-correlated, leading to ≈ 3000
objects with a measured magnitude in all the three
bands simultaneously. To search for compact clusters, we selected objects with sizes larger than the
PSF (sharpness >0.5), and magnitudes brighter
than mF814W . 22 in the final photometric catalog. By visually inspecting the selected objects in
all the images, we recognized several background
galaxies. We ended up with 12 candidate clusters
in our images. The location in the HRC image
of the selected candidate clusters is indicated in
Figure 5.
Intrinsic sizes of the candidate clusters were derived using the ISHAPE task in BAOLAB (Larsen
1999). ISHAPE models a source as an analytical
function convolved with the PSF of the image. For
each object, ISHAPE starts from an initial value
for the FWHM, ellipticity, and orientation. These
parameters are then iteratively adjusted until the
best fit between the observed profile and the model
convolved with the PSF is obtained. We adopted
a fitting radius of ≈ 4 × FWHM (PSF) in the
ISHAPE procedure, corresponding to 0.3” in the
I image. We also tested different input analytical
functions. A King (1962) profile with a concentration parameter c = 30 provides the best fit to the
data. The ISHAPE output includes the intrinsic
shape parameters which provide the best match to
the observed source profile, and a residual image.
The intrinsic effective radius, averaged over the
three bands, is listed for each cluster in Table 2.
The derived intrinsic effective radius distribution
is presented in Fig. 6.
We derived the magnitudes of the selected candidate clusters in U, B, V, I, J and H by performing aperture photometry with PHOT on the
ACS/HRC, WFPC2 and NIC2 data. We first produced for each filter a subtracted image where we
positive values of the (input–output) mags of the
recovered artificial stars: this is the signature of
a significant blending at the fainter mags, since a
star is recovered brighter when it overlaps another
star in the image.
2.3.
ACS/HRC data and the cluster sample
From the shape of the extended objects in the
WFPC2 F555W and F814W images of NGC 1705,
T01 suggested that 30 of them are likely candidate
star clusters (17 plus the SSC in the PC field, 4
in the WF2, 5 in the WF3, and 3 in the WF4),
while 42 are probably background galaxies. To
obtain a safer cluster selection in the central region of NGC 1705, we used the High Resolution
Channel (HRC) of the Advanced Camera for Surveys (ACS), whose excellent spatial resolution allows both a reliable identification of the star cluster population, and a study of its morphological
properties.
The HRC/ACS observations were performed
in August 2003 in the F330W (U), F555W (V),
and F814W (I) broad-band filters (program GTO9989). The exposures were taken with a 3-point
dithering pattern in each filter. The total expousure times are 680 s in U, and 420 s in both V
and I.
For each filter, we coadded with MULTIDRIZZLE (Koekemoer et al. 2002) the three dithered
frames, calibrated through the most up-to-date
version of the ACS calibration pipeline (CALACS),
into a single resampled image with a 0.02” pixel
size, i.e. ≈ 0.7 times the original HRC pixel size.
The field of view of each image is ≈ 26 × 29
arcsec2 . The MULTIDRIZZLE procedure also
corrects the ACS images for geometric distortion
and provides removal of cosmic rays and bad pixels.
PSF-fitting photometry with the DAOPHOT
package (Stetson 1987) in the IRAF2 environment
was performed on the final drizzled images.
To derive the PSF, we selected ∼ 30 stars in
each frame. The PSF was modeled with an analytic Moffat function plus additive corrections de2 IRAF
is distributed by the National Optical Astronomy
Observatories, which are operated by AURA, Inc., under
cooperative agreement with the National Science Foundation
5
removed all the stars around the candidate star
clusters. Then we performed aperture photometry
on the subtracted images at the cluster positions
inside a radius Re , given by the convolution of the
cluster intrinsic effective radius re with the PSF.
From the definition of effective radius, the total
magnitude is mT OT = m(r < Re ) − 0.75.
To test for systematics, we compared the photometry for all the objects in common between the
HRC and the WFPC2 catalogues. The comparison was performed only in the F555W and F814W
filters, because the F330W filter of the HRC data
is significantly different from the F380W filter of
the WFPC2 data. The cross-correlation provides
≈ 3000 stars common to the HRC and WFPC2
datasets. For a direct comparison, the HRC magnitudes were converted into the WFPC2 Vegamag system applying the transformations provided in Table 25 of Sirianni et al (2005). The
agreement between the HRC and WFPC2 photometries in the F555W filter is very good, with
a ∆(magHRC − magW F P C2 ) distribution peaked
around ∼ 0. The comparison in the F814W filter is instead less satisfactory, with a small offset
∆(magHRC −magW F P C2 ) ∼ 0.06. We checked for
possible problems in the data reduction process,
but were unable to identify the source of this slight
discrepancy. A possible cause for the observed systematics could be the use of two different photometric packages (Starfinder and Daophot) for the
reduction the WFPC2 and HRC datasets. In fact,
it has been shown that different photmetric packages used on the same image can yield magnitude
offsets up to ∼ 0.05 − 0.1 mag, especially when
dealing with fields with severe crowding and variable background due to unresolved sources (e.g.
Hill et al. (1998), Holtzman et al. (2006)).
We provide in Table 2 the cluster photometry
inside Re in the different filters: F330W (HRC),
F380W, F439W, F555W, F814W (WFPC2), and
F110W, F160W (NIC2). In this paper, we will use
the WFPC2 and NIC2 photometry to derive the
cluster ages and masses, for consistency with the
star formation history derived from the WFPC2
and NIC2 photometry of the resolved stars.
3.
the central region of NGC 1705, in addition to
the SSC. 10 of them have photometry in all the
U,B,V,I,J and H bands.
The color-color diagrams (U−V vs V−I, and
U-V vs J-H) are presented in Fig. 7. The data
are compared with the GALEV models (Anders
& Fritze-v. Alvensleben 2003) for different metallicites, and ages from 4 Myr to 10 Gyr. The models are based on the Padova isochrones (Bertelli et
al. 1994). The adopted IMF has a Salpeter slope,
between 0.1 M⊙ and 120 M⊙ . Fig. 7 shows that
SSP models of different metallicities are strongly
degenerate. For this reason we derive the cluster
ages assuming a metallicity of Z = 0.004, which
is consistent with the abundances of NGC 1705
HII regions and of the resolved stars of young and
intermediate age (T01, A03). The relative colorcolor diagram is presented in Fig. 8, where we also
show the effect of reddening on the Z = 0.004
models.
The cluster ages were derived by minimizing
the difference between models and data colors in
all possible combinations (15 colors), according to
a χ2 criterion, varying the age (between 4 Myr
and 10 Gyr), and the reddening (between E(BV)=0 and 1) of the models. Cluster masses were
then estimated from the mass-to-light ratios of the
best fitting models in the I band.
The results of the analysis are provided in Table 2, where we list the best-fitting ages, masses
and E(B − V ) values for the 12 candidate star
clusters. Clusters #6 and #8 are not well constrained by the fit, which allows for solutions in
the range 8−90 Myr, depending on the reddening. If we limit the fit to the bluer bands (U ,B
and V ), the best fit yields ages of the order of
100 Myr, in combination with low reddening, for
clusters #6 and #8. From Table 2 we notice that
these clusters are the least massive of our sample,
with masses of the order of 104 M⊙ for the old-age
solution, and masses as small as ∼ 3 × 103 M⊙ for
the young age solution. Thus their red light is
particularly prone to fluctuations on their evolved
star populations, which could explain why the fit
is poorly constrained.
The derived cluster age and mass distributions
are shown in Fig. 9. The distribution was obtained by adopting for clusters #6 and #8 the
oldest solutions of the full fit (≈ 80 and 90 Myr,
respectively). The 12 clusters span an age range
Star clusters properties
From the analysis of the HRC/ACS data, we
end up with 12 bona fide candidate clusters in
6
between ≈ 10 Myr and 1.4 Gyr, and masses between ≈ 104 and 105 M⊙ . The age distribution
appears peaked around ≈ 100 Myr, and there are
few clusters ≈ 1 Gyr old. Cluster 1 is as young as
≈ 16 Myr. It is very compact (re ≈ 1 pc), has a
mass of ≈ 6 × 104 M⊙ , and it is located at ≈ 20
pc from the central super star cluster. The two
most massive clusters (cluster #12, ≈ 7 × 104M⊙ ,
and cluster #9, ≈ 105 M⊙ ) in our sample are ≈
1.2 Gyr old.
Extragalactic young stellar clusters can be divided into three categories (e.g. Larsen & Richtler
2000, Billet, Hunter & Elmegreen 2002 and references therein), according to their luminosity: super star clusters [MV <
∼ − 10.5], populous clusters
[MV ≤ −8.5 for (B–V)0 ≥ −0.4 and MV ≤ −9.5
for (B–V)0 < −0.4], and regular ones. From
Table 2 one can see that the central region of
NGC 1705 contains one SSC, one populous cluster (cluster #1) and several regular ones. Notice
however that cluster #9 is actually more massive
than cluster #1, although not formally consistent
with the definition of populous cluster.
Billett, Hunter & Elmegreen (2002) have identified 15 clusters (plus the central SSC) in HSTWFPC2 images of NGC 1705 and discussed their
properties. This galaxy stands out of their sample
of 22 nearby late-type dwarfs for the highest cluster density, exceeded only by NGC 1569 (Hunter
et al 2000): the cluster population of NGC 1569
versus NGC 1705 count respectively 45 viz 15 normal clusters, 3 viz 1 SSCs, and 10 vs 1 populous clusters. Billett et al. (2002) also noticed
that, while in other galaxies the clusters are relatively separated, in NGC 1705 the majority of
them are found surrounding the central SSC, in
agreement with our results (see Figs.5 and 14).
A similar distribution has been found for the clusters in NGC 1569, from optical data (Hunter et al.
2000), and from HST J and H images (Origlia et
al. 2001). The bottom panel of our Fig. 14 shows
that the clusters trace well the spatial distribution of the intermediate-age (50–1000 Myr) stars
in NGC 1705. NGC 1705 and NGC 1569 are therefore similar exceptions among late-type dwarfs,
and their cluster properties probably reflect their
intense recent SF activity: NGC 1569 has a recent
SF rate (SFR) higher than in any other late-type
dwarf with resolved stellar populations studied so
far (Greggio et al. 1998, Angeretti et al. 2005);
second ranks NGC 1705, with a recent SFR a factor of a few lower (e.g. A03, Tosi 2007). This
confirms Larsen & Richtler (2000) suggestion of a
direct correlation between the number and luminosity of young clusters and the SF activity of the
hosting galaxies.
In the HRC field of view, we identify 7 clusters
in common with the Billett et al. (2002) sample
(our clusters # 3, 6, 7, 8, 10, 11, 12, corresponding respectively to their clusters # 12, 14, 13, 11,
10, 9, 7) . Billett et al. (2002) determined the
age for each cluster by comparing the integrated
cluster photometry in B, V and I with the models of Leitherer et al (1999). For our clusters #11
and #12, they derive an age of 1 Gyr, consistent
with our results. For clusters #6 and #8 they derive ages of 7 Myr, which are marginally consistent
with our solutions in the range (8 − 90) Myr. For
clusters #3, #7 and #10, instead, we derive ages
+9
+9
of 97+8
−49 , 101−60 , 80−17 Myr , which are fairly
older than their ages of 7, 15, and 7 Myr, respectively. This discrepancy does not arise from the
different photometry, not from the different procedure to account for the reddening (Billet et al.
(2002) assume an intrinsic reddening of 0.05 plus
a foreground reddening of ∼ 0.045, which is close
to the values that we derive for the three clusters).
Likely, the discrepancy results from the use of different SSP models, based on different stellar tracks
and on a different adopted metallicity (Billet et al
(2002) assume Z=0.008).
4.
The Colour-Magnitude Diagrams of the
resolved stars
With the resolved stars observed with WFPC2
and measured with the StarFinder procedure described in the previous sections, we have constructed the CMDs of each of the 8 concentric regions displayed in Fig.2. The U vs U-B, B vs B-V
and I vs V-I CMDs are shown in Fig. 10, where
regions 3-4 and regions 0-1-2 have been grouped
together for simplicity, since they present similar
CMDs. The number of objects in the CMDs is
indicated in each panel. These numbers do not
represent the actual stellar density in each region,
since they also depend on: i) camera spatial resolution (higher on the PC which covers the central regions), ii) completeness of the photometry
(higher for increasing galactocentric distance), iii)
7
area of the region (larger for increasing distance),
and iv) fraction of region covered by the two observing runs (smaller for increasing galactocentric
distance). As shown in Fig.2, the overlap of U ,B,V
and I frames is complete only for regions 6 and 7,
while, moving outwards, an increasing fraction of
the other regions lacks observations from one of
the two runs. Over the whole galaxy, only 4886
stars of the 40810 with F555W and F814W pho> 28), and
tometry, have also F439W (down to V ∼
2083 stars have both F380W and F439W (down
to U <
∼ 26).
For an immediate interpretation of the CMDs
in terms of stellar populations, we have superimposed in the central and right panels of Fig.10
the Padova stellar evolution tracks with metallicity Z=0.004 (Fagotto et al. 1994b), adopting from
T01 a distance of 5.1 Mpc, i.e. (m-M)0 =28.54,
and a reddening E(B–V)=0.045. It is apparent
that our CMDs are populated with stars in all the
evolutionary phases: main-sequence (MS), red supergiants, red giant branch (RGB) and asymptotic
giant branch (AGB) stars.
Star with masses lower than 2M⊙ are sampled
on the RGB, which is wide enough to be consistent
with an age range extending up to a Hubble time.
Unfortunately, the well known age-metallicity degeneracy of the RGB prevents us to firmly establish whether the red edge of the RGB corresponds
to metal poor stars as old as the Universe, or to
more metal rich ones ”only” a few Gyrs old. The
RGB component is only visible on the (V, I) CMD,
due to incompleteness at the shorter wavelengths,
as one can realize from the plotted tracks. Indeed, the U and B images are too shallow to reveal the faint, old stars. This is the reason why
the (U, B) and (B, V ) CMDs of the outer regions
cointain so little stars: these bands virtually show
only the gradient of the young population. In this
respect, the U, U-B and B, B-V diagrams do confirm the existence of a young stellar population
diffused over the whole galaxy, although more centrally concentrated, as usually found in late-type
dwarfs.
Common to all color combinations, we notice
that the depth of the CMDs is lowest in the central region: this is due to crowding. The magnitude difference between the faintest stars detected
in region 7 and in region 0-1-2 amounts to ∼ 3, 2
and 1 mags in the I,B and U bands respectively.
Actually, the resolved bluest stars are all bright,
(and young) and not strongly affected by crowding.
Taking all these effects into account, we can
conclude that the old population can be homogeneously distributed all over the galaxy, although
more easily measured in the more external, less
crowded, regions, while the youngest population
is more centrally concentrated, although present
everywhere up to the galaxy outskirts.
T01 found that the I, V-I CMDs of regions 7, 6
and 5 present an excess of stars along the 15M⊙
evolutionary track, i.e. with age <
∼ 15 Myr. This
peculiarity is confirmed in all bands, as shown in
Fig.3 where the CMDs of the 256 objects measured in all the 6 UBVIJH bands are plotted together with the 15M⊙ track by Fagotto (1994b).
In Fig.10 we plot this track as a thick line: the
excess of stars is apparent also in these CMDs,
and confined to the central regions. This feature
includes 20-25 stars (i.e. 10% of the plotted population) to the right of the blue plume, sometimes
showing up as a horizontal stripe at the top of the
stellar distribution (in the redder I and H CMDs),
sometimes as a sequence increasingly faint towards
redder colors (in the V and U CMDs). This ”windcone” is well confined in all the CMDs by the evolutionary tracks of stars with 20M⊙ and 15M⊙ ,
and corresponds to objects 10–15 Myr old, born
in the strong central burst identified by T01 and
A03, and coeval to the SSC. The unambiguous evidence of the presence and confinement of the 25
stars in the ”wind-cone” in all the CMDs of Fig.3
confirms that a very strong SF burst has started
15 Myr ago and has lasted only a few Myrs.
Fig. 3 also shows stars bluer than the ”windcone”, suggesting that younger stars exist. Many
of these objects are fairly well fitted by the dottedline in Fig.3, showing the evolutionary track of a
60M⊙ . Since the lifetime of a 60M⊙ is ∼4 Myr,
the bluest stars of Fig.3 must be of that age or
younger. This component (e.g. at B-V≤-0.3) is
present in the CMDs in Fig. 10 as well, where we
notice that it shows up all over the galaxy, out to
Region 0-1-2.
5.
Star formation history
The SFH of the resolved field population in
NGC 1705 was derived for the 8 regions of increas-
8
ing galactocentric distance by A03, applying the
method of synthetic CMDs described by Tosi et al.
(1991) and Greggio et al. (1998). For each region,
the SFH was modeled, looking for the best agreement between observed and simulated CMDs in
the V and I bands, and between the corresponding
luminosity functions. For an a posteriori consistency check, A03 also compared the near infrared
(J, H) CMDs with synthetic ones obtained assuming the SFH derived from the V and I data. They
showed that the agreement in the near infrared is
quite good as well.
Here, we want to check whether the SFH derived by A03 from the (V, I) data is consistent also
with the new data. To this purpose, we do not try
to model ex novo the SFH to reproduce the (B,
V) CMDs and LFs, but assume the SFH already
derived by A03, and check how it compares with
the observed B, B–V CMDs.
We briefly recall here that the synthetic CMDs
are produced via Monte Carlo extractions of
(mass, age) pairs for an assumed Initial Mass
Function (IMF), SF law, and initial and final
epochs of the SF activity. Each star is placed
in the theoretical log(L/L⊙ ), log Tef f plane via
interpolation on the Padova evolutionary tracks
(Fagotto et al. 1994a and b). Luminosity and
effective temperature are transformed into the
HST-Vegamag photometric system addopting the
Origlia & Leitherer (2000) conversion tables. Absolute magnitudes are converted into apparent
ones by applying reddening and distance modulus,
in this case E(B–V)=0.045 and (m-M)0 =28.54.
Then a completeness test is applied in order to
determine whether retaining or rejecting the synthetic star, based on the results of the artificial
star tests on the actual photometric data. Since we
performed the photometry independently in the B
and V bands, and then correlated the catalogues
(see Section 2), we require, in the simulations, that
the synthetic stars pass the test in both photometric bands. Photometric errors are assigned on
the basis of the distribution of the output–minus–
input magnitudes of the artificial stars. These
errors take into account the various instrumental
and observational effects, as well as systematic uncertainties due to crowding (i.e. blend of fainter
objects into an apparent brighter one).
The extraction of (mass, age) pairs is stopped
when the number of stars populating the syn-
thetic CMD (or portions of it) equals the observed one. The solution to this procedure is not
unique, and consists of several (SFH, IMF, metallicity, distance, reddening) combinations which are
in agreement with the observed CMD morphologies and the luminosity functions for different color
bins.
The synthetic CMDs are compared to the observed ones separately for regions 7, 6, 5, 3-4 and
0-1-2. It is important to recall that the B, B-V
CMDs sample only the area of intersection between the U, B and V, I data-sets. In particular, while there is a good spatial overlap for the
most central regions (7,6,5), sampled by the PC
camera, the superposition for the more external
regions, sampled by the WF cameras, is poorer.
The U, B data sample only ∼ 3/4 and ∼ 1/2
of the areas covered by regions 3-4 and 0-1-2 in
the V, I data, respectively. This implies that the
SFH presented in A03 must be scaled by these factors before comparison with the observed B, B-V
CMDs for regions 3–4 and 0–1–2. Of course this
re-scaled SFR is an approximation, because it is
possible that there are spatial variations in the
intensity of the SF within the same region. Table 3 summarizes the SFH for the different regions
of NGC1705. The table is the same as Table 6
in A03, except for the entries in regions 3–4 and
0–1–2, which have been scaled by 0.75 and 0.5,
respectively, to account for the areas of overlap.
The synthetic B, B-V CMDs and LFs obtained
assuming the SFH of Table 3 are presented in
Fig. 11. It is apparent that there is a good agreement between the simulations and the data, also
considering that we have not performed any finetuning. In particular, the comparison between the
LFs is statistically meaningful, as they have been
obtained through an average on several runs corresponding to the same SFH. This allows us to
minimize random effects due to the small number
of objects in the CMDs. Performing an average
on several runs, we recover 493, 2483, 1693, 196
and 179 stars in the synthetic CMDs for regions
7–6–5–34 and 012, respectively. For the same regions, the number of observed stars is 525, 2704,
1925, 236 and 149. Thus, simulations and data
agree within 10 % for regions 7–6–5 and within 20
% for regions 3–4 and 0–1–2.
In the simulations of Fig. 11, the objects present
in the B, B-V CMDs are mainly young stars in the
9
MS or blue-loop phases. Stars older than ∼ 1 Gyr
are barely detected in B, given the incompleteness
of our data. This implies that the new data can
provide a better insight on the recent evolution of
NGC 1705, but not on its earlier phases, which
were much better analysed with the V and I data
by A03, with a lookback time of about 5 Gyr.
The CMDs of Fig. 11 shows that the SFH derived by A03 provides a good agreement with the
new data. However, we performed additional simulations to better constrain and validate the SFH
during the last 50 Myr. From Fig. 11, we notice
that stars older than ∼ 50 Myr populate the CMD
at B > 23.5. Thus we focus on the portion of the
observed CMD brighter than B = 23.5 to constrain
the most recent SFH .
As a first check, we tested the need for a strong
SF activity started ∼3 Myr ago (B2), as derived
by A03. To this purpose we performed new simulations without B2. We obtained new synthetic
CMDs adopting the SFH of A03, but suppressing
the SF activity in the last 3 Myr; the stars of B2
were then re-distributed either between 10 and 15
Myr ago (this case is shown in Fig.12), or between
50 and 10 Myr ago. In both cases, the simulations
severely underestimate the number of blue stars
at 22.5 < B < 23.5, and thus confirm the need
for a strong SF activity started only few Myrs ago
and still ongoing. Only in the outermost regions
0-1-2 the significance of this result may be questioned: too few stars are measured in B and V in
the periphery to clearly distinguish between the
two scenarios.
We notice that despite the overall good agreement with the data, the simulations in Fig. 11
slightly underproduce the stars in the brightest
magnitude bins (B<21). A comparison with the
stellar evolution tracks indicate that these objects
are likely very massive stars that started to evolve
out of the MS phase after ≈ 3 Myr since their
birth, and that have a lifetime . 10 Myr. Thus,
we tested the need for an interruption in the SF
activity between burst B1 and burst B2 as derived
by A03. As a first check, we simulated the totality of the stars produced in the two bursts within
a single episode started 15 Myr ago and still ongoing. This translates into rates of 5.3 × 10−2 ,
3.5 × 10−2 , and 6.3 × 10−3 M⊙ yr−1 for Regions
7, 6 and 5 , respectively. The synthetic CMDs
and LFs for this scenario, shown in Fig.13, are
strongly inconsistent with the data for Regions 7
and 6, since they cause an overproduction of very
luminous blue stars. For example, in Reg. 7 the
simulation provides ≈ 90 stars with B − V < 0.4
and B < 22 against 35 stars observed, and 6 stars
with B < 20.5 against 0 observed. In Reg 5, instead, the data are consistent with a continuous
episode from 15 Myr to now at a rate of 6 × 10−3
M⊙ yr−1 .
Because of the small statistics at B<21, we did
not attempt to derive a best fit SFR between B1
and B2, but rather tried to derive an upper limit.
To this purpose, we simulated new CMDs where
we forced the SFR between B1 and B2 to assume
increasing values from 10−4 to 10−1 M⊙ yr−1 . The
rates of the other episodes were re-scaled from our
best-fit SFH in order to reproduce the number
of observed stars in each region. Then, we compared the data with the simulations focusing on
the brightest (B < 21) portion of the CMD, sampling the post-MS phase of massive (M > 20 M⊙ )
stars with ages between 3 and 10 Myr. To derive
a conservative upper limit for the interburst rate,
we required statistical consistency at a 95 % confidence level between the simulated and observed
counts at B < 20.5 and 20.5 < B < 21. This
provides an upper limit of ≈ 10−2 M⊙ yr−1 for
both Regions 7 and 6. For instance, a rate of 10−2
M⊙ yr−1 produces, over 10 runs, an average of
∼ 3.5 simulated counts at B < 20.5 in Reg. 7,
against 0 counts observed. The extracted counts
follow a Poisson distribution, and are consistent
with the observed counts with a probability <5
%. The derived upper limit of 10−2 M⊙ yr−1 is a
factor ≈ 4 and 16 lower than the rates derived in
Reg. 7 for B1 and B2, respectively, and a factor
≈ 2 and 10 lower than the rates of B1 and B2 in
Reg. 6. Our results indicate that the interburst
state was more likely characterized by a lower level
of star formation rather than by its complete cessation.
The occurrence of the two recent bursts, separated by only few Myrs, is particularly intringuing
since it suggests a feedback mechanism as the trigger for the more recent episode. To get insight into
this possible scenario, we investigated the spatial
distribution of the stars generated during B1 and
B2 in the PC field. With the help of the synthetic
CMD simulations performed, we selected in the
observed CMD stars younger than ≈ 5 Myr, be10
longing to B2 ((B−V), (V−I) . 0.3, 22.5<B<23.5,
22.5<I<24 ) and stars with ages (10-15) Myr, generated during B1 (I . 21), and plotted their spatial distribution in Fig.14. Intermediate-age ((501000) Myr) stars and candidate clusters are plotted as well for comparison. Fig.14 shows that
while B1 is confined to the most central region, the
current burst B2 is more extended, and presents
some ring/arc - like structures which remind of an
expanding shell. This suggests a scenario in which
SF occurred 10-15 Myr ago in the very center of
NGC 1705; then multiple supernovae explosions
generated a strong galactic wind and a superbubble that, expanding, shocked and compressed the
surrounding ISM; when the gas cooled after few
Myrs, new more external SF occured in the regions shocked by the superbubble (see also Burkert 2004).
6.
the rate in the interburst interval rather than for a
complete cessation of the activity. We derived an
upper limit of 10−2 M⊙ yr−1 for the rate allowed
between the two bursts, which is a factor ∼ 4 and
16 lower than the rates of B1 and B2 in the center of NGC 1705, respectively. This result agrees
with the conclusions of Lee et al. (2009) based on
the analysis of the Hα component of the 11HUGS
Survey. The authors showed that the complete
cessation of star formation generally does not occur in irregular galaxies, and is not characteristic
of the interburst phase.
The two bursts in NGC 1705 appear well separated in space: while B1 is centrally concentrated,
B2 occurs all over the galaxy, and presents ring
and arc-like structures reminiscent of an expanding shell. These results suggest a feedback scenario in which the most recent burst was triggered
by the expanding superbubble generated during
the previous (10–15) Myr burst. Hydrodynamical simulations (Burkert, private communication)
show that a few Myr interval between the two
bursts is sufficient for the shocked gas to cool down
and allow the formation of new stars.
The excellent spatial resolution of the HRC
allowed us to reliably identify 12 star clusters
(plus the SSC) in the central ∼ 26” × 29” region
of NGC 1705, and to study their morphological
properties. The clusters have intrinsic effective
radii from ∼ 1 to 6 pc. From the comparison of
their integrated WFPC2/NIC2 UBVIJH photometry with the GALEV population synthesis models, we derive cluster ages from ≈ 10 Myr to ≈ 1
Gyr, and masses between ≈ 104 and 105 M⊙ .
Some of the properties derived for NGC 1705 in
A03 and in this work are similar to those inferred
for all the other late-type dwarfs whose resolved
stellar populations have been studied so far, either
inside or outside the Local Group (for a summary,
see e.g. Tosi 2007). All of them have been shown
to be already forming stars at the lookback time
reached by the photometry, both irregulars (see
e.g. Aparicio, Gallart & Bertelli 1997 for Pegasus,
Dolphin et al. 2001 for the SMC, Smecker-Hane
et al. 2002 for the LMC, Clementini et al. 2003
for NGC 6822, Dolphin et al. 2003 for Sextans A,
Skillman et al. 2003 for IC 1613, Grocholski et
al. 2008 for NGC 1569, Cole et al. 2007 for Leo
A) and BCDs (see e.g. Lynds et al. 1998 for VIIZw403, Schulte-Ladbeck et al. 2000 for Mrk 178,
Discussion and conclusions
With the new data presented here (WFPC2
imaging in U and B, HRC imaging in U, V and
I) we have completed our multiband HST photometric analysis of NGC 1705, from the ultraviolet
to the near infrared. Since the new data are mostly
sensitive to bright blue stars, they provide a better insight on the recent evolution of NGC 1705,
but not on its old SFH. Epochs older than few
hundreds Myr were much better constrained from
the analysis of the V, I, J and H data by A03,
who concluded that NGC 1705 was already forming stars several Gyrs ago. Between ≈ 1 Gyr and
50 Myr ago, they also derived a ”fluctuating” star
formation of moderate strength, excluding long interruption in the SF activity.
In this paper, we applied the method of the synthetic CMDs to the new data to better constrain
the SFH of NGC 1705 in the last ∼ 50 Myr. We
confirm the presence of two strong young bursts,
as derived by A03 from longer wavelength data.
The older of the two bursts (B1) occurred between
∼ 10 and 15 Myr ago, and is coeval to the age of
the central SSC. The younger burst (B2) started
∼ 3 Myr ago, and it is still active. The stellar mass
produced by B2 amounts to ∼ 106 M⊙ , and it is a
factor of ∼ 3 lower for B1. The new data allowed
us also to test for possible SF activity between
the two bursts. The comparison of the data with
detailed simulations shows evidence for a drop of
11
Schulte-Ladbeck et al. 2001 for IZw36, Aloisi et
al. 2005 for SBS1415+437, Aloisi et al. 1999,
Ostlin 2000 and Aloisi et al. 2007 for IZw18). In
other words, no galaxy has been found yet with
evidence of having started to form stars only recently, not even the most metal-poor ones, IZw18
and SBS1415+437.
Also common to all the late-type dwarfs studied by means of the CMDs of their resolved stellar
populations is the result that their SFH has been
fairly continuous, with fluctuations in the SFR and
the possibility of short quiescent phases (gasping
SF regime, Marconi et al. 1995). What makes
NGC 1705 outstanding is the strength of its recent
SF episodes, comparable only to that of the other
very active and windy dwarf irregular NGC 1569.
It is interesting to emphasize that the strength of
the SF activity does not appear to be related to
the morphological classification of the galaxy: all
the other BCDs with inferred SFH have SFR comparable to those of nearby dwarf irregulars (i.e. of
the order of 0.005 M⊙ yr−1 kpc−2 , while the recent burst in NGC 1705 has a SFR density of 1
M⊙ yr−1 kpc−2 and that in NGC 1569 has > 4
M⊙ yr−1 kpc−2 ). In this respect dwarf irregulars
and BCDs do not differ from each other. One may
actually think, on the basis of the current sample, that BCDs are simply dwarf irregulars with
a recent SF activity strong enough to let them be
discovered in spite of the distance and the relative
faintness.
NGC 1705 and NGC 1569, in spite of being
differently classified, have many common features,
in addition to (or because of) the strength of their
recent SF activity: they both show observational
evidence of galactic winds, they both host SSCs
(1 NGC 1705, and 3, possibly 6, depending on the
definition, NGC 1569), and they both contain a
large number of star clusters (see also Billett et
al. 2002).
Galaxy evolution models (Romano et al. 2006),
computed adopting the SFH inferred from the
CMDs of the resolved stars, show that both in
NGC 1705 and NGC 1569 the strong SF activity
triggers violent galactic winds powered by Supernova explosions. The predicted winds eject total
masses of gas consistent with the observational estimates (MFDC, Martin, Kobulnicky & Heckman
2002). They remove efficiently from the galaxies
the metals produced by the Supernovae but only
tiny fractions of the interstellar medium, as expected on the basis of both observations and hydrodynamical models (e.g. De Young & Gallagher
1990, D’Ercole & Brighenti 1999, Mac Low & Ferrara 1999, Martin et al. 2002, Recchi et al. 2006).
These differential winds appear to be the only viable explanation to reconcile the strength of the
recent SF activity, and the long duration of the
previous one, with the current low metallicities
and high gas content observed in both galaxies.
The presence of one SSC, one populous cluster and several regular star clusters in the central
regions of NGC 1705 confirm its unusually high activity, surpassed only by the extreme one of NGC
1569. The fairly long age range of these clusters
also indicates a prolonged star formation activity.
Once again, NGC 1569 and NGC 1705 look fairly
similar to each other in their star and cluster formation efficiencies higher than in other late-type
dwarfs.
We are grateful to Luca Angeretti for his support on the synthetic CMD code and Livia Origlia
for providing the photometric conversion tables
to the Vegamag system. We thank A. Burkert,
R. P. van der Marel and U. Fritze-v. for useful discussions and P. Anders for providing the
GALEV models. We thank the anonymous referee
for the very useful comments to improve the paper. We acknowledge financial contribution from
INAF through PRIN-2005 and ASI-INAF through
contract I/016/07/0.
REFERENCES
Aloisi, A., Annibali, F., Mack, J., Tosi, M., van der
Marel, R.P., Clementini, G., Contreras, R.A.,
Fiorentino, G., et al. 2007, IAU Symp. 241,
310
Aloisi, A., Heckman, T.M., Hoopes, C.G., Leitherer, C., Savaglio, S., Sembach, K.R., 2005 in
Starbursts: from 30 Doradus to Lyman Break
galaxies; R. de Grijs, R.M. Gonzales-Delgado
eds (Springer, Dordrecht), p.P2
Aloisi, A., Tosi, M., & Greggio, L. 1999, AJ, 118,
302
Aloisi, A., van der Marel, R.P., Mack, J., Leitherer, C., Sirianni, M., & Tosi, M. 2005, ApJ,
631, L45
12
Anders, P., & Fritze-v. Alvensleben, U. 2003,
A&A, 401, 1063
L., Held, E. V., Romano, D., Sirianni, M., Tosi,
M., 2008, ApJL, 686, L79
Angeretti, L., Tosi, M., Greggio, L., Sabbi, E.,
Aloisi, A., & Leitherer, C. 2005, AJ, 129, 2203
Heckman, T.M., & Leitherer, C. 1997, AJ, 114, 69
Heckman, T.M., Sembach, K.R., Meurer, G.R.,
Strickland, D.K., Martin, C.L., Calzetti, D., &
Leitherer, C., 2001, ApJ, 554, 1021
Annibali, F., Greggio, L., Tosi, M., Aloisi, A., &
Leitherer, C., 2003, AJ, 126, 2752 (A03)
Aparicio, A., Gallart, C., & Bertelli, G., 1997a,
AJ, 114, 669
Hill, J.R. et al. 1998, ApJ, 496, 648
Baggett, S., Casertano, S., Gonzaga, S., & Ritchie,
C. 1997, ISR WFPC2 97-10
Holtzman, J.A., Burrows, C., Casertano, S., Hester, J., Trauger, J., Watson, A., & Worthey, G.
1995a, PASP, 107, 1065
Ho, L.C., & Filippenko, A.V. 1996, ApJ, 466, L83
Bertelli, G., Bressan, A., Chiosi, C., Fagotto, F.,
& Nasi, E. 1994, A&AS, 106, 275
Holtzman, J.A., et al. 1995b, PASP, 107, 156
Billett, O.H., Hunter, D.A., & Elmegreen, B.G.
2002, AJ, 123, 1454
Holtzman, J. A., Afonso, C., Dolphin, A. 2006,
ApJS, 166, 534
Biretta, J.A., et al. 2000, WFPC2 Instrument
Handbook, Version 5.0 (Baltimore: STScI)
Hunter, D.A., O’Connell, R.W., Gallagher, J.S. &
Smecker-Hane, T.A. 2000, AJ, 120, 2383
Burkert, A. 2004, The Formation and Evolution
of Massive Young Star Clusters, 322, 489
King, I. 1962, AJ, 67, 274
D’Ercole, A., & Brighenti, F. 1999, MNRAS, 309,
941
Koekemoer, A. M., Fruchter, A. S., Hook, R. N.,
& Hack, W. 2002, The 2002 HST Calibration
Workshop, 2002. Edited by Santiago Arribas,
Anton Koekemoer, and Brad Whitmore. Baltimore, MD: Space Telescope Science Institute,
2002., p.337, 337
De Young, D.S., &gallagher, J.S. 1990, ApJ, 356,
L15
Krist, J., & Hook, R., 1999, Tiny Tim User Manual Version 5.0, (Baltimore: STScI)
Diolaiti, E., Bendinelli, O., Bonaccini, D., Close,
L., Currie, D., Parmeggiani, G., 2000, A&AS,
147, 335
Larsen, S. S. 1999, A&AS, 139, 393
Dolphin, A.E., Saha, A., Skillman, E.D., et al.
2003, AJ, 126, 187
Lee, J. C., Kennicutt, R. C., José G. Funes, S. J.,
Sakai, S., & Akiyama, S. 2009, ApJ, 692, 1305
Dolphin, A.E., Walker, A.R., Hodge, P.W., Mateo, M., Olszewski, E.W., Schommer, R.A.,
Suntzeff, N.B.. 2001, ApJ, 562, 303
Leitherer, C., et al. 1999, ApJS, 123, 3
Clementini, G., Held, E.V., Baldacci, L., & Rizzi,
L. 2003, ApJ, 588, L85
Cole, A.A., et al. 2007, ApJ, 659, L20
Larsen, S.S. & Richtler, T. 2000, A&A, 354, 836
Lee, H., & Skillman, E.D. 2004, ApJ, 614, 698
Lynds, R., Tolstoy, E., O’Neil, E.J.Jr., & Hunter,
D.A. 1998, AJ116, 146
Fagotto, F., Bressan, A., Bertelli, G., & Chiosi, C.
1994a, A&AS, 104, 365
Mac Low, M.-M. & Ferrara, A. 1999, ApJ, 513,
142
Fagotto, F., Bressan, A., Bertelli, G., & Chiosi, C.
1994b, A&AS, 105, 29
Marconi, G., Tosi, M., Greggio, L., & Focardi, P.
1995, AJ, 109, 173
Fruchter, A.S., & Hook, R.N. 1998, PASP,
astro-ph/9808087
Martin, C.L., Kobulnicky, H.A., & Heckman,
T.M. 2002, ApJ, 574, 663
Greggio, L., Tosi, M., Clampin, M., De Marchi,
G., Leitherer, C., Nota, A., & Sirianni, M. 1998,
ApJ, 504, 725
Melnick, J., Moles, M., & Terlevich, R. 1985,
A&A, 149, L24
Meurer, G.R, Freeman, K.C., Dopita, M.A., Cacciari, C. 1992, AJ, 103, 60 (MFDC)
Grocholski, A. J., Aloisi, A., van der Marel, R. P.,
Mack, J., Annibali, F., Angeretti, L., Greggio,
13
O’Connell, R.W., Gallagher, J.S., & Hunter, D.A.
1994, ApJ, 433, 65
Origlia, L. & Leitherer, C. 2000, AJ, 119, 2018
Origlia, L., Leitherer, C., Aloisi, A., Greggio, L.,
Tosi, M. 2001, AJ, 122, 815
Ostlin, G. 2000, ApJ, 535, L99
Quillen, A.C., Ramirez, S.V., & Frogel, J.A., 1995,
AJ, 110, 205
Recchi, S., Hensler, G., Angeretti, L. & Matteucci,
F. A&A, 445, 875
Romano, D., Tosi, M., & Matteucci, F. 2006 MNRAS, 365, 759
Sirianni, M., Meurer, G., Homeier, N., Clampin,
M., Kimble, R., & The ACS Science Team 2005,
Starbursts: From 30 Doradus to Lyman Break
Galaxies, 329, 41
Schulte-Ladbeck, R.E., Hopp, U., Greggio, L. &
Crone, M.M. 2000 AJ, 120, 1713
Schulte-Ladbeck, R.E., Hopp, U., Greggio, L.,
Crone, M.M., & Drozdovsky, I.O. 2001, AJ,
121, 3007
Skillman, E.D., Tolstoy, E., Cole, A.A, Dolphin,
A.E., Saha, A., Gallagher, J.S., Dohm-Palmer,
R.C., Mateo, M. 2003, ApJ, 596, 253
Smecker-Hane, T.A., Cole, A.A., Gallagher, J.S.,
& Stetson, P.B. 2002, ApJ, 566, 239
Stetson, P. B. 1987, PASP, 99, 191
Storchi-Bergmann, T., Calzetti, D. & Kinney, A.
1994, ApJ429, 572
Tosi, M. 2007, in From Stars to Galaxies, A.Vallenari,
R.Tantalo, L.Portinari & A.Moretti eds, ASP
Conf.Ser., 374, p. 221
Tosi, M., Greggio, L., Marconi, G., & Focardi, P.
1991, AJ, 102, 951
Tosi, M., Sabbi, E., Bellazzini, M., Aloisi, A.,
Greggio, L., Leitherer, C., Montegriffo, P.,
2001, AJ, 123, 1271, T01
Whitmore, B., Heyer, I., & Casertano, S. 1999,
PASP, 111, 1559
This 2-column preprint was prepared with the AAS LATEX
macros v5.2.
14
Table 1
B-Band Completeness levels in each region from the artificial star tests.
mag
20
22
24
26
28
c7
c6
c5
c4
c3
c2
c1
c0
1.00
0.94
0.64
0.13
0.00
1.00
0.94
0.82
0.20
0.00
1.00
0.82
0.89
0.60
0.00
1.00
1.00
0.54
0.61
0.02
1.00
1.00
0.54
0.78
0.09
1.00
1.00
0.97
0.92
0.12
1.00
1.00
0.97
0.91
0.11
1.00
1.00
0.96
0.84
0.10
15
Table 2
NGC 1705 cluster properties
Cluster
re (”)
re (pc)a
MaF555W
mbF330W
mbF380W
mbF439W
mbF555W
mbF814W
mbF110W
mbF160W
E(B − V)
Age(Myr)
Mass (104 M⊙ )
SSCc
1
2
3
4
5
6∗
7
8∗
9
10
11
12
0.08
0.04
0.13
0.26
0.19
0.15
0.175
0.22
0.15
0.1
0.26
0.16
0.21
1.98
0.99
3.21
6.43
4.7
3.71
4.33
5.44
3.71
2.47
6.43
3.95
5.19
-13.8
-9.57
-7.53
-7.22
-7.54
-7.35
-6.66
-6.84
-6.85
-7.57
-7.22
-6.31
-7.19
13.6
17.77
21.23
21.93
21.29
21.61
21.89
22.26
21.84
21.84
21.59
23.08
22.30
–
18.72
21.61
21.83
21.60
21.66
22.20
22.41
22.01
21.85
21.76
23.21
22.33
–
19.01
21.64
21.74
21.62
21.62
22.23
22.38
22.02
21.72
21.77
23.05
22.24
14.9
19.08
21.43
21.58
21.42
21.30
21.99
22.12
21.80
21.08
21.55
22.34
21.46
15.0
18.89
20.79
21.18
20.88
20.67
21.60
21.57
21.43
20.21
21.09
21.33
20.52
–
18.68
20.51
–
–
20.01
21.52
21.31
21.28
19.71
20.76
20.84
20.06
–
18.18
19.83
–
–
19.19
21.16
20.60
20.81
18.95
20.22
20.14
19.36
0.06
0.0
0.1
0.05
0.1
0.0
≤0.15
0.1
≤0.2
0.0
0.1
0.0
0.0
12
16+2
−2
53+1
−4
97+8
−49
101+11
−14
356+202
−73
8 − 80
101+9
−60
8 − 90
1191+209
−210
80+9
−17
1427+705
−371
1260+410
−317
59
6.15+0.1
−1.3
1.74+0.1
−0.1
1.69+0.1
−0.8
2.41+0.1
−0.2
3.73+1.7
−0.5
0.3 − 1
1.28+0.1
−0.6
0.3 − 1.2
9.39+0.7
−1.3
1.94+0.1
−0.2
3.86+2.2
−0.5
6.94+2.7
−0.2
a
16
Correcting for (m − M )0 = 28.54, corresponding to a distance of 5.1 Mpc (Tosi et al. 2001), for a galactic
reddening E(B−V)=0.035, and for the intrinsic E(B−V) listed in this Table.
b
Apparent magnitudes measured within a radius R = re ∗ P SF
c
From Sirianni et al. (2005)
∗
For these clusters the fit is problematic and provides a wide range of solutions.
Table 3
Average SFRs∗ at various epochs
Region
7
6
5
3–4
0–1–2
total
∗
Area (kpc2 )
0.017
0.13
0.35
0.64
5.6
6.7
(0-3) Myr
(3-10) Myr
SFR (M⊙ yr −1 )
(10-15) Myr (15-50) Myr
(50-1000) Myr
0.16
0.11
0.032
0.005
0.0025
0.31
−
−
−
−
−
−
0.044
0.022
2.1 × 10−3
−
−
6.8 × 10−2
4.3 × 10−3
0.024
0.022
5.5 × 10−3
2.5 × 10−4
5.6 × 10−2
−
0.005
2.7 × 10−3
−
−
7.7 × 10−3
(1-5) Gyr
?
1.3 × 10−2
1.6 × 10−2
0.75 × 10−2
0.85 × 10−2
4.5 × 10−2
The rates are those derived in A03 and re-scaled to the area of overlap between the (V, I) and (U,B) datasets.
17
Fig. 1.— WFPC2 images of NGC 1705 in F439W and F814W. The different orientations of the two observing
runs unfortunately prevent the images to completely overlap as planned. North is up and East is to the left.
18
Fig. 2.— True color WFPC2 image of the regions of NGC 1705 where the F380W, F439W, F555W, and
F814W fields overlap. In the rest of the fields either 19
F380W and F439W are available (blue portions of the
figure), or the F555W and F814W (yellowish portions). Overimposed (in green) are the isophotal contour
levels adopted by TO01 to define 8 roughly concentric regions (see text for details).
Fig. 3.— CMDs in different bands of the 256 objects with measured U, B, V, I, J, H magnitudes (calibrated
in the HST-Vegamag system). Dots represent resolved stars and filled circles candidate clusters. The
evolutionary tracks of a 60M⊙ (dotted line) and of a 15M⊙ (solid line) with Z=0.004 (Fagotto et al. 1994b)
are overplotted in nine panels.
20
Fig. 4.— Magnitude difference (∆mag = input – output) versus input magnitude in the HST-Vegamag B
band of the artificial stars in the concentric regions. The standard deviations in 1 mag bins around B = 23,
25 and 27 are indicated for each region. The lines superimposed on the diagrams represent the local mean
∆m and the ±1 standard deviations.
21
Fig. 5.— Three-color (F330W (U, blue), F555W (V, green), and F814W (I, red)) composite image of
NGC 1705 obtained with ACS/HRC, showing a field of view of ≈ 26 × 29 arcsec2. The selected candidate
clusters are indicated on the image.
22
Fig. 6.— Intrinsic effective radius distribution of the 12 candidate star clusters from ACS/HRC data. A
distance of 5.1 Mpc was adopted for NGC 1705.
23
Fig. 7.— Left panel: U-V versus V-I diagram for the 12 selected candidate clusters. Right panel: U-V
versus J-H diagram for the 10 clusters that have photometry in all the UBVIJH bands. The lines are the
SSP GALEV models (Anders& Fritze-v. Alvensleben 2003) for different metallicities and ages from 4 Myr
to 10 Gyr. The position of the Log(age(yr))= 7,8,9,10 models in the color-color plane is indicated by the
black dots with the age labels for the metallicity Z = 0.004. No intrinsic reddening was applied.
24
Fig. 8.— Left panel: U-V versus V-I diagram for the 12 selected candidate clusters. Right panel: U-V
versus J-H diagram for the 10 clusters that have photometry in all the UBVIJH bands. The lines are the
SSP GALEV models (Anders& Fritze-v. Alvensleben 2003) for a metallicity Z = 0.004; the dotted one shows
the effect of a reddening of E(B − V ) = 0.2 in combination with a Cardelli et al. (1989) law. The position of
the Log(age(yr))= 7,8,9,10 models in the color-color plane is indicated by the black dots with the age labels
for E(B − V ) = 0.
25
Fig. 9.— Age (top panel) and mass (bottom panel) distributions obtained for the 12 candidate star clusters
by fitting the UBVIJH photometry with the GALEV models.
26
Fig. 10.— CMDs in the different colors of the concentric regions of NGC 1705 (labelled in the top-left corner
of the left-hand panels). Magnitudes are in the HST-Vegamag system. Overimposed in the middle and
right-hand panels are the Padova stellar evolution tracks for metallicity Z=0.004 (Fagotto et al. 1994b) and
masses from 0.9 to 30 M⊙ (the displayed masses are 30, 15, 9, 7, 5, 4, 3, 2, 1.8, 1.6, 1.4, 1.2, 1.0, and 0.9
M⊙ ). The 15 M⊙ track is plotted with the thick line. The number of stars in each CMD is indicated in the
bottom-right corner.
27
Fig. 11.— Comparison bewteen B,V data and simulations. The observed B vs B-V CMDs (in the HSTVegamag system) for different regions are displayed in the left-hand panels. Next to the observed CMDs,
we display the synthetic ones obtained by assuming the SFH derived by A03, reported in Table 3. In the
synthetic CMDs, the stars have been color-coded according to their age: blue = (0-3) Myr; cyan=(10-15)
Myr; green=(15-50) Myr; red=(50-1000) Myr; black= (1-14) Gyr. On the right we display the LFs. Dots
are for the data, continuous line for the simulations. From left to right, the LFs refer respectively to B-V
≤ 0.4, B-V> 0.4 and to the entire color range.
28
Fig. 12.— Synthetic CMDs obtained from the SFH of Tab. 3, but suppressing B2 (0-3 Myr). The stars of
B2 have been re-distributed into B1 (10 -15 Myr). The color-coding is the same as in Fig. 11.
29
Fig. 13.— Synthetic CMDs obtained from the SFH of Tab. 3 but without quiescent interval between the
two most recent bursts, and only for Regions 7, 6 and 5. From 15 Myr ago to now, the SFRs are 5.3 × 10−2 ,
3.5 × 10−2 , and 6.3 × 10−3 M⊙ yr−1 for Regions 7, 6 and 5, respectively. The simulated LFs are plotted with
a thick line. For Regions 7 and 6, we also display the LFs obtained assuming a rate of 10−2 M⊙ yr−1 in the
last 15 Myr (thin line). The color-coding is the same as in Fig. 11, with the addition of orange dots for the
stars with age (3-10) Myr. .
30
nopagenumbers
1
Fig. 14.— PC image of NGC 1705 in F439W (top) and spatial distribution of stars with different ages: .
5 Myr, (10-15) Myr, and (50-1000) Myr. The candiate clusters are overplotted on the resolved stars in the
bottom panel.
31