Mapping the properties of blue compact dwarf galaxies: integral field spectroscopy with PMAS
L.M. Cairos, N. Caon, C. Zurita, C. Kehrig, M. Roth, P. Weilbacher
aa r X i v : . [ a s t r o - ph . C O ] A p r Astronomy&Astrophysicsmanuscript no. cairos-8feb-astroph c (cid:13)
ESO 2018September 29, 2018
Mapping the properties of blue compact dwarf galaxies: integral field spectroscopy with PMAS
L. M. Cair´os , N. Caon , C. Zurita , C. Kehrig , M. Roth , and P. Weilbacher Astrophysikalisches Institut Potsdam, An der Sternwarte 16, D-14482 Potsdam, Germanye-mail: luzma; ckehrig; mmroth; [email protected] Instituto de Astrof´ısica de Canarias, E-38200 La Laguna, Tenerife, Spain and Departamento de Astrof´ısica, Universidad de laLaguna, E-38205, La Laguna, Tenerife, Spaine-mail: nicola.caon; [email protected]
Received 7 January 2010; accepted 29 March 2010
ABSTRACT
Context.
Blue compact dwarf (BCD) galaxies are low-luminosity, low-metal content dwarf systems undergoing violent bursts of starformation. They present a unique opportunity to probe galaxy formation and evolution and to investigate the process of star formationin a relatively simple scenario. Spectrophotometric studies of BCDs are essential to disentangle and characterize their stellar popula-tions.
Aims.
We perform integral field spectroscopy of a sample of BCDs with the aim of analyzing their morphology, the spatial distribu-tion of some of their physical properties (excitation, extinction, and electron density) and their relationship with the distribution andevolutionary state of the stellar populations.
Methods.
Integral field spectroscopy observations of the sample galaxies were carried out with the Potsdam Multi-ApertureSpectrophotometer (PMAS) at the 3.5 m telescope at Calar Alto Observatory. An area 16 ′′ × ′′ in size was mapped with a spa-tial sampling of 1 ′′ × ′′ . We obtained data in the 3590–6996 Å spectral range, with a linear dispersion of 3.2 Å per pixel. From thesedata we built two-dimensional maps of the flux of the most prominent emission lines, of two continuum bands, of the most relevantline ratios, and of the gas velocity field. Integrated spectra of the most prominent star-forming regions and of whole objects within theFOV were used to derive their physical parameters and the gas metal abundances. Results.
Six galaxies display the same morphology both in emission line and in continuum maps; only in two objects, Mrk 32 andTololo 1434 + ff er considerably. In general the di ff erent excitation maps fora same object display the same pattern and trace the star-forming regions, as expected for objects ionized by hot stars; only the outerregions of Mrk 32, I Zw 123 and I Zw 159 display higher [S ii ] / H α values, suggestive of shocks. Six galaxies display an inhomoge-neous dust distribution. Regarding the kinematics, Mrk 750, Mrk 206 and I Zw 159 display a clear rotation pattern, while in Mrk 32,Mrk 475 and I Zw 123 the velocity fields are flat. Key words. galaxies: starburst - galaxies: dwarf - galaxies: stellar content - galaxies: abundances
1. Introduction
Blue compact dwarf (BCD) galaxies are narrow emission-lineobjects, which undergo at the present time violent bursts of starformation (Sargent & Searle 1970). They are compact and low-luminosity objects (starburst diameter ≤ M B ≥ − Z ⊙ / ≤ Z ≤ Z ⊙ /
2) andhigh star-forming (SF) rates, able to exhaust their gas contenton a time scale much shorter than the age of the Universe.Initially it was hypothesized that BCDs were truly young galax-ies, forming their first generation of stars (Sargent & Searle1970; Lequeux & Viallefond 1980; Kunth et al. 1988), but thesubsequent detection of an extended redder stellar host galaxy inthe vast majority of them has shown that most BCDs are actuallyold systems (Loose & Thuan 1986; Telles 1995; Papaderos et al.1996; Cair´os et al. 2001a,b, 2002, 2003) undergoing recur-rent star-formation episodes (Thuan 1991; Mas-Hesse & Kunth1999).These galaxies present a unique opportunity to gain insightson central issues in contemporary galaxy research. Chemicallyunevolved nearby SF dwarfs like BCDs are an important linkto the early Universe and the epoch of galaxy formation, as they have been regarded as the local counterparts of thedistant subgalactic units (building blocks) from which largersystems are created at high redshifts (Kau ff mann et al. 1993;Lowenthal et al. 1997); the study of these systems hence pro-vides important insights into the star-formation process of dis-tant galaxies. Moreover, even though most BCDs are not gen-uinely young galaxies, their metal deficiency makes them usefulobjects to constrain the primordial He abundance and to mon-itor the synthesis and dispersal of heavy elements in a nearlypristine environment (Pagel et al. 1992; Masegosa et al. 1994;Izotov et al. 1997; Kunth & ¨Ostlin 2000). Blue compact dwarfsare also ideal laboratories for the study of the starburst phe-nomenon: as they are smaller and less massive than normalgalaxies, they cannot sustain a spiral density wave and do notsu ff er from disk instabilities, which considerably simplifies thestudy of the star formation process. Besides, the radiation emit-ted by their SF regions is less diluted by the stellar continuumthan in giant spiral galaxies, allowing for more precise studiesof element abundance ratios.However, and in spite of the great e ff ort done during the lasttwo decades on the field of BCDs, fundamental questions likethe mechanisms responsible for the ignition of their starburst, L. M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies: their evolutionary status or their SF histories are still far fromwell understood.To answer these questions it is of paramount importance tofirst disentangle and characterize the di ff erent components thatmake up a BCD galaxy. This is a demanding and di ffi cult task.At any location in the galaxy, the emitted flux is the sum of theemission from the local starburst, the flux produced by the neb-ula surrounding the young stars, and the emission from the un-derlying, old stellar population, all possibly modulated by dust(Cair´os et al. 2002, 2003, 2007). Substantial work in the field hasshown that photometry alone does not allow us to distinguishthe di ff erent components in BCDs (see Kunth & ¨Ostlin 2000;Cair´os et al. 2002). The properties of the SF knots in the samegalaxy may vary widely: accounting for the flux in emissionlines through broad-band filters and for the contribution of thestellar host is fundamental to derive the actual broad-band col-ors of the knots (Cair´os et al. 2002, 2007). On the other hand, thedust content (usually assumed to be negligible in BCDs) turnedout to be quite significant in several objects (Hunt et al. 2001;Cair´os et al. 2003; Vanzi & Sauvage 2004).The few spectrophotometric studies performed so far haveshown indeed that they are the right way to tackle the problem:combining high resolution broad- and narrow-band images withhigh-quality spatially resolved spectra does allow us to distin-guish the young stars from the older stars, derive the historyof the SF knots and constrain the evolutionary status of BCDs(Cair´os et al. 2002; Guseva et al. 2003a,b,c; Cair´os et al. 2007).That very few spectrophotometric analyses can be found in theliterature, and virtually all of them focused on one single object,is essentially due to the large amount of observing time that con-ventional observational techniques require. Acquiring images inseveral broad-band and narrow-band filters, plus a sequence oflong-slit spectra sweeping the region of interest translates intoobserving times of two or more nights per galaxy. Thus compre-hensive analysis of a statistically meaningful sample of BCDsbased on traditional imaging and spectroscopic techniques arein terms of observing time just not feasible. Moreover, these ob-servations usually su ff er from varying instrumental and atmo-spheric conditions, which makes combining all these data com-plicated. Long-slit spectroscopy has also the additional problemof the uncertainty on the exact location of the slit.On the other hand, it has been recently shown (Izotov et al.2006; Garc´ıa-Lorenzo et al. 2008; Kehrig et al. 2008;Vanzi et al. 2008; Lagos et al. 2009; James et al. 2009) thatthe state-of-the-art observational technique of integral fieldspectroscopy (IFS) o ff ers an alternative way to approach BCDsstudies in a highly e ff ective manner. IFS provides simultaneousspectra of each spatial resolution element under identicalinstrumental and atmospheric conditions. This is not only amore e ffi cient way of observing, but it also guarantees thehomogeneity of the dataset. In terms of observing time, IFSobservations of BCDs are one order of magnitude more e ffi cientthan traditional observing techniques. This implies that now, forthe first time, spectrophotometric studies of substantial samplesof BCD galaxies have become feasible.Consequently, we have undertaken a long-term project,which aims to map an extensive and representative sample ofBCDs by means of IFS. This galaxy sample, composed of about40 objects, has been chosen so as to span the large range in lu-minosities and morphologies found among the galaxies classi-fied as BCDs. The analysis of such a dataset will allow us to getinsights into basic questions of BCDs research, i.e. how to ef-fectively disentangle the old and young stellar populations, setconstraints of the age and SF history of the galaxies, study the triggering and propagation mechanisms of the star formation andinvestigate the metal abundance patterns.In the first two papers of this series (Cair´os et al. 2009a,b),we illustrated the full potential of this study by showing re-sults on two representatives BCDs, Mrk 1418 and Mrk 409,both observed with the Potsdam multi-aperture spectropho-tometer (PMAS), attached at the 3.5m telescope at Calar AltoObservatory. In this paper, the remaining objects observed withPMAS are studied. The whole sample will be analyzed in a se-ries of future publications.This paper is structured as follows: In Sect. 2 we describethe observations, the data reduction process and the method em-ployed to build the maps. In Sect. 3 we present the main re-sults of the work, that is, the flux, emission line and velocitymaps, as well as the results derived from the analysis of the in-tegrated spectra of the selected galaxy regions. These results arediscussed in Sect. 4 and summarized in Sect. 5.
2. Observations and data reduction
We present and analyze data of eight galaxies, all ofthem previously classified as BCDs. All objects exceptTololo 1434 +
032 are included in the Thuan & Martin (1981)BCD list, while Tololo 1434 +
032 appears cataloged as a BCDin Gil de Paz et al. (2003).Although strictly speaking one of the criteria that agalaxy has to fulfill to be classified as BCD is tohave M B ≥ −
18, it is worth mentioning that mostBCDs studies (Thuan & Martin 1981; Papaderos et al. 1996;Cair´os et al. 2001a,b; Bergvall & ¨Ostlin 2002; Kong & Cheng2002; Gil de Paz et al. 2003) include a relatively high fraction ofgalaxies with luminosities higher than this limit. Indeed, in prac-tice the term BCD designates a set of objects that have a verywide range in properties as luminosities ( − ≥ M B ≥ − Observations were carried out in 2007 March with the PMASinstrument, attached at the 3.5m telescope in the ObservatorioAstron´omico Hispano Alem´an Calar Alto (CAHA). PMAS isan integral field spectrograph, with a lens array of 16 ′′ × ′′ square elements, each 1 ′′ × ′′ in size in the configuration used,connected to a bundle of 256 optical fibers; the fibers are re-arranged to form a pseudoslit in the focal plane of the spec-trograph. The final spectrum is thus composed of 256 spaxels,where by “spaxel” we refer to each element of the 16 x 16fiber matrix. For a detailed description of the instrument seeRoth et al. (2005) and Kelz et al. (2006).A grating with 300 grooves per mm was used during the ob-servations in combination with a SITe ST002A 2K ×
4K CCDdetector. This setup provides a spectral range of 3590–6996 Å, . M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies: 3
Fig. 1.
Sloan Digital Sky Survey (SDSS) images of our sample of galaxies in the g-band; the field of view is 40 arcsec and thecentral boxes indicate the field of view covered by PMAS. North is up, east to the left.
Table 1.
The galaxy sample
Galaxy R.A. Decl. m B D M B A B (J2000) (J2000) (Mpc) (mag) (mag)Mrk 407 09 47 47.6 +
39 05 04 15 . ± .
49 27.2 − .
78 0.069Mrk 32 10 27 02.0 +
56 16 14 16 . ± .
11 16.4 − .
99 0.030Mrk 750 11 50 02.7 +
15 01 24 15 . ± .
54 05.2 − .
82 0.175Mrk 206 12 24 17.0 +
67 26 24 15 . ± .
07 24.3 − .
53 0.071Tololo 1434 +
032 14 37 08.9 +
03 02 50 16 . ± .
86 29.2 − .
42 0.149Mrk 475 14 39 05.4 +
36 48 22 16 . ± .
31 11.9 − .
97 0.052I Zw 123 15 37 04.2 +
55 15 48 15 . ± .
08 15.4 − .
50 0.062I Zw 159 16 35 21.0 +
52 12 53 15 . ± .
28 43.8 − .
56 0.125Notes: R.A., Decl., D and A B taken from the NED (http: // / ). Distances were computed using a Hubble constant of 73 kms − Mpc − and taking into account the influence of the Virgo Cluster, the Great Attractor and the Shapley supercluster. m B taken from HyperLeda(http: // leda.univ-lyon1.fr / ; Paturel et al. 2003) and M B computed from the tabulated values of m B and D . with a linear dispersion of 3.2 Å per pixel (the CCD was binned2 × + + Although several dedicated software packages have been devel-oped in the last years to reduce 3D-spectroscopic data (Becker2002; S´anchez 2006), we decided to process our data using stan- dard IRAF tasks. While using IRAF routines has, with respectto the use of dedicated pipelines, the main drawback of requir-ing a considerable amount of interactive work, which makes thewhole process somewhat slower, it has on the other hand theadvantage of allowing a complete and precise control of all theparameters involved in each data-reduction step.The reduction procedure includes the bias subtraction, imagetrimming, tracing and extraction of the individual spectra, wave-length and distortion calibrations, flat-fielding, combination ofthe individual galaxy frames, sky-subtraction and flux calibra-tion.The first step in the data reduction was the bias subtrac-tion. All the bias exposures were averaged to obtain a masterbias, which was then subtracted from the rest of the frames. Badcolumns were interpolated with the IRAF task fixpix .Next, apertures were defined and traced on the detector.Defining the apertures means to identify on the detector the spec-tra produced by the di ff erent fibers (that is, to find out how many IRAF is distributed by the National Optical AstronomyObservatories, which are operated by the Association of Universitiesfor Research in Astronomy, Inc., under cooperative agreement with theNational Science Foundation. L. M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies:
Table 2.
Log of the observations
Galaxy Exposure time airmass Seeing(s) (arcsec)Mrk 407 6900 1.14–1.01 1.4–1.8Mrk 32 6600 1.06–1.09 1.4–1.5Mrk 750 5400 1.20–1.09 1.9–2.0Mrk 206 6900 1.25–1.16 1.5–2.0Tololo 1434 +
032 6600 1.33–1.21 1.5–1.8Mrk 475 8400 1.03–1.18 1.5–1.9I Zw 123 3900 1.23–1.05 1.8–2.0I Zw 159 5100 1.09–1.04 1.6–1.8 and which pixels on the detector correspond to each fiber). Theapertures are a ff ected by the field distortions and / or by the op-tics of the system, and therefore each aperture does not line upalong the dispersion axis, but has a clear curvature. Hence, afterwe have defined the apertures at a given spectral position, eachof these loci must be traced along the spectral direction.Apertures were defined on well exposed continuum frameswith the IRAF task apall ; the task first finds the centers of eachfiber (the emission peaks) along the spatial axis at some speci-fied position, and then asks for the size of the extraction window,which we set to 6.4 pixels (the best compromise between includ-ing as much signal as possible without contamination by nearbyfibers). The apertures were then traced by fitting a polynomial tothe centroid along the dispersion axis. A fifth degree Legendrepolynomial was found to provide good fits, with a typical RMSof about 0.01 pixels.Once the apertures were defined and traced in the continuumframes, we again used apall to extract them in all the images.The extraction consists of summing the pixels along the spatialdirection into a final one-dimensional spectrum. After that, wehad the so-called “collapse” or “row-stacked” spectra: an image M × N , where M is the number of pixels in the dispersion direc-tion, and N is the number of fibers (256 for PMAS).Afterwards we performed the wavelength calibration and thedispersion correction. In order to calibrate in wavelength weused the tasks identify and reidentify : i) first, in the comparisonspectra (arc) we identified several emission features of a knownwavelength in a reference fiber; ii) second, a polynomial was fit-ted across the dispersion direction; the standard deviation (RMS)of the polynomial fit gives an estimate of the uncertainty in thewavelength calibration. We obtained typical RMS of about 0.01Å by fitting a fifth degree polynomial. iii) next, with reidentify we identified the emission lines in all the remaining fibers of thearc frame, using the selected one as a reference.Because of instrumental flexures, there are significant shifts(by up to two pixels) along both the spatial and the spectral di-rections, even among a sequence of consecutive exposures of thesame object. Spatial shifts can be taken care of by measuring theo ff set between the brightest fiber spectra in the galaxy spectrumand in the corresponding continuum, and applying the correctionin apall .As for the wavelength shifts, first we determined that theseshifts were independent of the wavelength itself by compar-ing arc spectra taken at di ff erent times and telescope positions.We also assessed that the relation between pixel coordinate andwavelength were essentially linear, with negligible deviationsfrom linearity.Then, in each sequence of spectra of a same target we mea-sured the shift in pixel on the spectral axis of the bright sky line at 5577 Å, relative to the first spectrum of the sequence. Thetransformation in wavelength was done with the task dispcor , byslightly modifying the starting and ending wavelengths so that w s = w s , + δ X · D , where w s is the starting wavelength, w s , isthe starting wavelength of the reference spectrum, δ X is the shiftalong the spectral direction, in pixels, and D is the actual disper-sion of the spectra. In this way the sky line ends up at exactly thesame position (within a few hundredths of pixels), which ensuresthat all the wavelength-calibrated spectra of the same object areat the same zeropoint.The wavelength calibrated data were corrected for response(detector pixel sensitivity variations as well as wavelength-dependent variations in the fibers transmission curves) by usingthe wavelength-calibrated continuum frames, and for throughput(variations in the whole responsivity of the lenslets and fibers) byusing the sky-flat exposures. Both steps were carried out simul-taneously by running the task msrep1 .After that the individual galaxy frames were corrected foratmospheric extinction (adopting the “summer extinction coef-ficients” published by S´anchez et al. 2007) and combined withthe task imcombine . The sky was subtracted from the final com-bined frame. For each object we took an o ff set sky exposure ofa shorter duration (typically 5 minutes). Sky spectra were pro-cessed in the same way as galaxy spectra. A one-dimensionalsky spectrum was produced by averaging the signal along thespatial direction with a sigma-clipping algorithm. The flux of thethree to four brightest sky lines was measured in both the finalsky spectrum and in the final galaxy spectrum to determine theappropriate scaling factor (with an accuracy of a few percents)by which to multiply the sky spectrum before subtracting it fromthe galaxy spectrum.Because the relative intensity of di ff erent sky lines varies no-ticeably on short time scales (Patat 2003), it is very di ffi cult tofind a scaling factor that applies equally well to all the sky linesand the sky continuum, and some fine tuning is required. The fi-nal scale factor was found by trial and error; we aimed at a valuethat minimized overall residuals in the sky lines (especially thoseclose to galaxy emission or absorption lines) even if it left largeresiduals in sky bright lines that were not a ff ecting any interest-ing spectral features.We must say here that minimizing sky-line residuals does notnecessarily imply the best match between the sky background inthe sky exposure and the sky background in the galaxy exposure.For this reason the uncertainties on the sky-subtracted galaxycontinuum may be relatively large (and di ffi cult to estimate), es-pecially in the outer regions of galaxies. While this does not atall a ff ect the emission-line parameters (flux, width, redshift), itcan clearly have a significant impact on the equivalent widths of . M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies: 5 the less luminous SF knots and on the outer, fainter spaxels inthe continuum maps.Spectra of spectrophotometric standards were reduced in thesame way, except that the sky spectrum to be subtracted wascomputed with the median of the outermost fibers. The inte-grated spectrum of the spectrophotometric standards was ob-tained by summing all the fibers within a radius of about 2FWHM (typically 3 to 4 arcsec) from the fiber with the high-est signal. The IRAF tasks standard and sensfunc were used toderive the sensitivity curves, after combining the data for the dif-ferent spectrophotometric stars observed in the same night.By comparing the sensitivity curves for di ff erent nights andstars in this same observing run, we can estimate that the relativeuncertainty on the calibration factor is generally equal or lessthan 2%, except blueward of 4000 Å, where the curve shows amarked change of slope and the uncertainty increases to about8%.No corrections for di ff erential atmospheric refraction (DAR)have been applied to our data. For the object observed at thehighest airmass (Tololo 1434 + ≃ . ff eren-tial shift between the bluest and the reddest wavelength in ourspectral range, measured on our data, is less than 1 arcsec (theshift between the H α and the H β lines is about 0.3 arcsec). Giventhe PMAS spaxel size of 1 square arcsec, the seeing ≥ . ′′
2, andthe fact that the diagnostic line ratios we compute involve emis-sion lines very close to each other in wavelength, we can safelyignore DAR e ff ects. In order to measure the relevant parameters of the emission lines(position, flux and width), they were fitted by a single Gaussian.The fit was carried out by the chi square minimization algorithmimplemented by C. B. Markwardt in the mpfitexpr
IDL library .The H β line, where an absorption component was present, wasfitted by two Gaussians.The continuum (typically 30–50 Å on both sides) was fittedby a straight line. Lines in a doublet were fitted imposing thatthey have the same redshift and width.Criteria like flux, error on flux, velocity and width were usedto do a first automatic assessment of whether to accept or rejecta fit. For instance, lines with too small (less than the instrumen-tal width) or too large widths were flagged as rejected, as wellas lines with a relative error on the flux of more than about 10%(the exact limits depend on the specific line and on the overallquality of the spectrum). These criteria were complemented by avisual inspection of all fits, which led to override in a few casesthe automated criteria decision (either accept a fit flagged as re-jected, or viceversa). Emission-line maps were constructed in the following way: theemission-line fit-procedure gives for each line and for each lineparameter (for instance flux) a table with the fiber ID number,the measured value and the acceptance / rejection flag. This tablewas then used to produce a 2D map, by using an IRAF scriptthat takes advantage of the fact that PMAS fibers are arranged ina regular 16x16 matrix. The script automatically converts ADUcounts into flux (erg s − cm − ) by multiplying by a wavelengthdependent conversion factor computed by using the sensitivitycurve described above. URL: http: // cow.physics.wisc.edu / ∼ craigm / idl / idl.html Continuum maps were obtained by summing the flux withinspecific wavelength intervals, selected so as to avoid emissionlines or strong residuals from the sky spectrum subtraction.Line ratio maps were simply derived by dividing the corre-sponding flux maps.
3. Results
IFS data provide within the field of view (FOV) of the instrumenta simultaneous mapping of the galaxy emission in a broad wave-length range. Therefore we can retrieve monochromatic maps atspecific wavelengths or co-added maps equivalent to broad- ornarrow-band images. We constructed emission-line intensity andcontinuum maps for the observed galaxies. Results are shown inFigs. 2–9.
To study the properties of the stellar component, we built contin-uum maps within selected wavelength intervals free from emis-sion lines (“pure continua”). Figs. 2–9 show the “blue” and “red”continuum maps for the sample galaxies, obtained by integrat-ing the spectrum in the regions 4500–4700 and 6000–6200 Årespectively.All objects except two show an overall regular morphol-ogy in the continuum, with a well defined central peak androughly circular isophotes. The exceptions are Mrk 750, whoseouter isophotes are elongated in the northeast direction, andTololo 1434 + In order to study the ionized gas morphology of the sam-ple galaxies we built continuum-subtracted emission line fluxmaps for the most prominent emission lines: [O ii ] λ β ,[O iii ] λ α , [N ii ] λ ii ] λλ , ff erentemission lines, as expected in objects ionized by stars. Six of theeight galaxies, namely Mrk 407, Mrk 750, Mrk 206, Mrk 475,I Zw 123 and I Zw 159, appear very compact in emission linesand have a single central starburst. In Mrk 32 the starburst isresolved into three smaller SF regions, aligned along a north-south axis, while in Tololo 1434 +
032 the current star formationactivity spreads all over the mapped region.
To investigate the excitation mechanisms acting in the galax-ies we computed the line ratio maps for [O iii ] λ / H β ,[N ii ] λ / H α , [S ii ] λλ , / H α and, in those cases inwhich the [O i ] λ i ] λ / H α .High excitation values correspond to high values of the[O iii ] / H β ratios and to low values of [N ii ] / H α and [S ii ] / H α .High values for [O iii ] / H β are expected when the ionization isproduced predominantly by UV photons, especially when theionization parameter is high. On the other hand, low excitationvalues can be associated with an ionizing mechanism di ff erentfrom photoionization (Veilleux & Osterbrock 1987). L. M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies:
Excitation maps for the galaxy sample are displayed inFigs. 2–9. All objects show the same pattern in the di ff erent ra-tio maps: they trace the regions of star formation, with [O iii ] / H β ([N ii ] / H α , [S ii ] / H α ) peaking (having a minimum) in the SFknots, and decreasing (increasing) with the distance to the centerof the region. This is the expected behavior in regions ionized byUV photons coming from massive stars: the metal-to-hydrogenline ratios change as a function of the ionization parameter ( U ),and therefore increasing the distance from the ionization sourcedecreases the value of U , lowering the [O iii ] / H β ratio and in-creasing the [S ii ] / H α ratio (Domgorgen & Mathis 1994). Interstellar extinction can be probed by comparing the observedratios of hydrogen recombination lines with their theoretical val-ues (Osterbrock & Ferland 2006). In the optical domain the ex-tinction is derived from the ratio of the di ff erent Balmer line se-ries to H β . In our case, although the H δ and H γ emission linesare usually visible, their weakness and that they are superposedon a strong underlying stellar absorption line make it impossi-ble to obtain reliable measurements of their flux for individualspaxels. Therefore the extinction maps have been derived fromthe H α / H β ratio.For T =
10 000 K and electron densities ∼
100 cm − , the the-oretical H α / H β ratio should be close to 2.86. Because extinctionis stronger at H β that at H α wavelengths, its e ff ect is to increasethe observed ratio.In computing the extinction map, we corrected the H β linefor underlying stellar absorption by fitting a Gaussian profileto its absorption wings. For the H α line, the absence of visi-ble absorption wings makes this decomposition impossible. Toaccount for the H α absorption component, several approachescould be adopted in principle. The most popular one is to set theequivalent width of the H α absorption, W (H α ) abs , to the samevalue as found for H β . A second more conservative approachis to set it to some fixed value (for instance 2 Å) or simply toassume that the absorption in H α is negligible.While the first strategy might be appropriate when dealingwith with integrated spectra, the relatively high uncertainties onthe measurements of W (H β ) abs makes it unfeasible for a spaxel-to-spaxel correction, thus we adopted here the more conservativeapproach (setting W (H α ) abs = α / H β line ratio maps of the sample galaxies are displayedin Figs. 2–9. Six out of the eight objects have significant inter-stellar extinction values — up to E ( B − V ) = . E ( B − V ) = . C (H β )) — and a patchy dust distri-bution. We also produced maps of the [S ii ] λ / [S ii ] λ − . The density maps of the sample galaxies are dis-played in Figs. 2–9. We studied the kinematics of the ionized gas using the[O iii ] λ α emission lines. We fitted the peak wave-length of the above emission lines with a Gaussian to obtain theradial velocity of the ionized gas at each spaxel. Due to the low spectral resolution of our data (about 6.8 Å FWHM) no reliablevelocity dispersion measurements could be obtained.The velocity fields are shown in Figs. 2–9; in these maps redcolors represent redshifts and blue colors blueshifts. The meanuncertainty in the velocity data, estimated from the scatter ofvelocity maps built on the 5577 Å skyline (which in our spectrahas intensities comparable with the H α and [O iii ] λ − in the centralregions and increases outwards with decreasing line emissionintensity. In this section we present results of the integrated spectroscopy.For each galaxy we extracted a one-dimensional spectrum of theSF regions in our maps and the integrated spectrum inside thewhole mapped FOV. Six galaxies have just one nuclear SF re-gion; only in Mrk 32 and Tololo 1434 +
032 we identified two ormore SF regions.Because of the heterogeneity of the sample in terms of ap-parent luminosity, distance, morphology, brightness of the SFregions, and signal-to-noise ratio of the observed spectra, thereis no unique, clear-cut criterion for delineating the area of the SFregions. Thus we followed a more pragmatic approach, based onthe specific morphology of the SF knots: we integrated within aboundary that follows the shape of the SF knot with a minimumarea of ∼
20 square arcsec (that is an equivalent radius of abouttwice the seeing FHWM). For galaxies with multiple SF knots,the above criterion was relaxed; see the discussion on each indi-vidual object (Sect. 4). The SF knots are outlined and labeled inthe H α maps in Figs. 2–9.As for the integrated spectrum of a galaxy within the PMASFOV, in order to not degrade its signal-to-noise ratio unneces-sarily by including outer spaxels with no gaseous emission, weonly summed those spaxels with measured H α emission (that isthose shown in the H α flux maps).Figs. 10 and 11 display the spectra of the sample galaxies.In general all spectra are dominated by bright, narrow emissionlines, indicating an important contribution of ionizing stars, butwith significant di ff erences among the objects: several galaxiesdisplay a very high continuum with strong absorption features,whereas some of them show a flat spectrum, characteristic ofan OB population. A more detailed description of the spectralcharacteristics of the individual objects is provided in Sect. 4. The higher S / N ratio of the integrated spectra allows us on onehand a more accurate measurement of the Balmer line fluxes,and on the other a more careful and reliable determination of theextinction coe ffi cient.For each spectrum we measured fluxes and equivalent widthsof the emission lines using the Gaussian profile fitting option inthe IRAF task splot . In order to obtain reliable values of Balmerfluxes in emission we must take into account the underlying stel-lar absorption (McCall et al. 1985; Diaz 1988). To do that, wefollowed two di ff erent approaches, depending on the character-istics of the spectra.When the absorption wings around the Balmer lines were notvisible, we assumed that the equivalent width in absorption isthe same for all the lines. We first adopted an initial estimate forthe absorption equivalent width, EW abs , corrected the measuredfluxes, and computed the extinction coe ffi cient C (H β ) through . M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies: 7 a least-square fit to the Balmer decrement. We then varied thevalue of EW abs , until we found the one that provided the bestmatch (e.g. the minimum scatter in the fit) between the correctedand the theoretical line ratios. A more detailed description ofthis method can be found in Izotov et al. (1994) and Cair´os et al.(2007).We adopted the “case B” Balmer recombination decrementfor T e =
10 000 K and N e = cm − (Brocklehurst 1971) andthe Cardelli et al. (1989) reddening curve.In those cases where absorption wings around the Balmerlines were visible, we simultaneously fitted an absorption andan emission component. We then applied the same method asbefore, varying the equivalent width in absorption only in thoselines in which it could not be fitted.In several cases the values of the H δ and H γ fluxes aredoubtful due to their intrinsic weakness and the large uncer-tainties in the correction for the underlying stellar absorption.In these cases we computed C (H β ) directly from the H α / H β ra-tio: (i) if the absorption in H β was fitted, we set EW(H α ) abs = EW(H β ) abs ; (ii) if not, following McCall et al. (1985) we set theequivalent width of both lines in absorption to 2 Å.Reddening-corrected intensity ratios and equivalent widthsfor the di ff erent spatial regions are listed in Tables 3–6. Physical properties and ionic abundances were derived fromthe reddening-corrected emission line fluxes, following the5-level atom fivel program in the IRAF nebular package(De Robertis et al. 1987; Shaw & Dufour 1995).Electron densities were measured from the emission line ra-tio [S ii ] λ / λ T e [O iii ]) were de-rived from the [O iii ] λ / ( λ + λ iii ] λ / N. Inthese cases, T e [O ii ] was calculated from the relation between T e [O ii ] and T e [O iii ] provided in Pilyugin et al. (2006).We then adopted T e [O ii ] for the calculation of N + , O + andS + abundances, and T e [O iii ] for the calculation of O + andNe + abundances. We used [Ne iii ] λ ii ] λ iii ] λλ , ii ] λλ , ii ] λλ , / O = Ne ++ / O ++ andN / O = N + / O + , respectively, and the total oxygen abundance wascalculated as O / H = (O + / H + + O ++ / H + ) .To obtain oxygen abundances in those knots in which[O iii ] λ
4. Discussion
The galaxy Mrk 407 is included in the Petrosian et al. (2007)Atlas of Markarian galaxies, where it appears classified as an S0. J , H , and K NIR surface brightness photometry was published inCair´os et al. (2003), but the quality of these data was insu ffi cientfor assessing whether or not this galaxy has an older stellar lowsurface brightness component underlying the SF regions. The only spectroscopic data available in the literature are its redshiftand the equivalent widths and flux ratios of the strongest lines(Ugryumov et al. 1998).The PMAS FOV covers an area of 2 . × . ii regions(Veilleux & Osterbrock 1987). The [O iii ] / H β map also peakssoutheast of the central SF region, at the same spatial locationas the extinction maximum.The galaxy has an inhomogeneous extinction pattern, with adust lane crossing it southwest–northeast, where H α / H β reachesvalues of up to 5.5 — which translates into an interstellar red-dening E ( B − V ) of up to 0.6. In the rest of the galaxy, the valuesof the H α / H β ratio are closer to the theoretical value of 2.86.Both the nuclear and the integrated spectra of Mrk 407 dis-play strong emission lines atop a blue continuum, with severalabsorption features (high order Balmer lines in absorption andpronounced absorption wings in H δ , H γ and H β ; see Fig. 10);the absorption features are indicative of a substantial contribu-tion from older stars. As we could not detect [O iii ] λ . Z ⊙ , which places Mrk 407 in the highmetallicity BCD group.While the H α and [O iii ] velocity maps are quite noisy, theyboth seem to indicate an overall rotation around an axis roughlyoriented on the northwest-southeast direction, with a velocityamplitude of about 30–40 km s − . With M B = − .
99, Mrk 32 is one of the faintest galaxiesin our sample. In the Petrosian et al. (2007) Atlas it is clas-sified as an Im / BCD. Broad-band imaging and optical spec-troscopy were published in Hunter & Elmegreen (2006) andHunter & Ho ff man (1999), respectively.The PMAS FOV covers an area of 1 . × .
27 kpc with asampling of 80 pc per spaxel. Two-dimensional maps of Mrk 32are shown in Fig. 3.In the continuum the galaxy shows an overall regular be-havior, with elliptical isophotes, much elongated in the north-south direction. Although the intensity distribution clearly in-creases towards the central regions, there is not a clear peak, buta somehow di ff use maximum, whose position seems to be dis-placed to the west when moving towards red wavelengths. Allthe emission-line maps display a similar pattern, with the SFknots roughly aligned in the south-north direction, and all, ex-cept [O ii ], peak at the northeast knot (A, as labeled in Fig. 3);none of the three SF knots seen in emission-lines coincides withthe central intensity peak in the continuum maps.The excitation maps trace the three SF regions; the max-imum in [O iii ] / H β (the minimum in the other maps) is lo-cated in the brightest knot A. Line ratio maps display valuestypical of H ii regions, aside from the southern regions whose L. M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies: higher values for [O i ] / H α and [S ii ] / H α are more characteristicof LINERS / Seyfert galaxies (log[O i ] / H α ≥ − .
3, log[S ii ] / H α ≥ − .
4; Veilleux & Osterbrock 1987); this suggests that anothermechanism, most probably shocks, is contributing to the gas ex-citation.The H α / H β ratio map displays a noisy pattern, with most ofthe spaxels showing values close to 2.86; however, several spax-els located in the outer regions and close to knot C have highervalues — consistent with the higher reddening values derived forthe integrated spectrum of knot C.We produced the integrated spectrum of knot AB (because oftheir closeness, knots A and B were lumped together althoughthey seem to be two separate SF regions) and C (whose smallsize is dictated by the need of keeping it separated from AB),and of the integrated galaxy spectrum. All three spectra are dom-inated by young stars, with strong emission lines atop a bluecontinuum; in knot C, and in the integrated spectrum, absorptionwings around Balmer lines are detected. We also have a marginaldetection of the Wolf-Rayet (WR) bump at λ / N ratio of the WRbump is too low for a reliable measurement of its flux.We found a low interstellar extinction, E ( B − V ) ≤ . E ( B − V ) = .
48 value reported byHunter & Ho ff man (1999).For knot AB (the only one where we could measure the[O iii ] λ + log(O / H) = .
56 (about 1 / Z ⊙ ), which places Mrk 32 amongthe extremely metal-deficient BCDs — galaxies having 12 + log(O / H) ≤ . T e -based abundances(see for instance Shi et al. 2005, where discrepancies on the or-der of 0.45 dex have been found). However, this considerabledisagreement may also arise from the relatively large measure-ment uncertainties for the faint [O iii ] λ / N ratios are systematically biased towardshigher values). Deep spectroscopic observations are required toderive accurate T e -based abundances.The velocity field appears essentially flat, with perhaps a hintof a small increase to the southwest. The galaxy Mrk 750 is an extremely faint ( M B = − . / Im in Petrosian et al. (2007). It be-longs to the sample of low metallicity galaxies (Izotov & Thuan1999; Izotov et al. 2007), and it is also a well-known Wolf-Rayet galaxy: Kunth & Joubert (1985) first pointed out the broadHe ii λ iii λ iv λ α observationswere published by M´endez & Esteban (2000). From high spa-tial resolution H i synthesis observations, van Zee et al. (2001)found that the neutral gas extends to approximately twice theoptical diameter of the galaxy, peaking in the central region ofstar formation.Our IFU data cover a region of 400 ×
400 pc, with a sam-pling of 25 pc per spaxel. Two-dimensional maps of Mrk 750are shown in Fig. 4. The galaxy has a similar morphology both in the emissionline and in the continuum maps, with roughly circular isophotesand a single central peak, whose position is the same in all maps(emission lines and continuum). However, while in the emissionline maps the isophotes are circular at all intensity levels, in thecontinuum maps the outer ones appear elongated to the north-east, indicating that the central, bigger knot of Mrk 750 is con-necting with a continuum source. This peculiar morphology wasinterpreted as an interaction sign in M´endez & Esteban (2000).The excitation maps all display the same pattern and all tracethe central SF knot. The ratio values in the four maps are consis-tent with ionization by hot stars.In the extinction map we clearly distinguish a dust patchdominating the southeast part of the galaxy, with H α / H β valuesup to 3.8 — E ( B − V ) up to 0.27.The integrated and nuclear spectra are both blue, OB youngstars dominated spectra with very prominent emission lines; nei-ther Balmer absorption wings nor absorption lines are visible.In agreement with previous results, we detected the blue WolfRayet bump around 4640Å (see Fig. 10); this feature is promi-nent in ten spaxels, located all in the nuclear region of the galaxy.We also have a marginal detection of C iv λ iii ] flux mapwith crosses and squares for the blue and the red bumps respec-tively (see Fig. 4).We computed the fluxes and equivalent width of the blueWR bump in the nuclear spectrum by a plain integration ofthe signal in the corresponding spectral region after fitting andsubtracting the underlying continuum. Because the WR bumpsare a blend of both genuine WR features and nebular lines, ourmeasurement is actually an upper limit (the S / N ratio and theresolution of the spectrum do not allow us to fit and subtractseparately the nebular lines). The extinction-corrected flux is F (WR blue ) ∼ × − ergs cm − s − , and the equivalent widthis EW(WR blue ) ∼ T e based method, 12 + log(O / H) = . ± .
03, excellentlyagrees with the value 12 + log(O / H) = . ± .
02 reported byIzotov & Thuan (1998). There is also good agreement with thevalues we derived with empirical calibrations.Both the H α and [O iii ] velocity maps, while somewhatnoisy, display a rotation pattern around an axis roughly ori-ented east-west, with an amplitude of about 20 km s − ; this re-sult qualitatively agrees with the H i kinematics published byvan Zee et al. (2001). The galaxy Mrk 206 is classified as a BCD in Petrosian et al.(2007). Intensities and equivalent widths of hydrogen and oxy-gen emission lines as well as oxygen abundances are publishedin Kniazev et al. (2004).Our IFU data cover a region of 1 . × . . M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies: 9 The extinction pattern on the other hand is highly inhomoge-neous, with a substantial amount of dust located on the northeastregion of the galaxy, with H α / H β ratios up to 6 — E ( B − V ) upto 0.7.The galaxy displays a flat, typical H ii region spectrum,with no evident absorption features. With a metallicity of about0 . Z ⊙ it is the galaxy with the highest metallicity in our sample.However, as we fail to detect [O iii ] λ T e based method abundance of 12 + log(O / H) = . ≈ . α / H β map. The northeast side isapproaching, the southwest side is receding; the velocity ampli-tude inside the mapped region is about 50–60 km s − . Tololo 1434 +
032 is a low-metallicity BCD (Izotov et al. 2007;Kniazev et al. 2004). B and R broad-band photometry was pub-lished in Doublier et al. (1997) and Gil de Paz et al. (2003), andH α imaging in Gil de Paz et al. (2003).The PMAS FOV covers a region of 2 . × . +
032 are shown in Fig. 6.The galaxy shows a clumpy morphology both in emissionlines and in the continuum. In the continuum maps we resolvetwo major emission peaks. The strongest is located in the south-east, while the other is displaced about 9 arcsec (1.3 kpc) to thenorthwest. All emission lines maps have the same structure: fourSF knots are distributed in a roughly circular pattern, with thepeak of emission located in knot A, whose position coincideswith the continuum peak (knots are labeled in the H α map inFig. 6). Northeast of A there is a smaller knot, B, while twofainter knots, C and D, are seen in the north side: neither coin-cides spatially with the secondary continuum peak. (Because oftheir small size and luminosity, C and D were lumped togetherto obtain their integrated spectrum.)The excitation maps trace the regions of star formation anddisplay in the whole field of view values typical of H ii regions.The galaxy shows an homogeneous extinction pattern withvalues close to the theoretical value of 2.86 across the wholeFOV.The spectra of the resolved SF knots are very similar and alsoresemble well the integrated spectrum: they are all flat, youngstar-dominated spectra with no evidence of absorption features.The oxygen abundance we found for the brightest knot, usingthe T e method, 12 + log(O / H) = . ± .
07, compares wellwith the value derived in Kniazev et al. (2004), 12 + log(O / H) = . ± .
04. The values derived through empirical calibrationsare also very similar.Both the H α and the [O iii ] velocity maps seem to marginallyindicate a low amplitude rotation ( .
20 km s − ) around anortheast-southwest axis. The galaxy Mrk 475 is a low luminosity object, includedin Izotov’s sample of metal poor galaxies (Izotov et al. 1994,2007), and is classified as a BCD in Petrosian et al. (2007). B and R broad-band surface brightness photometry and H α imag-ing were published by Gil de Paz et al. (2003). It is a WR galaxy, where nebular and broad He ii λ iv λ ×
920 pc, with a spatialsampling of 58 pc per spaxel. Two-dimensional maps of Mrk 475are displayed in Fig. 7.This is a compact, regular object, with a single central SFknot. Emission lines and continuum maps display all the samemorphology.The ionization maps show all the same complex pattern:while in the western galaxy regions the excitation ratio decreaseswith the distance from the central SF region, in the eastern partit displays a constant value. The four ratio maps show valuestypical of excitation by hot stars.The extinction map is inhomogeneous and has a peak in thesouthwest; extinction values are moderate, with H α / H β peakingaround 3.The galaxy displays a blue spectrum, with prominent emis-sion lines and no visible absorption features. We detected theblue WR bump in eight spaxels (marked with crosses in Fig. 7);in two of them the red WR bump is also visible (squares inFig. 7). We measured the flux and equivalent width of bothWR bumps in the nuclear spectrum in the same way we did forMrk 750, finding fluxes of ∼ × − and ∼ × − , andequivalent widths of 15 and 4 Å for the blue and the red bumprespectively.The oxygen abundance we found for the nuclear region, us-ing the T e method, 12 + log(O / H) = . ± .
02, agrees well withthe value of 7 . ± .
04 reported by Izotov et al. (1994). Valuesderived using the empirical calibrations are slightly higher, butstill in good agreement (di ff erence ≤ .
15 dex).Both the H α and the [O iii ] kinematical maps show a flat ve-locity field. The galaxy I Zw 123 is classified as a BCD in Petrosian et al.(2007) and belongs also to the metal-poor BCD class(Izotov et al. 1997; Izotov & Thuan 1999). I Zw 123 is a rela-tively well-studied object: optical surface photometry was pub-lished in several papers (Cair´os et al. 2001a,b; Gil de Paz et al.2003; Caon et al. 2005; Amor´ın et al. 2007), NIR photometry inCair´os et al. (2003). A thorough spectroscopic study, includingan analysis of its stellar content in terms of population synthesismodels, was published in the series of papers by Kong & Cheng(2002), Kong et al. (2002), Kong et al. (2003), Kong (2004) andShi et al. (2005).The PMAS FOV covers an area of 1 . × . ii ] / H α map, where the outer regions haverelatively high values (log[S ii ] / H α ≥ − . α / H β ratio im-plies an E ( B − V ) of about 0.3 mag. The galaxy exhibits a blue spectrum, with prominent emis-sion lines and no evident absorption features. The oxygen abun-dance that we find by applying the T e method, 12 + log(O / H) = . ± .
03 agrees well with the value reported in Izotov & Thuan(1999), 12 + log(O / H) = . ± . α and [O III] velocity maps the velocity fieldappears flat (the higher velocities seen in the outermost spaxelsin the [O III] map are most likely due to noise). In Petrosian et al. (2007) I Zw 159 is classified as a BCD / Irr.Optical surface brightness photometry was published inDoublier et al. (1997), Doublier et al. (1999), Gil de Paz et al.(2003) and Gil de Paz & Madore (2005). It is also includedin the spectroscopy study of BCDs by Kong & Cheng (2002),Kong et al. (2002), Kong et al. (2003), Kong (2004) andShi et al. (2005), and is the only object in our sample in com-mon with the sample of galaxies studied by means of IFS byPetrosian et al. (2002).The PMAS FOV covers a region of 3 . × . iii ] λ / H β and [N ii ] λ / H α maps the values are con-sistent with photoionization by young stars in the whole galaxy,in the [S ii ] / H α and [O i ] / H α maps the outer regions of the galaxyhave higher values (log[O i ] / H α ≥ − .
0, log[S ii ] / H α ≥ − . E ( B − V ) is as highas 0.45 mag.The integrated spectra are blue, with some absorption wingsaround the Balmer lines. With an oxygen abundance 12 + log(O / H) ≃ .
3, this galaxy also belongs to the high-metallicityBCDs branch.The velocity field in the H α and [O iii ] maps shows a clearoverall rotation along an axis roughly oriented south-north, withan amplitude of about 70–80 km s − , in broad agreement withthe velocity field published by Petrosian et al. (2002).
5. Summary and conclusions
We present here what is to our knowledge the most extensiveIFS analysis of a sample of BCDs. This study is based on PMASdata, which cover a wavelength range of 3590-6996 Å, with alinear dispersion of 3.2 Åper pixel, and map an area 16 ′′ × ′′ with a spatial sampling of 1 ′′ × ′′ .For all the sample galaxies we produced an atlas of two-dimensional maps: two continuum bands, the brightest emis-sion lines (i.e. [O ii ] λ β , [O iii ] λ i ] λ α ,[N ii ] λ ii ] λλ , iii ] / H β , [O i ] / H α , [N ii ] / H α , [S ii ] / H α and H α / H β )as well as the velocity field of the ionized gas. Integrated spec-troscopic properties of the most prominent SF regions and of thewhole galaxy have been also derived.From this work we highlight the following results: 1. All the objects except Mrk 750 and Tololo 1434 +
032 ex-hibit a mostly regular morphology in the continuum, withone (or several for Mrk 32 and Tololo 1434 + +
032 displays a clumpy continuum morphology.All the galaxies show a similar morphology in the di ff erentmapped emission lines, as expected for objects ionized byhot stars, and for most of the galaxies the emission line mor-phology traces also the stellar component. Only for Mrk 32and Tololo 1434 +
032 we found that the distribution of thegaseous emission di ff ers considerably from that of the stel-lar component. Spatial discrepancies in the distribution ofemission lines and continuum are interpreted as signs of aspatial migration of the SF over the history of the galax-ies (Petrosian et al. 2002). However, small spatial o ff sets be-tween continuum and emission line peaks, as those seenin Tololo 1434 + ff erent excitation maps produced for the same galaxiesdisplay a similar pattern and trace the regions of star forma-tion as expected in objects ionized by hot stars. In three outof the eight sample galaxies, namely Mrk 32, I Zw 123 andI Zw 159, higher values of [S ii ] / H α in the outer galaxy re-gions suggest shocks.3. Six out of the eight objects display inhomogeneous extinc-tion maps, with interstellar reddening values E ( B − V ) vary-ing across the galaxy from ≤ . ff erent regions of the galaxy.4. All SF regions in the sample galaxies have low electron den-sities, ranging from ≤
100 to 320 cm − , typical of classicalH ii regions.5. The oxygen abundances in the present objects range from12 + log(O / H) = .
56 to 8.44 ( Z = / Z ⊙ to Z = . Z ⊙ ).We measured for the first time the oxygen abundances ofMrk 407 and Mrk 32. The galaxy Mrk 407 is found to bea relatively high metallicity BCD, while the oxygen abun-dance found for Mrk 32 from the [O iii ] λ + log(O / H) ≤ .
6, are excel-lent laboratories for galaxy formation and evolution stud-ies, as they allow us to study chemical compositions andstellar populations in conditions approaching those of dis-tant protogalactic systems. However, they are also very dif-ficult to find, and at the present time only about 30 ex-tremely metal-deficient BCDs are known (Kunth & ¨Ostlin2000; Kniazev et al. 2004; Papaderos et al. 2008).6. Wolf-Rayet features were measured in three out of the eightgalaxies; a marginal detection was reported for Mrk 32.7. Three galaxies display a clear rotation pattern (Mrk 750,Mrk 206, I Zw 159); for Mrk 407 and Tololo 1434 + . M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies: 11 This paper is part of a larger project that aims to map of theproperties of an externsive and representative sample of BCDsby means of IFS. Results for five luminous BCDs were pub-lished in Garc´ıa-Lorenzo et al. (2008), and results for the galax-ies Mrk 409 and Mrk 1418, also observed with PMAS, have beenshown in Cair´os et al. (2009a) and Cair´os et al. (2009b) respec-tively. The global properties of the whole sample will be dis-cussed in a forthcoming publication.
Acknowledgements.
L. M. Cair´os and C. Kehrig acknowledge the Alexandervon Humboldt Foundation. N. Caon and C. Zurita are grateful for thehospitality of the Astrophysikalisches Institut Potsdam. This research hasmade use of the NASA / IPAC Extragalactic Database (NED), which is op-erated by the Jet Propulsion Laboratory, Caltech, under contract with theNational Aeronautics and Space Administration. We acknowledge the us-age of the HyperLeda database (http: // leda.univ-lyon1.fr). This work hasbeen partially funded by the spanish “Ministerio de Ciencia y Innovaci´on”through grants AYA 2007 67965 and HA2006-0032, and under the Consolider-Ingenio 2010 Program grant CSD2006-00070: First Science with the GTC(http: // / consolider-ingenio-gtc / ). References
Amor´ın, R. O., Mu˜noz-Tu˜n´on, C., Aguerri, J. A. L., Cair´os, L. M., & Caon, N.2007, A&A, 467, 541Becker, T. 2002, PhD thesisBergvall, N. & ¨Ostlin, G. 2002, A&A, 390, 891Brocklehurst, M. 1971, MNRAS, 153, 471Cair´os, L. M., Caon, N., Garc´ıa-Lorenzo, B., et al. 2007, ApJ, 669, 251Cair´os, L. M., Caon, N., Garc´ıa-Lorenzo, B., V´ılchez, J. M., & Mu˜noz-Tu˜n´on,C. 2002, ApJ, 577, 164Cair´os, L. M., Caon, N., Papaderos, P., et al. 2009a, ApJ, 707, 1676Cair´os, L. M., Caon, N., Papaderos, P., et al. 2003, ApJ, 593, 312Cair´os, L. M., Caon, N., V´ılchez, J. M., Gonz´alez-P´erez, J. N., & Mu˜noz-Tu˜n´on,C. 2001a, ApJS, 136, 393Cair´os, L. M., Caon, N., Zurita, C., et al. 2009b, A&A, 507, 1291Cair´os, L. M., V´ılchez, J. M., Gonz´alez P´erez, J. N., Iglesias-P´aramo, J., & Caon,N. 2001b, ApJS, 133, 321Caon, N., Cair´os, L. M., Aguerri, J. A. L., & Mu˜noz-Tu˜n´on, C. 2005, ApJS, 157,218Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ, 345, 245Conti, P. S. 1991, ApJ, 377, 115De Robertis, M. M., Dufour, R. J., & Hunt, R. W. 1987, JRASC, 81, 195Diaz, A. I. 1988, MNRAS, 231, 57Domgorgen, H. & Mathis, J. S. 1994, ApJ, 428, 647Doublier, V., Caulet, A., & Comte, G. 1999, A&AS, 138, 213Doublier, V., Comte, G., Petrosian, A., Surace, C., & Turatto, M. 1997, A&AS,124, 405Garc´ıa-Lorenzo, B., Cair´os, L. M., Caon, N., Monreal-Ibero, A., & Kehrig, C.2008, ApJ, 677, 201Gil de Paz, A. & Madore, B. F. 2005, ApJS, 156, 345Gil de Paz, A., Madore, B. F., & Pevunova, O. 2003, ApJS, 147, 29Guseva, N. G., Izotov, Y. I., & Thuan, T. X. 2000, ApJ, 531, 776Guseva, N. G., Papaderos, P., Izotov, Y. I., et al. 2003a, A&A, 407, 75Guseva, N. G., Papaderos, P., Izotov, Y. I., et al. 2003b, A&A, 407, 91Guseva, N. G., Papaderos, P., Izotov, Y. I., et al. 2003c, A&A, 407, 105Hunt, L. K., Vanzi, L., & Thuan, T. X. 2001, A&A, 377, 66Hunter, D. A. & Elmegreen, B. G. 2006, ApJS, 162, 49Hunter, D. A. & Ho ff man, L. 1999, AJ, 117, 2789Hunter, D. A. & Thronson, Jr., H. A. 1995, ApJ, 452, 238Izotov, Y. I., Schaerer, D., Blecha, A., et al. 2006, A&A, 459, 71Izotov, Y. I. & Thuan, T. X. 1998, ApJ, 500, 188Izotov, Y. I. & Thuan, T. X. 1999, ApJ, 511, 639Izotov, Y. I., Thuan, T. X., & Lipovetsky, V. A. 1994, ApJ, 435, 647Izotov, Y. I., Thuan, T. X., & Lipovetsky, V. A. 1997, ApJS, 108, 1Izotov, Y. I., Thuan, T. X., & Stasi´nska, G. 2007, ApJ, 662, 15James, B. L., Tsamis, Y. G., Barlow, M. J., et al. 2009, MNRAS, 398, 2Kau ff mann, G., White, S. D. M., & Guiderdoni, B. 1993, MNRAS, 264, 201Kehrig, C., V´ılchez, J. M., S´anchez, S. F., et al. 2008, A&A, 477, 813Kelz, A., Verheijen, M. A. W., Roth, M. M., et al. 2006, PASP, 118, 129Kniazev, A. Y., Pustilnik, S. A., Grebel, E. K., Lee, H., & Pramskij, A. G. 2004,ApJS, 153, 429Kobulnicky, H. A., Kennicutt, Jr., R. C., & Pizagno, J. L. 1999, ApJ, 514, 544Kong, X. 2004, A&A, 425, 417Kong, X., Charlot, S., Weiss, A., & Cheng, F. Z. 2003, A&A, 403, 877 Kong, X. & Cheng, F. Z. 2002, A&A, 389, 845Kong, X., Cheng, F. Z., Weiss, A., & Charlot, S. 2002, A&A, 396, 503Kunth, D. & Joubert, M. 1985, A&A, 142, 411Kunth, D., Maurogordato, S., & Vigroux, L. 1988, A&A, 204, 10Kunth, D. & ¨Ostlin, G. 2000, A&A Rev., 10, 1Lagos, P., Telles, E., & Melnick, J. 2007, A&A, 476, 89Lagos, P., Telles, E., Mu˜noz-Tu˜n´on, C., et al. 2009, AJ, 137, 5068Lequeux, J. & Viallefond, F. 1980, A&A, 91, 269Loose, H.-H. & Thuan, T. X. 1986, in Star-forming Dwarf Galaxies and RelatedObjects, 73–88Lowenthal, J. D., Koo, D. C., Guzman, R., et al. 1997, ApJ, 481, 673Maiz-Apellaniz, J., Mas-Hesse, J. M., Munoz-Tunon, C., Vilchez, J. M., &Castaneda, H. O. 1998, A&A, 329, 409Mas-Hesse, J. M. & Kunth, D. 1999, A&A, 349, 765Masegosa, J., Moles, M., & Campos-Aguilar, A. 1994, ApJ, 420, 576McCall, M. L., Rybski, P. M., & Shields, G. A. 1985, ApJS, 57, 1M´endez, D. I. & Esteban, C. 2000, A&A, 359, 493Osterbrock, D. E. & Ferland, G. J. 2006, Astrophysics of gaseous nebulae andactive galactic nuclei (Astrophysics of gaseous nebulae and active galactic nu-clei, 2nd. ed. by D.E. Osterbrock and G.J. Ferland. Sausalito, CA: UniversityScience Books, 2006)Pagel, B. E. J., Simonson, E. A., Terlevich, R. J., & Edmunds, M. G. 1992,MNRAS, 255, 325Papaderos, P., Guseva, N. G., Izotov, Y. I., & Fricke, K. J. 2008, A&A, 491, 113Papaderos, P., Loose, H.-H., Thuan, T. X., & Fricke, K. J. 1996, A&AS, 120,207Patat, F. 2003, A&A, 400, 1183Paturel, G., Petit, C., Prugniel, P., et al. 2003, A&A, 412, 45Petrosian, A., McLean, B., Allen, R. J., & MacKenty, J. W. 2007, ApJS, 170, 33Petrosian, A. R., Movsessian, T., Comte, G., Kunth, D., & Dodonov, S. 2002,A&A, 391, 487Pettini, M. & Pagel, B. E. J. 2004, MNRAS, 348, L59Pilyugin, L. S., V´ılchez, J. M., & Thuan, T. X. 2006, MNRAS, 370, 1928Roth, M. M., Kelz, A., Fechner, T., et al. 2005, PASP, 117, 620S´anchez, S. F. 2006, Astronomische Nachrichten, 327, 850S´anchez, S. F., Aceituno, J., Thiele, U., P´erez-Ram´ırez, D., & Alves, J. 2007,PASP, 119, 1186Sargent, W. L. W. & Searle, L. 1970, ApJ, 162, L155 + Schaerer, D., Contini, T., & Pindao, M. 1999, A&AS, 136, 35Shaw, R. A. & Dufour, R. J. 1995, PASP, 107, 896Shi, F., Kong, X., Li, C., & Cheng, F. Z. 2005, A&A, 437, 849Telles, J. E. 1995, PhD thesis, , Univ. Cambridge, (1995)Thuan, T. X. 1991, Observations and Models of Blue Compact Dwarf Galaxies(Massive Stars in Starbursts), 183– + Thuan, T. X. & Martin, G. E. 1981, ApJ, 247, 823Ugryumov, A. V., Pustilnik, S. A., Lipovetsky, V. A., Izotov, Y. I., & Richter, G.1998, A&AS, 131, 295van Zee, L., Salzer, J. J., & Skillman, E. D. 2001, AJ, 122, 121Vanzi, L., Cresci, G., Telles, E., & Melnick, J. 2008, A&A, 486, 393Vanzi, L. & Sauvage, M. 2004, A&A, 415, 509Veilleux, S. & Osterbrock, D. E. 1987, ApJS, 63, 295
Fig. 2.
Two-dimensional maps for Mrk 407. Continuum maps in the “emission line free” intervals 4500–4700 Å (blue) and6000-6200 Å (red); emission line flux maps: [O ii ] λ β , [O iii ] λ α , [N ii ] λ ii ] λλ , iii ] λ / H β , [N ii ] λ / H α , [S ii ] λλ , / H α (ionization ratios), H α / H β (interstellar extinction) and[S ii ] λ / [S ii ] λ α and [O iii ] λ − ergs cm − s − . The outline of the region within which the integrated nuclear spectrum was obtained (see Sect. 3.4) isshown in the H α map. . M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies: 13 A BC
Fig. 3.
Same as Fig. 2 for Mrk 32. The [O i ] λ i ] λ / H α ionization ratio map are also included. Theoutline of the identified SF knots (see Sect. 3.4) is shown in the H α map. Spaxels with a marginal detection of the blue WR bumphave been marked by crosses in the [O iii ] map. Fig. 4.
Same as Fig. 2 for Mrk 750. Maps of [O iii ] λ i ] λ i ] λ / H α ionization ratio are also included.Spaxels in which the WR feature was detected were marked in the [O iii ] λ . M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies: 15 Fig. 5.
Same as Fig. 2 for Mrk 206. Maps of [O i ] λ i ] λ / H α ionization ratio are also included. AB C D
Fig. 6.
Same as Fig. 2 for Tololo 1434 + α map. . M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies: 17 Fig. 7.
Same as Fig. 2 for Mrk 475. Maps of [O i ] λ i ] λ / H α ionization ratio are also included. Spaxels inwhich the WR feature has been detected have been marked in the [O iii ] λ Fig. 8.
Same as Fig. 2 for I Zw 123. Spaxels in which the WR blue bump was detected were marked in the [O iii ] λ . M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies: 19 Fig. 9.
Same as Fig. 2 for I Zw 159. Maps of [O i ] λ i ] λ / H α ionization ratio are also included. Fig. 10.
Nuclear and integrated spectra for Mrk 407, Mrk 750 and Mrk 206; for Mrk 32, spectra of the identified SF regions and theintegrated spectrum. The inset in the Mrk 750 figure shows in detail the blue WR bump region in the nuclear spectrum. Spectra areshown in logarithmic scale and are o ff set for clarity. The interval between large tickmarks is 1 dex (0.05 dex in the inset). . M. Cair´os et al.: Mapping the properties of blue compact dwarf galaxies: 21 Fig. 11.
Nuclear spectra and integrated spectra for Mrk 475, I Zw 123 and I Zw 159; for Tololo 1434 + L . M . C a i r´ o s e t a l . : M a pp i ng t h e p r op e r ti e s o f b l u ec o m p ac t d w a rf g a l a x i e s : Table 3.
Reddening-corrected line ratios, normalized to H β ( ∗ ) for the sample of galaxies λ Ion Mrk 407 Mrk 32Nuclear Integrated Knot AB Knot C Integrated F λ − W λ F λ − W λ F λ − W λ F λ − W λ F λ − W λ ii ] 4 . ± .
54 43 . ± . . ± .
22 49 . ± . . ± .
31 34 . ± . . ± .
56 33 . ± . . ± .
71 26 . ± . iii ] — — — — 0 . ± .
08 6 . ± . + HeI — — — — — — — — — —3968 [Ne iii ] — — — — — — — — — —4101 H δ — — — — 0 . ± .
03 4 . ± . γ . ± .
05 2 . ± . . ± .
04 9 . ± . . ± .
04 4 . ± . . ± .
08 4 . ± . iii ] — — — — 0 . ± .
02 1 . ± . i — — — — — — — — — —4861 H β . ± .
00 10 . ± . . ± .
00 10 . ± . . ± .
00 24 . ± . . ± .
00 11 . ± . . ± .
00 9 . ± . iii ] 1 . ± .
06 12 . ± . . ± .
12 10 . ± . . ± .
04 25 . ± . . ± .
06 11 . ± . . ± .
09 9 . ± . iii ] 3 . ± .
13 35 . ± . . ± .
23 26 . ± . . ± .
11 70 . ± . . ± .
12 33 . ± . . ± .
16 23 . ± . i — — — — 0 . ± .
02 4 . ± . . ± .
03 2 . ± . i ] — — — — 0 . ± .
01 3 . ± . . ± .
02 1 . ± . . ± .
02 2 . ± . ii ] — — — — — — — — — —6563 H α . ± .
16 53 . ± . . ± .
32 46 . ± . . ± .
12 107 . ± . . ± .
18 63 . ± . . ± .
26 46 . ± . ii ] 0 . ± .
02 4 . ± . . ± .
04 3 . ± . . ± .
02 5 . ± . . ± .
03 5 . ± . . ± .
04 3 . ± . i . ± .
01 0 . ± . ii ] 0 . ± .
02 6 . ± . . ± .
06 6 . ± . . ± .
02 15 . ± . . ± .
03 6 . ± . . ± .
06 8 . ± . ii ] 0 . ± .
02 4 . ± . . ± .
05 5 . ± . . ± .
02 11 . ± . . ± .
03 5 . ± . . ± .
05 6 . ± . F (H β ) = . ± . F (H β ) = . ± . F (H β ) = . ± . F (H β ) = . ± . F (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . F (H β ) =
1. Equivalent widths of Balmer lines are corrected for underlying stellar absorption. The reddening coe ffi cient, C (H β ), E ( B − V ) (derived as 0 . × C (H β )) and the reddening-correctedH β flux, F (H β )( × − ergs cm − s − ) are listed for each region. The quoted uncertainties account for measurement, flux-calibration and reddening coe ffi cient errors. . M . C a i r´ o s e t a l . : M a pp i ng t h e p r op e r ti e s o f b l u ec o m p ac t d w a rf g a l a x i e s : Table 4.
Reddening-corrected line ratios, normalized to H β ( ∗ ) λ Ion Mrk 750 Mrk 206Nuclear Integrated Nuclear Integrated F λ − W λ F λ − W λ F λ − W λ F λ − W λ ii ] 2 . ± .
17 106 . ± . . ± .
23 86 . ± . . ± .
33 97 . ± . . ± .
27 219 . ± . iii ] 0 . ± .
03 28 . ± . . ± .
05 24 . ± . + HeI 0 . ± .
01 12 . ± . . ± .
03 12 . ± . iii ] 0 . ± .
02 18 . ± . . ± .
03 17 . ± . δ . ± .
01 19 . ± . . ± .
02 19 . ± . . ± .
03 4 . ± . γ . ± .
01 42 . ± . . ± .
03 28 . ± . . ± .
02 12 . ± . . ± .
05 5 . ± . iii ] 0 . ± .
01 5 . ± . . ± .
02 4 . ± . i . ± .
01 3 . ± . . ± .
01 3 . ± . β . ± .
00 114 . ± . . ± .
00 68 . ± . . ± .
00 30 . ± . . ± .
00 20 . ± . iii ] 1 . ± .
04 280 . ± . . ± .
05 107 . ± . . ± .
02 21 . ± . . ± .
06 14 . ± . iii ] 5 . ± .
12 801 . ± . . ± .
14 323 . ± . . ± .
05 64 . ± . . ± .
13 43 . ± . i . ± .
01 17 . ± . . ± .
01 11 . ± . . ± .
01 5 . ± . . ± .
02 4 . ± . i ] 0 . ± .
01 5 . ± . . ± .
01 3 . ± . ii ] — — — — 0 . ± .
01 9 . ± . . ± .
03 5 . ± . α . ± .
07 984 . ± . . ± .
14 325 . ± . . ± .
08 140 . ± . . ± .
23 78 . ± . ii ] 0 . ± .
01 29 . ± . . ± .
01 19 . ± . . ± .
02 26 . ± . . ± .
06 16 . ± . i . ± .
01 5 . ± . . ± .
01 5 . ± . . ± .
01 1 . ± . ii ] 0 . ± .
01 37 . ± . . ± .
01 23 . ± . . ± .
01 16 . ± . . ± .
04 10 . ± . ii ] 0 . ± .
01 29 . ± . . ± .
01 17 . ± . . ± .
01 12 . ± . . ± .
03 7 . ± . F (H β ) = . ± . F (H β ) = . ± . F (H β ) = . ± . F (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . L . M . C a i r´ o s e t a l . : M a pp i ng t h e p r op e r ti e s o f b l u ec o m p ac t d w a rf g a l a x i e s : Table 5.
Reddening-corrected line ratios, normalized to H β ( ∗ ) λ Ion Tololo 1434 +
032 Mrk 475Knot A Knot B Knot CD Integrated Nuclear Integrated F λ − W λ F λ − W λ F λ − W λ F λ − W λ F λ − W λ F λ − W λ ii ] 2 . ± .
22 153 . ± . . ± .
43 123 . ± . . ± .
28 490 . ± . . ± .
71 64 . ± . . ± .
12 107 . ± . . ± .
28 139 . ± . iii ] 0 . ± .
05 17 . ± . . ± .
11 13 . ± . . ± .
19 11 . ± . . ± .
03 44 . ± . . ± .
09 51 . ± . + HeI 0 . ± .
03 7 . ± . . ± .
02 19 . ± . iii ] 0 . ± .
03 12 . ± . . ± .
07 9 . ± . . ± .
02 30 . ± . δ . ± .
02 11 . ± . . ± .
05 8 . ± . . ± .
01 28 . ± . . ± .
03 18 . ± . γ . ± .
03 27 . ± . . ± .
05 22 . ± . . ± .
10 8 . ± . . ± .
10 7 . ± . . ± .
01 57 . ± . . ± .
03 46 . ± . iii ] 0 . ± .
01 2 . ± . . ± .
01 10 . ± . . ± .
02 7 . ± . i — — — — — — — — 0 . ± .
01 4 . ± . ii — — — — — — — — 0 . ± .
01 4 . ± . β . ± .
00 70 . ± . . ± .
00 49 . ± . . ± .
00 19 . ± . . ± .
00 25 . ± . . ± .
00 143 . ± . . ± .
00 91 . ± . iii ] 1 . ± .
05 118 . ± . . ± .
08 86 . ± . . pm .
11 26 . ± . . ± .
10 35 . ± . . ± .
04 258 . ± . . ± .
06 148 . ± . iii ] 4 . ± .
13 307 . ± . . ± .
21 254 . ± . . ± .
27 73 . ± . . ± .
25 104 . ± . . ± .
13 799 . ± . . ± .
18 411 . ± . i . ± .
01 9 . ± . . ± .
01 19 . ± . . ± .
01 9 . ± . i ] — — — — — — — — — — — —6548 [N ii ] 0 . ± .
01 7 . ± . . ± .
01 3 . ± . α . ± .
11 340 . ± . . ± .
20 207 . ± . . ± .
31 94 . ± . . ± .
26 102 . ± . . ± .
08 636 . ± . . ± .
12 347 . ± . ii ] 0 . ± .
01 14 . ± . . ± .
02 9 . ± . . ± .
05 6 . ± . . ± .
04 4 . ± . . ± .
01 21 . ± . . ± .
01 9 . ± . i . ± .
01 4 . ± . . ± .
01 7 . ± . . ± .
01 4 . ± . ii ] 0 . ± .
01 23 . ± . . ± .
04 19 . ± . . ± .
06 11 . ± . . ± .
05 9 . ± . . ± .
01 32 . ± . . ± .
02 24 . ± . ii ] 0 . ± .
01 17 . ± . . ± .
03 13 . ± . . ± .
05 7 . ± . . ± .
05 7 . ± . . ± .
01 26 . ± . . ± .
02 21 . ± . F (H β ) = . ± . F (H β ) = . ± . F (H β ) = . ± . F (H β ) = . ± . F (H β ) = . ± . F (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . . M . C a i r´ o s e t a l . : M a pp i ng t h e p r op e r ti e s o f b l u ec o m p ac t d w a rf g a l a x i e s : Table 6.
Reddening-corrected line ratios, normalized to H β ( ∗ ) λ Ion IZw 123 IZw 159Nuclear Integrated Nuclear Integrated F λ − W λ F λ − W λ F λ − W λ F λ − W λ ii ] 1 . ± .
17 55 . ± . . ± .
34 37 . ± . . ± .
27 79 . ± . . ± .
49 87 . ± . iii ] 0 . ± .
05 22 . ± . . ± .
15 13 . ± . . ± .
05 6 . ± . + HeI 0 . ± .
02 4 . ± . iii ] 0 . ± .
02 11 . ± . δ . ± .
01 12 . ± . . ± .
02 6 . ± . γ . ± .
02 25 . ± . . ± .
04 14 . ± . . ± .
02 13 . ± . . ± .
03 9 . ± . iii ] 0 . ± .
01 3 . ± . i . ± .
01 3 . ± . β . ± .
00 66 . ± . . ± .
00 44 . ± . . ± .
00 33 . ± . . ± .
00 24 . ± . iii ] 1 . ± .
05 131 . ± . . ± .
09 73 . ± . . ± .
02 31 . ± . . ± .
03 19 . ± . iii ] 5 . ± .
14 406 . ± . . ± .
24 226 . ± . . ± .
07 95 . ± . . ± .
09 61 . ± . i . ± .
01 9 . ± . . ± .
01 4 . ± . . ± .
01 3 . ± . i ] — — — — 0 . ± .
01 3 . ± . . ± .
02 3 . ± . ii ] — — — — 0 . ± .
01 7 . ± . α . ± .
07 323 . ± . . ± .
19 204 . ± . . ± .
07 153 . ± . . ± .
11 104 . ± . ii ] 0 . ± .
01 15 . ± . . ± .
02 11 . ± . . ± .
01 18 . ± . . ± .
02 13 . ± . i . ± .
01 4 . ± . . ± .
01 1 . ± . ii ] 0 . ± .
01 16 . ± . . ± .
03 17 . ± . . ± .
01 18 . ± . . ± .
02 15 . ± . ii ] 0 . ± .
01 14 . ± . . ± .
03 12 . ± . . ± .
01 14 . ± . . ± .
02 11 . ± . F (H β ) = . ± . F (H β ) = . ± . F (H β ) = . ± . F (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . C (H β ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . E ( B − V ) = . ± . L . M . C a i r´ o s e t a l . : M a pp i ng t h e p r op e r ti e s o f b l u ec o m p ac t d w a rf g a l a x i e s : Table 7.
Physical parameters and chemical abundances
Parameter Mrk 407 Mrk 32 Mrk 750 Mrk 206Nuclear Integrated Knot AB Knot C Integrated Nuclear Integrated Nuclear Integrated N e ([S ii ]) (cm − ) 188 197 110 154 169 < <
100 154 < T e ([O ii ]) (10 K) — — 1 . ± .
22 — — 1 . ± .
02 1 . ± .
09 — — T e ([O iii ]) (10 K) — — 1 . ± .
31 — — 1 . ± .
03 1 . ± .
13 — —12 + log(O / H) – ( T e ) — — 7 . ± .
10 — — 8 . ± .
03 8 . ± .
07 — —12 + log(Ne ++ / H + ) — — 6 . ± .
18 — — 7 . ± .
04 7 . ± .
12 — —12 + log(S + / H + ) — — 5 . ± .
09 — — 5 . ± .
02 5 . ± .
06 — —12 + log(N + / H + ) — — 6 . ± .
12 — — 6 . ± .
02 6 . ± .
06 — —log(N / O) — — − . ± .
20 — — − . ± . − . ± .
12 — —12 + log(O / H) – (N2) 8.26 8.24 8.16 8.24 8.22 8.17 8.17 8.44 8.4812 + log(O / H) – (O3N2) 8.24 8.24 8.17 8.25 8.25 8.10 8.11 8.41 8.42log([O iii ]5007 / H β ) 0 . ± .
02 0 . ± .
04 0 . ± .
02 0 . ± .
02 0 . ± .
03 0 . ± .
01 0 . ± .
01 0 . ± .
01 0 . ± . ii ]6584 / H α ) − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . ii ]6717 + / H α ) − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . i ]6300 / H α ) — — − . ± . − . ± . − . ± . − . ± .
03 — − . ± .
04 —(*) Notes: T e ([O ii ]) derived from the relation: T e ([O ii ]) = . × T e ([O iii ]) + .
26 found by Pilyugin et al. (2006); T e ([O iii ]): electron temperature measured from [O iii ] λ + log(O / H) – ( T e ): direct O / H abundance derived from T e ([O iii ]);12 + log(O / H) – (N2): O / H derived from the N2 index (Pettini & Pagel 2004); the associated uncertainty is ± .
38; 12 + log(O / H) – (O3N2): O / H derived from the O3N2 index (Pettini & Pagel 2004); the associated uncertainty is ± . . M . C a i r´ o s e t a l . : M a pp i ng t h e p r op e r ti e s o f b l u ec o m p ac t d w a rf g a l a x i e s : Table 8.
Physical parameters and chemical abundances
Parameter Tololo 1434 +
032 Mrk 475 IZw 123 IZw 159Knot A Knot B Knot CD Integrated Nuclear Integrated Nuclear Integrated Nuclear Integrated N e ([S ii ]) (cm − ) < < <
100 150 147 323 287 <
100 113 < T e ([O ii ]) (10 K) 1 . ± .
08 — — — 1 . ± .
03 1 . ± .
09 1 . ± .
03 — — — T e ([O iii ]) (10 K) 1 . ± .
11 — — — 1 . ± .
04 1 . ± .
13 1 . ± .
05 — — —12 + log(O / H) – ( T e ) 8 . ± .
07 — — — 7 . ± .
02 7 . ± .
07 8 . ± .
03 — — —12 + log(Ne ++ / H + ) 7 . ± .
12 — — — 7 . ± .
05 7 . ± .
13 7 . ± .
06 — — —12 + log(S + / H + ) 5 . ± .
06 — — — 5 . ± .
02 5 . ± .
06 5 . ± .
03 — — —12 + log(N + / H + ) 6 . ± .
06 — — — 5 . ± .
02 5 . ± .
08 6 . ± .
03 — — —log(N / O) − . ± .
12 — — — − . ± . − . ± . − . ± .
07 — — —12 + log(O / H) – (N2) 8.13 8.14 8.23 8.15 8.06 8.07 8.15 8.17 8.32 8.3412 + log(O / H) – (O3N2) 8.08 8.10 8.21 8.12 8.01 8.02 8.07 8.10 8.30 8.31log([O iii ]5007 / H β ) 0 . ± .
01 0 . ± .
02 0 . ± .
03 0 . ± .
03 0 . ± .
01 0 . ± .
02 0 . ± .
01 0 . ± .
02 0 . ± .
01 0 . ± . ii ]6584 / H α ) − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . ii ]6717 + / H α ) − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . − . ± . i ]6300 / H α ) — — — — — — — — − . ± . − . ± ..