1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass{emulateapj}
3:
4: \newcommand{\vdag}{(v)^\dagger}
5: \newcommand{\myemail}{hwu@bao.ac.cn}
6:
7:
8: \shorttitle{METALLICITY AND AGN ACTIVITY ON MIR EMISSION}
9: %
10: \shortauthors{Wu, et al.}
11: \def\comment#1{\bf ???: #1}
12:
13: %\received{}
14: %\accepted{}
15:
16: \begin{document}
17:
18: \title{EFFECTS OF METALLICITY AND AGN ACTIVITY ON THE MID-INFRARED DUST
19: EMISSION OF GALAXIES}
20:
21: \author{
22: Hong Wu\altaffilmark{1,2},
23: Yi-Nan Zhu\altaffilmark{1,3},
24: Chen Cao\altaffilmark{1,3},
25: Bo Qin\altaffilmark{1}
26: }
27:
28: \email{hwu@bao.ac.cn}
29: \altaffiltext{1}{National Astronomical Observatories, Chinese Academy of Sciences, A20 Datun Road, Beijing 100012, P.R.\ China}
30: \altaffiltext{2}{Visiting Scholar, Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI 96822}
31: \altaffiltext{3}{Graduate School, Chinese Academy of Sciences, Beijing, 100039, P.R.\ China}
32:
33:
34: \begin{abstract}
35: Using a sample of the $Spitzer$ SWIRE-field galaxies whose optical
36: spectra are taken from the Data Release 4 of the Sloan Digital Sky Survey,
37: we study possible correlations between the Mid-InfraRed (MIR) dust emission from these
38: galaxies and both their metallicities and AGN activities. We find that
39: both metallicity and AGN activity are well correlated with the
40: following ratios: PAH(Polycyclic Aromatic Hydrocarbon)-to-star,
41: VSG(Very Small Grain)-to-star and PAH-to-VSG, which can
42: be characterized by $\nu L_{\nu}[8\mu m({\rm dust})]/\nu L_{\nu}[3.6\mu m]$,
43: $\nu L_{\nu}[24\mu m]/\nu L_{\nu}[3.6\mu m]$, and
44: $\nu L_{\nu}[8\mu m({\rm dust})]/\nu L_{\nu}[24\mu m]$ respectively.
45: We argue that our MIR-metallicity correlation could be explained by
46: either the amount of dust (ongoing dust formation) or dust destruction
47: (PAHs and VSGs could be destroyed by hard and intense radiation fields),
48: and that the MIR-AGN correlation could arise due to either PAH destruction
49: or an enhanced VSG continuum by the central AGN.
50:
51: \end{abstract}
52:
53: \keywords { infrared: galaxies -- galaxies: starburst -- galaxies: abundances -- galaxies: active}
54:
55: \section{INTRODUCTION}
56: \label{sec intro}
57:
58: In galaxies, the MIR bands not only contain a large number of
59: atomic, ionic and molecular lines, but also cover various dust
60: features from grains of different size \citep{sturm02}. These
61: include the features related to the polycyclic aromatic hydrocarbons (PAHs)
62: and the dust continuum from the very small grains (VSGs).
63: Recent work
64: \citep{calzetti05,wu05,vogler05,alonso06,elbaz02,boselli04,forster04,roussel01}
65: have shown that the MIR dust emission are well correlated with the star
66: formation rate (SFR) indicators, such as H$\alpha$, Pa$\alpha$,
67: radio, UV, and far-infrared in either star forming regions in
68: galaxies or galaxies themselves as a whole. With further advantages
69: of having less extinction, the MIR dust luminosities are ideal
70: tracers of star formation and can be widely used to estimate the SFR.
71:
72: The MIR emission in galaxies could, however, be affected by many
73: factors. For example, early observations on the ground as well as
74: recent space observations by the Infrared Space Observatory ($ISO$)
75: and the $Spitzer$ Space Telescope \citep{werner04} have shown that the PAH
76: emission is generally suppressed in low-metallicity galaxies
77: \citep{roche91,madden00,galliano05,madden06,wuyl06}, or even
78: missing in the most metal-poor galaxies
79: \citep{houck04b,engelbracht05,thuan1999}. \citet{siebenmorgen04} have
80: demonstrated the absence of PAHs in the nuclei of AGN-hosting
81: galaxies. Using the $Spitzer$ Infrared Spectrograph \citep[IRS,][]{houck04a} spectra,
82: \citet{weedman05} have confirmed that the PAH
83: emission is much weaker in AGNs.
84: Moreover, the AGN activity could also steepen the shape of the MIR
85: VSG continuum in some active galaxies \citep{verma2005}.
86:
87: To quantitatively examine the factors that could affect the MIR
88: emission in galaxies, we need a sample of galaxies with both MIR
89: measurements and optical spectra. The {\it Spitzer} Wide-area
90: InfraRed Extragalactic survey \citep[SWIRE;][]{lonsdale03} and the
91: Sloan Digital Sky Survey \citep[SDSS;][]{stoughton02} provide
92: such opportunities. The Infrared Array Camera \citep[IRAC,][]{fazio04} 8$\mu$m band, which
93: just covers the main PAH features, can be used to trace the PAH
94: emission, while the Multiband Imaging Photometer \citep[MIPS,][]{rieke04} 24$\mu$m band, which
95: covers the continuum, can trace the VSG emission. The SDSS spectra
96: provide us information of both the metallicities and AGN activities
97: of galaxies. In the present paper, we mainly focus on the
98: correlations between the MIR emission and metallicity/AGN-activity.
99:
100: In Section 2, we describe the data, data reduction, and the sample.
101: The spectral classification, the metallicity and the definition of
102: AGN activity are presented in Section 3. The correlation
103: analysis between the MIR and metallicity/AGN-activity are shown in
104: Section 4. Finaly, Section 5 and 6 are discussions and summary.
105: Throughout the paper we adopt a $\Lambda$CDM cosmology
106: with $\Omega_M=0.3$, $\Omega_{\Lambda}=0.7$, and $H_0=70$.
107:
108: %
109: \section{DATA AND SAMPLE}
110: %
111: The $Spitzer$ SWIRE fields \citep{lonsdale03} of ELAIS-N1, ELAIS-N2,
112: and Lockman Hole have been imaged by both instruments: IRAC
113: \citep[3.6, 4.5, 5.8, 8.0~$\mu$m;][]{fazio04} and MIPS
114: \citep[24$\mu$m;][]{rieke04}. These fields have also been completely
115: or partly covered by the SDSS Data Release 4
116: \citep[DR4;][]{adelman06}. The overlap area between the three SWIRE
117: fields and the SDSS DR4 field used in this paper is about 15 degree$^2$.
118: The Basic Calibrated Data (BCD) images of the IRAC four bands were
119: obtained from the {\it Spitzer} Science Center, which include
120: flat-field corrections, dark subtraction, linearity and flux
121: calibrations \citep{fazio04}. The IRAC images (in all four IRAC
122: bands) were mosaiced from the BCD images after pointing refinement,
123: distortion correction and cosmic-ray removal with a final pixel
124: scale of 0.6\arcsec \ as described by \citet{huang04} and
125: \citet{wu05}, whilst the MIPS 24$\mu$m images were in a similar way
126: but with a final pixel scale of 1.225\arcsec \citep{wen07,cao07}. The
127: FWHMs of point spread functions (PSFs) in mosaiced images are around
128: 2\arcsec \ for the IRAC four bands and 6\arcsec \ for the MIPS
129: 24$\mu$m band. Matching the sources detected by the SExtractor
130: \citep{bertin96} in all the five bands with the 2MASS sources
131: \citep{cutri03} ensured the astrometric uncertainties to be less than 0.1$\arcsec$.
132:
133: The MIR photometries were obtained by the SExtractor with an
134: aperture of 6\arcsec \ for the IRAC four bands and the MIPS 24$\mu$m
135: band. The apertures of 6\arcsec \ were used to match the FWHM of the PSF
136: of the MIPS 24$\mu$m band, though the fiber aperture of the SDSS spectra is
137: 3\arcsec. All these magnitudes are in the AB magnitude system
138: \citep{oke83}. The values of -0.20, -0.17, -0.23, -0.41 magnitudes
139: for 3.6, 4.5, 5.8, 8.0~$\mu$m bands were adopted to correct the
140: apertures of 6\arcsec \ to 24\arcsec \ respectively. Comparing with
141: the model colors of the IRAC magnitudes and 2MASS $K_S$ magnitude
142: \citep{cutri03} for the stars with $J$-$K_S$ $\le$ 0.3 as
143: \citet{lacy05} did for sources in the {\it Spitzer} extragalactic First
144: Look Survey (xFLS) field, small additional corrections were made
145: for all the four IRAC bands. And these give us the calibration
146: errors in the four IRAC bands better than 0.08 mag. A factor of
147: -1.21 magnitude was adopted to correct the aperture from 6\arcsec \
148: to 30\arcsec \ for the 24$\mu$m band. An additional correction of -0.196
149: mag was also performed for the 24$\mu$m sources. This included a
150: factor of 1.15 of calibration and a factor of 1.04 of color
151: correction \citep{surace05}. The calibration accuracy of the 24$\mu$m photometry
152: is better than 10\% \citep{rieke04}.
153:
154: To build our galaxy sample, we first selected the galaxies of three
155: SWIRE fields from the main galaxy sample of the SDSS-DR4
156: \citep{adelman06} and then matched them with the MIR sources
157: detected in the $Spitzer$ SWIRE fields with a matching radius of
158: 2$\arcsec$. 14 out of 1628 (about 1\%) of the SDSS galaxies were
159: mis-matched with the MIR sources, due to the 2\arcsec \ matching
160: radius. To obtain reliable line ratios and to avoid possible
161: contaminations from the old stellar population, we only kept the sources
162: with a reliable H$\beta$ flux (S/N $\ge 5$) and a higher H$
163: {\alpha}$ equivalent width (less than $-5.0$, the minus here means
164: emission), where the H$ {\alpha}$ line emission flux used here was
165: from BC03 \citep{bc03} continuum-subtracted. \footnote{MPA emission
166: line catalog, which is available at {\it
167: http://www.mpa-garching.mpg.de/SDSS/DR4/raw\ data.html}} This gives
168: our sample of 600 galaxies. All these galaxies have 8$\mu$m fluxes,
169: and most have 24$\mu$m fluxes. The redshifts of the galaxies in
170: the sample are from 0.004 to 0.18.
171:
172: \section{SPECTRAL CLASSIFICATION, METALLICITY AND AGN ACTIVITY}
173:
174: Our spectral classification was based on the traditional
175: line-diagnostic diagram [OIII]/H$\beta$ versus [NII]/H$\alpha$,
176: \citep{veilleux87} as in Figure~\ref{fig1}. The dotted curve in the
177: figure is the parametrized curve by \citet{kauffmann03} while the
178: dashed curve is the theoretical boundary for starburst by
179: \citet{kewley01}. The galaxies below the dotted curve are
180: HII galaxies, the galaxies between the dotted curve and dashed curve are
181: mixture types \citep{wu98} or composites \citep{kewley06}, and
182: galaxies above the dashed curve are narrow line AGNs. There are
183: altogether 472 HII galaxies, 99 composites and 29 AGNs.
184:
185: The metallicities of HII galaxies were from the metallicity catalog
186: of SDSS DR4 \citep{tremonti04} \footnote{Available at {\it
187: http://www.mpa-garching.mpg.de/SDSS/DR4/Data/oh\ catalogue.html}}.
188: Here we adopted the median of the likelihood distribution for 12 + Log
189: O/H. 329 HII galaxies, which have measured metallicities in the
190: \citet{tremonti04} catalog, were selected for the following
191: analysis. 297 of them have 24$\mu$m fluxes. We defined dwarf
192: galaxies as the galaxies whose absolute B magnitudes were greater
193: than $-18$. Here the B magnitudes were calculated from the SDSS $g$-
194: and $r$- magnitudes \citep{smith02}. The dwarf galaxies are marked
195: with pluses in Figure~\ref{fig1}.
196:
197: Figure~\ref{fig2} shows the metallicities of HII galaxies as a
198: function of the internal reddening.
199: The internal reddening (characterized by E(B-V))
200: was derived from the Balmer decrement H$ {\alpha}$/H$ {\beta}$
201: \citep{calzetti01}. The figure suggests that dusty galaxies tend to
202: be metal rich.
203:
204: To quantify AGN activity, here we define a distance $d_{AGN}$ in dex
205: (see the red dotted line in Figure 1 as an example) in the traditional
206: line-diagnostic diagram [OIII]/H$ {\beta}$ versus [NII]/H$ {\alpha}$
207: to represent AGN activity. This is quite similar to the star-forming
208: distance defined by \citet{kewley06} in the diagnostic diagram
209: [OIII]/H$ {\beta}$ versus [OI]/H${\alpha}$ or [SII]/H${\alpha}$.
210: Considering that the composites could be dominated by star formation
211: and that this would weaken possible contributions from the central AGN,
212: the $d_{AGN}$ is defined as the distance of an AGN from its position ($x_p$,$y_p$)
213: to the \citet{kewley01}'s curve along the parallel of the best linear
214: fitting of all AGNs (solid line) in Figure~\ref{fig1}, and
215: can be formulized as :
216: %
217: \begin{equation}
218: d_{AGN} = 0.507 ( 5.8 x_p + y_p - 3.196 +A )
219: \end{equation}
220: \begin{equation}
221: A = \sqrt{14.152 + ( y_p - 5.8 x_p + 1.536 )^2},
222: \end{equation}
223: where $x=log$[NII]/H$\alpha$ and $y=log$[OIII]/H$\beta$.
224: Larger distances correspond to stronger AGN
225: activities. After removing the two most deviated AGNs, 27 AGNs between
226: the two dash-dotted lines were selected for our following analysis. Three
227: of them do not have 24$\mu$m fluxes.
228:
229:
230: \section{ANALYSIS}
231:
232: \subsection{MIR Ratios Versus Metallicity}
233:
234: To correct the redshift effect, we adopted the SED of the normal HII
235: galaxy NGC~3351 from IRS observations in the $Spitzer$ Legacy
236: Program SINGS \citep{kennicutt03} as the template for the
237: K-corrections for our sample galaxies in both the 24$\mu$m and
238: 8$\mu$m bands \citep{wu05}. In order to avoid the strong PAH
239: emission, we used the 3.6$\mu$m band to estimate the stellar
240: contribution in the 8.0$\mu$m band. A factor of 0.26 \citep{wu05} was
241: used to scale the stellar continuum of 3.6$\mu$m to that of 8$\mu$m.
242: After subtracting the corresponding stellar continuum contribution,
243: we obtained the flux of 8$\mu$m dust emission denoted by
244: 8$\mu$m(dust), as in \citet{wu05}. The 8$\mu$m(dust) includes the
245: dust emission from both the PAHs and VSGs. Considering that the
246: contribution of the stellar continuum in the 24$\mu$m band is quite
247: small and negligible, we have not removed the stellar continuum from
248: the 24$\mu$m band flux.
249:
250: It is well known that the mass of a galaxy is well related to its
251: metallicity \citep{pagel81}. To avoid possible effects caused by
252: different mass variations, we used the MIR ratios $\nu L_{\nu}[8\mu
253: m(dust)]/\nu L_{\nu}[3.6\mu m]$ and $\nu L_{\nu}[24\mu m]/\nu L_ {\nu}[3.6\mu m]$
254: instead of the MIR luminosities in our analysis.
255: These ratios can be explained as the MIR dust emission in mass unit
256: \citep{wen07}, since the 3.6$\mu$m luminosity can approximately
257: represent the stellar mass \citep{smithbj07,hancock07}.
258:
259: Figure~\ref{fig3} shows the MIR ratios $\nu L_{\nu}[8\mu
260: m(dust)]/\nu L_{\nu}[3.6\mu m]$, $\nu L_{\nu}[24\mu m]/\nu L_{\nu}[3.6 \mu m]$,
261: and $\nu L_{\nu}[8 \mu m(dust)]/\nu L_{\nu}[24
262: \mu m]$ as functions of metallicity in the log-log space. We plotted
263: the mean values (in diamonds) and the 1-$\sigma$ standard deviation
264: bars with a metallicity bin of 0.2. The plus symbols represent the
265: dwarf galaxies. We can see that the MIR ratios and metallicity are
266: correlated. Both $\nu L_{\nu}[8 \mu m(dust)]/\nu L_{\nu}[3.6 \mu m]$
267: and $\nu L_{\nu}[24 \mu m]/\nu L_{\nu}[3.6 \mu m]$ ratios show
268: nearly linear correlations with metallicity in the log-log space,
269: and the Spearman correlations are tight with the probabilities of
270: null-hypothesis being 2.7$\times$10$^{-17}$ and 5.2$\times$10$^{-7}$
271: respectively. Almost all the dwarf galaxies here obey the same
272: correlation as the normal galaxies do, except Mrk1434 which has the
273: lowest metallicity in our sample. Although the MIR ratio
274: $\nu L_{\nu}[8 \mu m(dust)]/\nu L_{\nu}[24 \mu m]$ also correlates with
275: metallicity, this correlation only exists for galaxies with
276: metallicities lower than 8.7. For galaxies with higher metallicities,
277: the $\nu L_{\nu}[8 \mu m(dust)]/\nu L_{\nu}[24 \mu m]$ ratios almost
278: remain constant. Compared with the above two correlations, the
279: Spearman correlation between $\nu L_{\nu}[8 \mu m(dust)]/\nu L_{\nu}[24
280: \mu m]$ and metallicity is much weaker, with the probability of
281: null-hypothesis being 0.02.
282:
283: \subsection{MIR Ratios Versus AGN Distance $d_{AGN}$}
284:
285: Figure~\ref{fig4} shows the MIR ratios as functions of AGN activity.
286: Here we used the distance $d_{AGN}$ as defined in Section 3 to
287: characterize AGN activity. The MIR ratio $\nu L_{\nu}[8\mu
288: m(dust)]/\nu L_{\nu}[3.6\mu m]$ seems to mildly increase with the
289: distance $d_{AGN}$. The probability of null-hypothesis of the
290: Spearman correlation is 0.3, indicating that the correlation is
291: poor. The ratio $\nu L_{\nu}[24\mu m]/\nu L_{\nu}[3.6 \mu m]$ is
292: well correlated with $d_{AGN}$. The probability of null-hypothesis
293: of such a correlation is 3.6$\times$10$^{-4}$, indicating that the
294: 24$\mu$m emission in mass unit tend to be stronger with the level of
295: AGN activity. On the contrary, the MIR ratio $\nu L_{\nu}[8 \mu
296: m(dust)]/\nu L_{\nu}[24 \mu m]$ is anti-correlated with $d_{AGN}$.
297: The probability of null-hypothesis is 1.7$\times$10$^{-5}$. The
298: correlation between $\nu L_{\nu}[8 \mu m(dust)]/\nu L_{\nu}[24 \mu
299: m]$ and $d_{AGN}$ is the tightest among the three. Even the most
300: discrepant point in Figure~\ref{fig4} (a) and (b) also follows such a
301: correlation.
302:
303: \subsection{MIR Ratios Versus Internal Reddening}
304:
305: To explore possible influence of the internal reddening on the MIR
306: emission, the MIR ratios were also plotted against E(B-V)
307: in Figure~\ref{fig5}. For HII galaxies, both ratios,
308: $\nu L_{\nu}[8\mu m(dust)]/\nu L_{\nu}[3.6\mu m]$ and $\nu L_{\nu}[24\mu m]/\nu L_{\nu}[3.6 \mu m]$,
309: show tight correlations with E(B-V).
310: The probabilities of null-hypothesis of the Spearman
311: correlation are 3.1$\times$10$^{-22}$ and 3.4$\times$10$^{-10}$
312: respectively. Both correlations indicate that the dusty galaxies
313: have relatively strong 8$\mu$m or 24$\mu$m dust emission. Such
314: correlations can also be expected in Figure~\ref{fig3}, since the
315: metallicities of HII galaxies are also correlated with their
316: internal reddening (Figure~\ref{fig2}). However, $\nu L_{\nu}[8 \mu
317: m(dust)]/\nu L_{\nu}[24 \mu m]$ does not show any correlation with
318: E(B-V). The probability of null-hypothesis is 0.16. Therefore, this
319: ratio does not depend on the internal reddening.
320:
321: For AGNs, all the three MIR ratios do not correlate with reddening.
322: The Spearman correlations show that the probabilities of
323: null-hypothesis are 1.00, 0.16, and 0.09 respectively,
324: indicating that the internal reddening does not affect all the three
325: MIR ratios of AGNs.
326:
327: \section{DISCUSSIONS}
328: \label{sec discuss}
329:
330: \subsection{Aperture Effect}
331:
332: Since the galaxies in our sample cover a redshift range from 0.004
333: to 0.17, the 6\arcsec \ aperture corresponds to a physical size from
334: several hundred pc to several tens kpc. Will the aperture affect the
335: correlations we obtained? We examined this by grouping the galaxies
336: into several redshift bins, as shown in Figure~\ref{fig3} and
337: Figure~\ref{fig4}. We find that either the HII galaxies
338: (Figure~\ref{fig3}) or AGNs (Figure~\ref{fig4}) with different
339: redshifts follow the same correlations.
340:
341: To further explore the aperture affect, we plotted the MIR ratios as
342: functions of redshift in Figure~\ref{fig6}. The mean MIR ratios (except for dwarf galaxies)
343: and the corresponding standard deviations with a redshift bin of 0.02
344: were also plotted. There is only a small variation of the mean MIR
345: ratios (less than 0.2 dex) with redshift.
346:
347: Meanwhile, we ''virtually'' placed a template galaxy (e.g., the SINGS
348: galaxy NGC~3351) in different redshifts. The 6\arcsec \ aperture
349: corresponds to different physical sizes at different redshifts.
350: Therefore, we did the aperture photometry for the MIR images of the SINGS
351: galaxy NGC~3351 with a set of apertures, whose physical sizes are
352: equal to those of the 6\arcsec \ aperture at different redshifts. As
353: a result, we can obtain the MIR ratios of NGC~3351 as functions of
354: redshift or aperture, which are plotted as the dotted curves in
355: Figure~\ref{fig6}. Here we only gave the MIR ratios of NGC~3351 at
356: the rest-frame and do not consider the K-correction. From the
357: figure, all the three MIR ratios of the template NGC~3351 only vary
358: by about 0.1 to 0.2 dex with redshift.
359:
360: Both the sample statistics and template modelling show that the
361: aperture effect on the MIR ratios is weaker than the correlations shown in
362: Figure~\ref{fig3}. Therefore, we believe that the aperture effect
363: would not significantly affect our correlations obtained from
364: Figure~\ref{fig3} and Figure~\ref{fig4}.
365:
366: \subsection{Metallicity and Dust Emission}
367:
368: Although the VSG continuum also contributes to the $Spitzer$ 8$\mu$m
369: band \citep{smith07}, in HII galaxies, the strongest 7.7$\mu$m PAH
370: features still dominate the emission in this band. However, the VSG
371: continuum dominates the $Spitzer$ MIPS 24$\mu$m band. Therefore, the
372: correlations between the MIR ratios and metallicity in
373: Figure~\ref{fig3} reveal the relationships between the ratios of
374: PAH-to-star/VSG-to-star/PAH-to-VSG and metallicity. Here, the
375: 3.6$\mu$m luminosity can approximately represent the stellar mass
376: \citep{smithbj07,hancock07}. All these ratios increase with
377: metallicity, though the details are different (e.g., different
378: slopes). Apparently the dust properties in the HII galaxies seem to
379: depend on metallicity. However, the metallicity itself is often
380: related to many other factors, such as mass (the mass-metallicity
381: relation, i.e., lower mass dwarf galaxies often have lower
382: metallicities), and radiation field (in a low metallicity environment,
383: the radiation field is often hard). Hence, it is necessary to
384: explore the underlying mechanisms.
385:
386: In fact, all these MIR behaviors mainly depend on two main factors:
387: the local radiation field and the dust. The properties of the
388: radiation field include the hardness, which can be characterized by
389: the MIR [NeIII]/[NeII] \citep{Thornley00,wuyl06,madden06}, and the
390: intensity, which can be characterized by either the MIR luminosity
391: density \citep{engelbracht05,wuyl06} or star formation rate
392: \citep{rosenberg06}. The very hard and intense radiation field could
393: destroy the PAHs \citep{galliano05}, and even the VSGs
394: \citep{contursi00}.
395:
396: The dust properties include the amount of dust,
397: and the fractions and spatial distributions of the PAHs and VSGs.
398: Galaxies with lower or vanishing PAH or VSG emission could
399: contain less amount of dust, possibly because they have low
400: metallicities, which lack material to form dust, or their masses are too low
401: to retain the dust against the radiation pressure and winds,
402: or they are too young, to have had enough time to form PAHs \citep{hogg05,madden06,draine07b}.
403: Different fractions and distributions of the PAHs and VSGs could result
404: in different PAH-to-VSG ratios.
405:
406: The positive correlations between the PAH/VSG-to-star ratios and
407: metallicity obtained from Figure~\ref{fig3} could be explained by the
408: amount of dust. It could be expected from the relations between the
409: reddening and metallicity (Figure~\ref{fig2}) or the PAH/VSG-to-star
410: ratios (Figure~\ref{fig5} (a) and (b)). In fact, the internal reddening
411: is directly related to the amount of dust if assuming a
412: constant dust-to-gas ratio, due to the tight correlation between E(B-V)
413: and H column density \citep{bohlin78,draine03}. Therefore, the low
414: PAH/VSG-to-star ratio of lower metallicity galaxies could not be due to
415: that their masses are too low to retain the amount of dust, because
416: some of the normal galaxies also follow the same correlation
417: as lower metallicity galaxies do in Figure~\ref{fig3}.
418: It could be that they are so young that a large
419: amount of dust has not had enough time to be released. This is
420: supported by Figure~\ref{fig2}.
421:
422: As for the PAH-to-VSG ratio, if
423: assuming that the AGB stars are the dominant carbon reservoir to
424: form the PAHs \citep{dwek98} with an age of star formation less than
425: 500Myr when the 4M$ {\odot}$ has entered the AGB stage\citep{madden06},
426: the PAH-to-VSG ratio could increase with time. After that, the ratio
427: could approach a constant. This seems to be consistent with
428: Figure~\ref{fig3}(c), in which the value of $\nu L_{\nu}[8 \mu
429: m(dust)]/\nu L_{\nu}[24 \mu m]$ increases with metallicity and then
430: remains constant beyond solar metallicity. Such an explanation is
431: quite similar to the explanation of the low dust-to-gas ratios for
432: low metallicity SINGS galaxies by \citet{draine07b}.
433: Also the behavior of the PAH-to-VSG ratio with time can be well described
434: by the Figure 4 of \citet{galliano06} recent work on
435: chemical evolution modeling of carbon and silicate grains.
436:
437: Another explanation could be that the hard and intense radiation
438: field from the young stars in the lower metallicity galaxies can
439: destroy not only the PAHs but also the VSGs. Since the PAHs are much
440: more sensitive to hard radiation field than the VSGs are, the slope of
441: the PAH-metallicity relation is much steeper than that of the VSGs.
442: This results in a decreasing PAH-to-VSG ratio in the lower
443: metallicity galaxies (Figure~\ref{fig3}). Such a destruction effect
444: is further supported by Figure~\ref{fig7}, which shows the
445: anti-correlation between $\nu L_{\nu}[8 \mu m(dust)]/\nu L_{\nu}[24
446: \mu m]$ and the 24$\mu$m intensity for the sample of HII galaxies,
447: though such an anti-correlation is not as steep as that for AGNs.
448: The 24$\mu$m intensity is obtained from the ratio of the 24$\mu$m
449: luminosity to the physical area in the aperture of 6\arcsec.
450:
451: We thus conclude that both effects of dust formation and destruction can
452: explain the correlations between the MIR ratios and metallicity.
453: This is consistent with the conclusions for PAHs by \citet{wuyl06}
454: and \citet{draine07a}.
455:
456: \subsection{AGN Activity and Dust Emission}
457:
458: The MIR spectra of AGNs were already noted to be void of PAH
459: features \citep{roche85, roche91, aitken85, genzel00, laurent00,
460: siebenmorgen04}. This could be explained by the destruction of PAHs
461: due to the hard radiation field of the central source
462: \citep{madden06}. It is also supported by recent $Spitzer$ results
463: given by \citet{dale06} and \citet{armus07}, who used [OIV]/[NeII]
464: as the indicator of the hardness of radiation field. However,
465: Figure~\ref{fig4}(a) shows that the PAH-to-star ratio apparently
466: does not have any tendency to decrease but rather increases with AGN
467: activity. \citet{draine07b} explained the SINGS AGNs as
468: low-luminosity AGNs, which have little effect on the 8$\mu$m dust
469: emission so that they show no evidence of PAH suppression. This is
470: however not the case for our sample of AGNs, since all of them have
471: higher H$ {\alpha}$ equivalent widths according to our selection in Section
472: 2. The amount of dust cannot account for that, since the internal
473: reddening does not correlate with any of the three MIR ratios
474: (Figure~\ref{fig5}). We suggest that the AGNs not only can destroy
475: the PAHs in the central regions but could also possibly excite the
476: PAH emission in the outer regions \citep{smith07}.
477:
478: Another explanation is that the central AGN raises the level of the
479: VSG continuum in the 8$\mu$m band. Similarly the sharp increase of
480: the VSG-to-star ratio with AGN activity in Figure~\ref{fig4}(b) may
481: also be due to the enhanced 24$\mu$m VSG continuum by the powerful AGN.
482: Although the outer PAH emission could be excited by the central AGN,
483: both the destruction of PAHs and the enhancement of the 24$\mu$m VSG continuum
484: by the AGN would result in the tight anti-correlation
485: between the PAH-to-VSG ratio and AGN activity.
486:
487: \subsection{Radiation Field and Dust Emission}
488:
489: One of the important factors in explaining the correlations between
490: the MIR ratios and metallicity or AGN activity is the radiation
491: field, which could be provided by either young stars or AGNs.
492: \citet{hogg05} suggested that [OIII]/H$ {\beta}$ could be an
493: indicator of radiation field hardness.
494:
495: Figure~\ref{fig8} shows the
496: three MIR ratios as functions of [OIII]/H$ {\beta}$. For AGNs, the
497: correlations between the MIR ratios and [OIII]/H$ {\beta}$ are
498: similar to those between the MIR ratios and $d_{AGN}$. So, the AGN
499: distance $d_{AGN}$ is directly related to the radiation field. For
500: HII galaxies, all these MIR ratios show the anti-correlation with
501: [OIII]/H$ {\beta}$. With [OIII]/H$ {\beta}$ increasing, the
502: PAH/VSG-to-star ratios of the HII galaxies and AGNs show different
503: features. Hence, the PAH/VSG-to-star ratios seem to depend on not
504: only the radiation field, but also some other important factor as
505: well. As discussed above, in the HII galaxies. the amount of dust
506: would play an important role.
507: However the PAH-to-VSG ratio, which
508: describes the fraction of two different dust components, does not
509: depend on the amount of dust but rather the radiation field. So, the
510: PAH-to-VSG ratio of the HII galaxies and AGNs would present the similar behaviors
511: along [OIII]/H$ {\beta}$, as shown in Figure~\ref{fig8} (c).
512:
513:
514: In a hard radiation field environment ([OIII]/H$ {\beta}$ $>$ 1),
515: the PAH-to-VSG ratio decreases sharply with increasing [OIII]/H$
516: {\beta}$, whether in HII galaxies or in AGNs. However, the PAH-to-VSG
517: ratio remains constant in a weak radiation field environment
518: ([OIII]/H$ {\beta}$ $<$ 1). We conclude that the hard radiation field,
519: rather than metallicity or AGN activity, is the most direct factor to determine
520: the PAH-to-VSG ratio.
521:
522: \section{SUMMARY}
523:
524: We have constructed a sample of galaxies from both the SDSS Data
525: Released 4 and the $Spitzer$ SWIRE fields to study possible effects
526: of both metallicity and AGN activity on the MIR dust emission.
527: Based on this sample we have found that:
528:
529: The MIR ratios: PAH-to-star, VSG-to-star and PAH-to-VSG, which can
530: be characterized respectively by $\nu L_{\nu}[8\mu m(dust)]/\nu L_{\nu}[3.6\mu m]$,
531: $\nu L_{\nu}[24\mu m]/\nu L_{\nu}[3.6\mu m]$, \\ and
532: $\nu L_{\nu}[8\mu m(dust)]/\nu L_{\nu}[24\mu m]$, are found to be
533: positively correlated with the metallicities of galaxies. The above
534: correlations could be explained by either the amount of dust (ongoing
535: dust formation) or dust destruction (either the PAHs or VSGs could
536: be destroyed by the hard and strong radiation field).
537:
538: The VSG-to-star and PAH-to-VSG ratios are strongly correlated or
539: anti-correlated with AGN activity. However, the PAH-to-star ratio
540: poorly depends on AGN activity. This may be due to the PAH destruction
541: or enhanced VSG continuum by the central AGN.
542:
543: \section{ACKNOWLEDGEMENTS}
544:
545: We thank Jia-sheng Huang, Zhong Wang, David Sanders, and Yan-Chun Liang
546: for helpful discussions. Also many thanks to anonymous referees.
547: H.W. is grateful to the Institute for Astronomy, University of Hawaii for hospitality.
548: This work was
549: supported by the National Science Foundation of China under grants
550: 10273012, 10333060, and 10473013, and a CAS grant KJCX3-SYW-N2.
551: It is, in part, based on observations made with the Spitzer Space Telescope, which is
552: operated by Jet Propulsion Laboratory of the California Institute of
553: Technology under NASA Contract 1407.
554:
555: Funding for the SDSS and SDSS-II has been provided by the Alfred
556: P. Sloan Foundation, the Participating Institutions, the National Science
557: Foundation, the U.S. Department of Energy, the National Aeronautics
558: and Space Administration, the Japanese Monbukagakusho, the Max Planck
559: Society, and the Higher Education Funding Council for England. The SDSS Web Site is http://www.sdss.org/.
560:
561: The SDSS is managed by the Astrophysical Research Consortium for the Participating Institutions. The Participating Institutions are the American Museum of Natural History, Astrophysical Institute Potsdam, University of Basel, University of Cambridge, Case Western Reserve University, University of Chicago, Drexel University, Fermilab, the Institute for Advanced Study, the Japan Participation Group, Johns Hopkins University, the Joint Institute for Nuclear Astrophysics, the Kavli Institute for Particle Astrophysics and Cosmology, the Korean Scientist Group, the Chinese Academy of Sciences (LAMOST), Los Alamos National Laboratory, the Max-Planck-Institute for Astronomy (MPIA), the Max-Planck-Institute for Astrophysics (MPA), New Mexico State University, Ohio State University, University of Pittsburgh, University of Portsmouth, Princeton University, the United States Naval Observatory, and the University of Washington.
562:
563:
564: %
565: \begin{thebibliography}{99}
566:
567: \bibitem[Adelman-McCarthy et al.(2006)]{adelman06} Adelman-McCarthy, J.~K., et al.\ 2006, \apjs, 162, 38
568: \bibitem[Aitken \& Roche(1985)]{aitken85} Aitken, D.~K., \& Roche, P.~F.\ 1985, \mnras, 213, 777
569: \bibitem[Alonso-Herrero et al.(2006)]{alonso06} Alonso-Herrero, A., Rieke, G.~H., Rieke, M.~J., Colina, L., P{\'e}rez-Gonz{\'a}lez, P.~G., \& Ryder, S.~D.\ 2006, \apj, 650, 835
570: \bibitem[Armus et al.(2007)]{armus07} Armus, L., et al.\ 2007, \apj, 656, 148
571: \bibitem[Bertin \& Arnouts(1996)]{bertin96} Bertin, E., \& Arnouts, S.\ 1996, \aaps, 117, 393
572: \bibitem[Bohlin et al.(1978)]{bohlin78} Bohlin, R.~C., Savage, B.~D., \& Drake, J.~F.\ 1978, \apj, 224, 132
573: \bibitem[Boselli et al.(2004)]{boselli04} Boselli, A., Lequeux, J., \& Gavazzi, G.\ 2004, \aap, 428, 409
574: \bibitem[Bruzual \& Charlot(2003)]{bc03} Bruzual, G., \& Charlot, S.\ 2003, \mnras, 344, 1000
575: \bibitem[Calzetti et al.(2005)]{calzetti05} Calzetti, D., et al.\ 2005, \apj, 633, 871
576: \bibitem[Calzetti(2001)]{calzetti01} Calzetti, D.\ 2001, \pasp, 113, 1449
577: \bibitem[Cao \& Wu(2007)]{cao07} Cao, C., \& Wu, H.\ 2007, \aj, 133, 1710
578: \bibitem[Contursi et al.(2000)]{contursi00} Contursi, A., et al.\ 2000, \aap, 362, 310
579: \bibitem[Cutri et al.(2003)]{cutri03} Cutri, R.~M., et al.\ 2003, The IRSA 2MASS All-Sky Point Source Catalog, NASA/IPAC
580: \bibitem[Dale et al.(2006)]{dale06} Dale, D.~A., et al.\ 2006, \apj, 646, 161
581: \bibitem[Draine \& Li(2007)]{draine07a} Draine, B.~T., \& Li, A.\ 2007, \apj, 657, 810
582: \bibitem[Draine et al.(2007)]{draine07b} Draine, B.~T., et al.\ 2007, ArXiv Astrophysics e-prints, arXiv:astro-ph/0703213
583: \bibitem[Draine(2003)]{draine03} Draine, B.~T.\ 2003, \araa, 41, 241
584: \bibitem[Dwek(1998)]{dwek98} Dwek, E.\ 1998, \apj, 501, 643
585: \bibitem[Elbaz et al.(2002)]{elbaz02} Elbaz, D., Cesarsky, C.~J., Chanial, P., Aussel, H., Franceschini, A., Fadda, D., \& Chary, R.~R.\ 2002, \aap, 384, 848
586: \bibitem[Engelbracht et al.(2005)]{engelbracht05} Engelbracht, C.~W., Gordon, K.~D., Rieke, G.~H., Werner, M.~W., Dale, D.~A., \& Latter, W.~B.\ 2005, \apjl, 628, L29
587: \bibitem[Fazio et al.(2004)]{fazio04} Fazio, G. G. et al., 2004, \apjs, 154, 10
588: \bibitem[F{\"o}rster Schreiber et al.(2004)]{forster04} F{\"o}rster Schreiber, N.~M., Roussel, H., Sauvage, M., \& Charmandaris, V.\ 2004, \aap, 419, 501
589: \bibitem[Galliano et al.(2005)]{galliano05} Galliano, F., Madden, S.~C., Jones, A.~P., Wilson, C.~D., \& Bernard, J.-P.\ 2005, \aap, 434, 867
590: \bibitem[Galliano(2006)]{galliano06} Galliano, F.\ 2006, ArXiv Astrophysics e-prints, arXiv:astro-ph/0610852
591: \bibitem[Genzel \& Cesarsky(2000)]{genzel00} Genzel, R., \& Cesarsky, C.~J.\ 2000, \araa, 38, 761
592: \bibitem[Hancock et al.(2007)]{hancock07} Hancock, M., Smith, B.~J., Struck, C., Giroux, M.~L., Appleton, P.~N., Charmandaris, V., \& Reach, W.~T.\ 2007, \aj, 133, 676
593: \bibitem[Hogg et al.(2005)]{hogg05} Hogg, D.~W., Tremonti, C.~A., Blanton, M.~R., Finkbeiner, D.~P., Padmanabhan, N., Quintero, A.~D., Schlegel, D.~J., \& Wherry, N.\ 2005, \apj, 624, 162
594: \bibitem[Houck et al.(2004)]{houck04a} Houck, J.~R., et al.\ 2004a, \apjs, 154, 18
595: \bibitem[Houck et al.(2004)]{houck04b} Houck, J.~R., et al.\ 2004b, \apjs, 154, 211
596: \bibitem[Huang et al.(2004)]{huang04} Huang, J.-S., et al.\ 2004, \apjs, 154, 44
597: \bibitem[Kauffmann et al.(2003)]{kauffmann03} Kauffmann, G., et al.\ 2003, \mnras, 346, 1055
598: \bibitem[Kennicutt et al.(2003)]{kennicutt03} Kennicutt, R.~C., Jr., et al.\ 2003, \pasp, 115, 928
599: \bibitem[Kewley et al.(2001)]{kewley01} Kewley, L.~J., Dopita, M.~A., Sutherland, R.~S., Heisler, C.~A., \& Trevena, J.\ 2001, \apj, 556, 121
600: \bibitem[Kewley et al.(2006)]{kewley06} Kewley, L.~J., Groves, B., Kauffmann, G., \& Heckman, T.\ 2006, \mnras, 372, 961
601: \bibitem[Lacy et al.(2005)]{lacy05} Lacy, M., et al.\ 2005, \apjs, 161, 41
602: \bibitem[Laurent et al.(2000)]{laurent00} Laurent, O., Mirabel, I.~F., Charmandaris, V., Gallais, P., Madden, S.~C., Sauvage, M., Vigroux, L., \& Cesarsky, C.\ 2000, \aap, 359, 887
603: \bibitem[Lonsdale et al.(2003)]{lonsdale03} Lonsdale, C.~J., et al.\ 2003, \pasp, 115, 897
604: \bibitem[Madden(2000)]{madden00} Madden, S.~C.\ 2000, New Astronomy Review, 44, 249
605: \bibitem[Madden et al.(2006)]{madden06} Madden, S.~C., Galliano, F., Jones, A.~P., \& Sauvage, M.\ 2006, \aap, 446, 877
606: \bibitem[Oke \& Gunn(1983)]{oke83} Oke, J.~B., \& Gunn, J.~E.\ 1983, \apj, 266, 713
607: \bibitem[Pagel \& Edmunds(1981)]{pagel81} Pagel, B.~E.~J., \& Edmunds, M.~G.\ 1981, \araa, 19, 77
608: \bibitem[Rieke et al.(2004)]{rieke04} Rieke, G.~H., et al.\ 2004, \apjs, 154, 25
609: \bibitem[Roche et al.(1991)]{roche91} Roche, P.~F., Aitken, D.~K., Smith, C.~H., \& Ward, M.~J.\ 1991, \mnras, 248, 606
610: \bibitem[Roche \& Aitken(1985)]{roche85} Roche, P.~F., \& Aitken, D.~K.\ 1985, \mnras, 213, 789
611: \bibitem[Rosenberg et al.(2006)]{rosenberg06} Rosenberg, J.~L., Ashby, M.~L.~N., Salzer, J.~J., \& Huang, J.-S.\ 2006, \apj, 636, 742
612: \bibitem[Roussel et al.(2001)]{roussel01} Roussel, H., Sauvage, M., Vigroux, L., \& Bosma, A.\ 2001, \aap, 372, 427
613: \bibitem[Siebenmorgen et al.(2004)]{siebenmorgen04} Siebenmorgen, R., Kr{\"u}gel, E., \& Spoon, H.~W.~W.\ 2004, \aap, 414, 123
614: \bibitem[Smith et al.(2007a)]{smithbj07} Smith, B.~J., Struck, C., Hancock, M., Appleton, P.~N., Charmandaris, V., \& Reach, W.~T.\ 2007a, \aj, 133, 791
615: \bibitem[Smith et al.(2007b)]{smith07} Smith, J.~D.~T., et al.\ 2007b, \apj, 656, 756
616: \bibitem[Smith et al.(2002)]{smith02} Smith, J.~A., et al.\ 2002, \aj, 123, 2121
617: \bibitem[Stoughton et al.(2002)]{stoughton02} Stoughton, C., et al.\ 2002, \aj, 123, 485
618: \bibitem[Sturm et al.(2002)]{sturm02} Sturm, E., Lutz, D., Verma, A., Netzer, H., Sternberg, A., Moorwood, A.~F.~M., Oliva, E., \& Genzel, R.\ 2002, \aap, 393, 821
619: \bibitem[Surace et al.(2005)]{surace05} Surace J.~A., et al., 2005, The SWIRE Data Release 2: Image Atlases and Source Catalogs for ELAIS-N1, ELAIS-N2, XMM-LSS, and the Lockman Hole
620: \bibitem[Thornley et al.(2000)]{Thornley00} Thornley, M.~D., Schreiber, N.~M.~F., Lutz, D., Genzel, R., Spoon, H.~W.~W., Kunze, D., \& Sternberg, A.\ 2000, \apj, 539, 641
621: \bibitem[Thuan et al.(1999)]{thuan1999} Thuan, T.~X., Sauvage, M., \& Madden, S.\ 1999, \apj, 516, 783
622: \bibitem[Tremonti et al.(2004)]{tremonti04} Tremonti, C.~A., et al.\ 2004, \apj, 613, 898
623: \bibitem[Veilleux \& Osterbrock(1987)]{veilleux87} Veilleux, S., \& Osterbrock, D.~E.\ 1987, \apjs, 63, 295
624: \bibitem[Verma et al.(2005)]{verma2005} Verma, A., Charmandaris, V., Klaas, U., Lutz, D., \& Haas, M.\ 2005, Space Science Reviews, 119, 355
625: \bibitem[Vogler et al.(2005)]{vogler05} Vogler, A., Madden, S.~C., Beck, R., Lundgren, A.~A., Sauvage, M., Vigroux, L., \& Ehle, M.\ 2005, \aap, 441, 491
626: \bibitem[Weedman et al.(2005)]{weedman05} Weedman, D.~W., et al.\ 2005, \apj, 633, 706
627: \bibitem[Wen et al.(2007)]{wen07} Wen, X.~Q., Wu, H., Cao, C., Xia, X.~Y., 2007, ChJAA, 7, 187
628: \bibitem[Werner et al.(2004)]{werner04} Werner, M. W. et al., 2004, \apjs, 154, 1
629: \bibitem[Wu et al.(1998)]{wu98} Wu, H., Zou, Z.~L., Xia, X.~Y., \& Deng, Z.~G.\ 1998, \aaps, 132, 181
630: \bibitem[Wu et al.(2005)]{wu05} Wu, H., Cao, C., Hao, C.-N., Liu, F.-S., Wang, J.-L., Xia, X.-Y., Deng, Z.-G., \& Young, C.~K.-S.\ 2005, \apjl, 632, L79
631: \bibitem[Wu et al.(2006)]{wuyl06} Wu, Y., Charmandaris, V., Hao, L., Brandl, B.~R., Bernard-Salas, J., Spoon, H.~W.~W., \& Houck, J.~R.\ 2006, \apj, 639, 157
632:
633: \end{thebibliography}
634:
635: \clearpage
636:
637: \begin{figure}
638: \centerline{\includegraphics[height=0.9\textwidth]{f1.eps}}
639: \caption{The line-diagnostic diagram [OIII]/H$\beta$ versus [NII]/H$\alpha$.
640: The dotted curve is the parametrized curve by
641: \citet{kauffmann03} and the dashed curve is the theoretical boundary
642: for starburst \citep{kewley01}. Under the dotted curve, the HII galaxies
643: are labelled as solid circles and dwarf galaxies as pluses.
644: The composites are between the dotted and dashed curves and
645: are labelled as open triangles. The AGNs are above the dashed curve
646: and are labelled as open boxes. To characterize AGN activity, we
647: define the AGN distance $d_{AGN}$ as the distance from these boxes
648: to the dashed curve along the best-fit solid line.} \label{fig1}
649: \end{figure}
650:
651: \clearpage
652:
653: \begin{figure}
654: \centerline{\includegraphics[height=0.9\textwidth]{f2.eps}}
655: \caption{The metallicities of HII galaxies as a function of their internal reddening.
656: The plus symbols are dwarf galaxies.
657: } \label{fig2}
658: \end{figure}
659:
660: \clearpage
661:
662: \begin{figure}
663: \centerline{\includegraphics[height=0.9\textwidth]{f3.eps}}
664: \caption{The MIR ratios (a) $\nu L_{\nu}[8\mu m(dust)]/\nu L_{\nu}[3.6\mu m]$, ]
665: (b) $\nu L_{\nu}[24\mu m]/\nu L_{\nu}[3.6 \mu m]$,
666: and (c) $\nu L_{\nu}[8 \mu m(dust)]/\nu L_{\nu}[24 \mu m]$ versus
667: metallicity. The plus symbols represent the dwarf galaxies and the
668: solid circles are normal galaxies. The diamonds and bars give the mean
669: values and 1-$\sigma$ standard deviations with a metallicity bin of
670: 0.2. Different colors show galaxies in different redshift bins.}
671: \label{fig3}
672: \end{figure}
673:
674: \clearpage
675:
676: \begin{figure}
677: \centerline{\includegraphics[height=0.9\textwidth]{f4.eps}}
678: \caption{The MIR ratios (a) $\nu L_{\nu}[8\mu m(dust)]/\nu L_{\nu}[3.6\mu m]$,
679: (b) $\nu L_{\nu}[24\mu m]/\nu L_{\nu}[3.6 \mu m]$,
680: and (c) $\nu L_{\nu}[8 \mu m(dust)]/\nu L_{\nu}[24 \mu m]$ versus
681: the distance $d_{AGN}$. Different colors represent AGNs in different
682: redshift bins. The dotted lines represent the best linear fitting. }
683: \label{fig4}
684: \end{figure}
685:
686: \clearpage
687:
688: \begin{figure}
689: \centerline{\includegraphics[height=0.9\textwidth]{f5.eps}}
690: \caption{The MIR ratios (a) $\nu L_{\nu}[8\mu m(dust)]/\nu L_{\nu}[3.6\mu m]$,
691: (b) $\nu L_{\nu}[24\mu m]/\nu L_{\nu}[3.6 \mu m]$,
692: and (c) $\nu L_{\nu}[8 \mu m(dust)]/\nu L_{\nu}[24 \mu m]$
693: as functions of reddening. The symbols are the same as in Figure~\ref{fig1}.
694: The diamonds and bars give the mean values and 1-$\sigma$ standard deviations
695: of HII galaxies with an E(B-V) bin of 0.1.
696: } \label{fig5}
697: \end{figure}
698:
699: \clearpage
700:
701: \begin{figure}
702: \centerline{\includegraphics[height=0.9\textwidth]{f6.eps}}
703: \caption{The MIR ratios (a) $\nu L_{\nu}[8\mu m(dust)]/\nu L_{\nu}[3.6\mu m]$,
704: (b) $\nu L_{\nu}[24\mu m]/\nu L_{\nu}[3.6 \mu m]$,
705: and (c) $\nu L_{\nu}[8 \mu m(dust)]/\nu L_{\nu}[24 \mu m]$
706: as functions of redshift.
707: The diamonds and bars are the mean values and 1-$\sigma$ standard deviations with a redshift bin
708: of 0.02.
709: The dotted curves show the variation of the MIR ratios of NGC~3351 with
710: redshift (aperture). Dward galaxies are labelled as plus symbols.
711: } \label{fig6}
712: \end{figure}
713:
714: \clearpage
715:
716:
717: \begin{figure}
718: \centerline{\includegraphics[height=0.9\textwidth]{f7.eps}}
719: \caption{The MIR ratio $\nu L {\nu}[8 \mu m(dust)]/\nu L {\nu}[24 \mu m]$
720: as a function of the 24$\mu$m intensity. The 24$\mu$m intensity is defined as
721: the 24$\mu$m luminosity in unit physical area and is obtained from the ratio of
722: the 24$\mu$m luminosity to the physical area of aperture 6 \arcsec \ in
723: each galaxy.
724: The symbols are the same as in Figure~\ref{fig1}.
725: } \label{fig7}
726: \end{figure}
727:
728: \clearpage
729:
730: \begin{figure}
731: \centerline{\includegraphics[height=0.9\textwidth]{f8.eps}}
732: \caption{The MIR ratios (a) $\nu L_{\nu}[8\mu m(dust)]/\nu L_{\nu}[3.6\mu m]$,
733: (b) $\nu L_{\nu}[24\mu m]/\nu L_{\nu}[3.6 \mu m]$,
734: and (c) $\nu L_{\nu}[8 \mu m(dust)]/\nu L_{\nu}[24 \mu m]$
735: as functions of the emission line ratio log([OIII]/H$ {\beta}$).
736: The symbols are the same as in Figure~\ref{fig1}.
737: } \label{fig8}
738: \end{figure}
739:
740: \end{document}
741: