2 Observations and reduction

2.1 Observations

 

 

Table 1: Journal of observations.

Object	Date	Exposure	Wavelength	Dispersion	Seeing	Airmass	PA
		time [s]	Range [Å]	[Å/pixel]	[arcsec]		[degree]
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
UM 151	1.02.2000	2 $\times$ 1800	$3700\div7000$	0.81/0.54	1.8	1.50	0
UM 408	2.02.2000	2 $\times$ 1800	$3700\div7000$	0.81/0.54	1.4	1.30	6
A 1228+12	20.01.2001	2 $\times$ 1800	$3700\div7400$	2.4	1.7	1.18	74

   

Table 2: Line intensities of the studied galaxies.

	A 1228+12		UM 151		UM 408
$\lambda_{0}$(Å) Iona	F($\lambda$)/F(H$\beta$)b	I($\lambda$)/I(H$\beta$)c	F($\lambda$)/F(H$\beta$)b	I($\lambda$)/I(H$\beta$)c	F($\lambda$)/F(H$\beta$)b	I($\lambda$)/I(H$\beta$)c
3727 [O II]	0.9503 $\pm$ 0.0722	0.9259 $\pm$ 0.0778	2.4257 $\pm$ 0.1997	2.6919 $\pm$ 0.2466	1.4502 $\pm$ 0.1280	2.1700 $\pm$ 0.2045
3835 H9	--	--	--	--	0.0590 $\pm$ 0.0187	0.0860 $\pm$ 0.0380
3868 [Ne III]	0.3791 $\pm$ 0.0308	0.3693 $\pm$ 0.0324	--	--	0.3812 $\pm$ 0.0409	0.5390 $\pm$ 0.0601
3889 He I + H8	0.1634 $\pm$ 0.0170	0.1958 $\pm$ 0.0275	0.1262 $\pm$ 0.0259	0.1815 $\pm$ 0.0528	0.1474 $\pm$ 0.0246	0.2082 $\pm$ 0.0434
3967 [Ne III] + H7	0.2254 $\pm$ 0.0196	0.2555 $\pm$ 0.0288	0.1020 $\pm$ 0.0285	0.1470 $\pm$ 0.0527	0.2077 $\pm$ 0.0294	0.2842 $\pm$ 0.0507
4101 H$\delta$	0.2210 $\pm$ 0.0196	0.2492 $\pm$ 0.0279	0.2409 $\pm$ 0.0320	0.2947 $\pm$ 0.0518	0.2113 $\pm$ 0.0230	0.2749 $\pm$ 0.0425
4340 H$\gamma$	0.4605 $\pm$ 0.0360	0.4798 $\pm$ 0.0416	0.4086 $\pm$ 0.0389	0.4588 $\pm$ 0.0553	0.4251 $\pm$ 0.0362	0.5056 $\pm$ 0.0492
4363 [O III]	0.1055 $\pm$ 0.0130	0.1028 $\pm$ 0.0131	--	--	0.0851 $\pm$ 0.0170	0.1002 $\pm$ 0.0202
4471 He I	0.0321 $\pm$ 0.0073	0.0313 $\pm$ 0.0073	0.0271 $\pm$ 0.0144	0.0273 $\pm$ 0.0151	0.0266 $\pm$ 0.0113	0.0302 $\pm$ 0.0129
4686 He II	0.0403 $\pm$ 0.0078	0.0392 $\pm$ 0.0078	--	--	--	--
4713 [Ar IV] + He I	0.0208 $\pm$ 0.0073	0.0203 $\pm$ 0.0073	--	--	--	--
4713 [Ar IV]	0.0163 $\pm$ 0.0080	0.0159 $\pm$ 0.0080	--	--	--	--
4861 H$\beta$	1.0000 $\pm$ 0.0771	1.0000 $\pm$ 0.0800	1.0000 $\pm$ 0.0829	1.0000 $\pm$ 0.0899	1.0000 $\pm$ 0.0758	1.0000 $\pm$ 0.0789
4959 [O III]	1.5651 $\pm$ 0.1213	1.5251 $\pm$ 0.1213	0.5509 $\pm$ 0.0493	0.5270 $\pm$ 0.0490	1.7714 $\pm$ 0.1318	1.7174 $\pm$ 0.1291
5007 [O III]	4.6938 $\pm$ 0.3588	4.5736 $\pm$ 0.3590	1.6881 $\pm$ 0.1343	1.6069 $\pm$ 0.1328	5.6848 $\pm$ 0.4445	5.4322 $\pm$ 0.4294
5876 He I	0.0992 $\pm$ 0.0107	0.0967 $\pm$ 0.0109	0.1227 $\pm$ 0.0211	0.1077 $\pm$ 0.0194	0.1538 $\pm$ 0.0311	0.1161 $\pm$ 0.0238
6300 [O I]	0.0230 $\pm$ 0.0096	0.0224 $\pm$ 0.0096	0.0636 $\pm$ 0.0171	0.0540 $\pm$ 0.0152	0.0584 $\pm$ 0.0232	0.0399 $\pm$ 0.0161
6312 [S III]	0.0146 $\pm$ 0.0087	0.0143 $\pm$ 0.0087	0.0161 $\pm$ 0.0119	0.0137 $\pm$ 0.0105	0.0309 $\pm$ 0.0231	0.0211 $\pm$ 0.0159
6364 [O I]	0.0085 $\pm$ 0.0055	0.0083 $\pm$ 0.0055	0.0203 $\pm$ 0.0145	0.0171 $\pm$ 0.0128	--	--
6548 [N II]	0.0102 $\pm$ 0.0068	0.0100 $\pm$ 0.0068	0.1455 $\pm$ 0.0234	0.1212 $\pm$ 0.0206	0.0342 $\pm$ 0.0306	0.0222 $\pm$ 0.0201
6563 H$\alpha$	2.6763 $\pm$ 0.2002	2.6217 $\pm$ 0.2185	3.4333 $\pm$ 0.2673	2.8819 $\pm$ 0.2533	4.3736 $\pm$ 0.3226	2.8264 $\pm$ 0.2289
6584 [N II]	0.0310 $\pm$ 0.0229	0.0302 $\pm$ 0.0229	0.4404 $\pm$ 0.0383	0.3661 $\pm$ 0.0353	0.0957 $\pm$ 0.0245	0.0615 $\pm$ 0.0160
6678 He I	0.0250 $\pm$ 0.0071	0.0244 $\pm$ 0.0071	0.0379 $\pm$ 0.0144	0.0313 $\pm$ 0.0124	0.0506 $\pm$ 0.0220	0.0319 $\pm$ 0.0140
6717 [S II]	0.0936 $\pm$ 0.0109	0.0912 $\pm$ 0.0113	0.4766 $\pm$ 0.0409	0.3926 $\pm$ 0.0378	0.2467 $\pm$ 0.0342	0.1545 $\pm$ 0.0222
6731 [S II]	0.0645 $\pm$ 0.0096	0.0628 $\pm$ 0.0099	0.3423 $\pm$ 0.0328	0.2816 $\pm$ 0.0298	0.1635 $\pm$ 0.0260	0.1021 $\pm$ 0.0168
7065 He I	0.0173 $\pm$ 0.0062	0.0168 $\pm$ 0.0063	--	--	--	--
7136 [Ar III]	0.0430 $\pm$ 0.0083	0.0419 $\pm$ 0.0085	--	--	--	--
C(H$\beta$) dex	0.00 $\pm$ 0.10		0.20 $\pm$ 0.10		0.57 $\pm$ 0.10
EW(abs) Å	2.45 $\pm$ 1.08		0.90 $\pm$ 0.67		0.05 $\pm$ 0.94
F(H$\beta$)d	232 $\pm$ 13		68 $\pm$ 4		33 $\pm$ 2
EW(H$\beta$,emis) Å	93 $\pm$ 5		20 $\pm$ 2		50 $\pm$ 3

^a The rest-frame wavelength in Å; ^b observed flux ratio; ^c corrected flux ratio; ^d in units of 10^-16 ergs s^-1 cm^-2.

The spectra of UM 151 and UM 408 were obtained with the TWIN spectrograph attached to the Cassegrain focus of the 3.5 m telescope at the Calar Alto Observatory (Spain) as supplementary objects to the main program devoted to the detailed spectroscopy of the HSS (Hamburg/SAO Survey, Ugryumov et al. 2001, and references therein) blue compact galaxies. Parameters of these observations are shown in Table 1. The setup used for TWIN was the T07 grating in second order for the blue and T06 in first order for the red arm, that provided dispersions of 54 Å mm^-1 and 36 Å mm^-1 respectively. We have used the CCD detectors SITE12a-11 and SITe6a-11 for the blue and red arms with the 5500 Å beam splitter and a slit width of $2\farcs1$ for UM 151 and $1\farcs2$ for UM 408. The resulting FWHM spectral resolution measured on strong lines were 3.1 Å and 2.5 Å in the blue and red, for UM 151, and 2.9 and 2.6 Å, for UM 408. The scale along the slit was $0\farcs56$ pix^-1.

Figure 1: 1D-spectra in the observed wavelength scale for the three galaxies discussed in the paper. In the bottom of each spectra the scaled down versions are drawn to show the relative intensities of strong lines. For UM 151 the continuum with Balmer absorption lines is shown separately, shifted down along the ordinate (Y) axis by two flux units (marked as "Y-2'').

The spectroscopic data for A 1228+12 (RMB 132) were obtained with the 6 m telescope of the Special Astrophysical Observatory of the Russian Academy of Sciences (SAO RAS). Parameters of these observations are shown in Table 1. The long-slit spectrograph (LSS) (Afanasiev et al. 1995) was used with a Photometrics CCD detector of 1024$\times$1024 pixels with a $24\times24~\mu$m pixel size. Observations were conducted mainly with the software package NICE in MIDAS, as described by Kniazev & Shergin (1995). The scale along the slit was $0\farcs39$ pix^-1. A grating with 651 grooves mm^-1 and a slit width of $2\arcsec$ were used, giving a FWHM spectral resolution of 8 Å. Two 0.5-hour spectra were obtained, one after the other, each in the spectral range of 3700-6100 Å and 5000-7400 Å, with the same pointing and long slit orientation. Spectra were extracted from the same region and the two spectra were combined to get the full spectrum of the object for analysis.

For each night we obtained biases, flat-fields and illumination correction images. Comparison lamps of Fe-Ne and Ar-Ne-He were used for wavelength calibration for the Calar Alto and SAO data, respectively. For flux calibrations, spectrophotometric standard stars from Bohlin (1996) for the 6 m telescope observations and Oke (1990) for the 3.5 m telescope were used. Average sensitivity curves were produced for each night with rms deviations of $\sim$5% in the whole spectral blue + red range.

2.2 Reduction

Standard reduction procedures were used with the IRAF package. Once 2D spectra were wavelength calibrated and sky subtracted, flux calibration was performed by using the average sensitivity curves.

The 1D spectra were extracted with the apertures of $3.6\times2.0\arcsec$, $6.2\times2.1\arcsec$ and $6.2\times2.1\arcsec$, respectively for the galaxies A 1228+12, UM 151 and UM 408. The 1D final spectra are shown in Fig. 1. The continuum determination and the measurements of the flux and equivalent width (EW) of spectral lines were performed with MIDAS (for details, see e.g., Kniazev et al. 2000). EWs for individual emission lines were measured with the standard MIDAS procedure INTEGRATE/LINE. The flux and equivalent width of the blended lines were measured using Gaussian decomposition fitting. In both cases the background was drawn by two methods: manually and with the use of the automatic procedure, with the help of the algorithm, described in detail by Shergin et al. (1996). The results of both cases were compared. The errors of the sensitivity curve and those of the line intensities have been combined in quadrature and propagated to calculate element abundances.

In particular, for the A 1228+12 spectrum with $\sim$7 Å resolution, deblending was performed for H$\gamma$/[O III]$\lambda$4363, H$\alpha$/[N II]$\lambda$6548, 6584, [O I]$\lambda$6300/[S II]$\lambda$6312. With these procedures the redshift and the line width were derived first for the stronger line of the blend. For the Gaussian fitting of the fainter blend components, these parameters have been fixed with the values derived for the stronger component. For the fitting of [N II]$\lambda$6548, 6584 lines we also fixed the intensity ratio of the two lines as 1:3, expected from theory (e.g. Aller 1984). The uncertainties of these fitted values were determined from the residual noise of the spectrum near the lines under analysis. These uncertainties were combined, as well as for all other measured line intensities, with the other error components (see below). Therefore, the derived errors of [N II] lines can be large, and their intensity ratio in the table is theoretical. For the TWIN spectra the spectral resolution is sufficient to measure each line separately. While in the spectrum of UM 408 [N II]-lines are rather faint, their line ratio is occasionally close to the theoretical value. In addition to the noise of the underlying continuum, quoted errors in the line intensities include two more components: one comes from Poisson statistics of photon flux, the other comes from the uncertainties of the sensitivity curve, contributing a few per cent to all lines. The line intensity errors presented in Table 2 incorporate all three components, and thus should be reliable estimators for other derived physical parameters and chemical abundances in the H II regions considered. An independent check of the reliability of the cited errors is the good consistency between our results and the results of Kinman & Davidson (1981) for A 1228+12 (see Sect. 4.3). Both sets of line intensities are consistent within the cited errors, if their extinction correction is accounted for. The latter can be overestimated, since Kinman & Davidson indicate a mismatch in the continuum level for independent red and blue spectra. Another factor leading to small differences is that Kinman & Davidson did not account for the underlying Balmer absorption. In the present work we determined underlying Balmer absorption at H$\beta$ of EW(H$\beta$) $\sim$ 2.5 Å. We derived from the spectrum of A 1228+12 the value of C(H$\beta$) = 0. This is consistent within the cited uncertainties ( $\sigma_{C({\rm H}\beta)}=$ 0.10) with the minimum value of C(H$\beta$) = 0.043 following from the Galaxy extinction in this direction, A_B= 0.12 mag (see Table 4). We have checked the effect of the change of C(H$\beta$) from zero to 0.043 on the derived element abundances. The O/H value does not change at all. The values of log (N/O), (Ne/O), (S/O) and (Ar/O) change by only 0.02-0.03 dex, which is many times smaller than their cited uncertainties.

 

 

Table 3: Abundances in the studied galaxies.

Value	A 1228+12	UM 408
$T_{\rm e}$(OIII)(K)	16100 $\pm$ 1100	14800 $\pm$ 1400
$T_{\rm e}$(OII)(K)	14300 $\pm$ 900	13600 $\pm$ 1200
$T_{\rm e}$(SIII)(K)	15100 $\pm$ 900	14000 $\pm$ 1200
$N_{\rm e}$(SII)(cm-3)	10 $\pm$ 10	10 $\pm$ 10
O+/H+($\times$105)	0.936 $\pm$ 0.186	2.518 $\pm$ 0.689
O++/H+($\times$105)	4.198 $\pm$ 0.770	6.050 $\pm$ 1.546
O+++/H+($\times$105)	0.254 $\pm$ 0.098	-
O/H($\times$105)	5.388 $\pm$ 0.798	8.568 $\pm$ 1.693
12+log(O/H)	7.73 $\pm$ 0.06	7.93 $\pm$ 0.09
N+/H+($\times$107)	2.470 $\pm$ 1.472	5.491 $\pm$ 1.439
ICF(N)	5.758	3.402
log(N/O)	-1.58 $\pm$ 0.27	-1.66 $\pm$ 0.14
Ne++/H+($\times$105)	0.737 $\pm$ 0.143	1.369 $\pm$ 0.381
ICF(Ne)	1.284	1.416
log(Ne/O)	-0.76 $\pm$ 0.11	-0.65 $\pm$ 0.15
S+/H+($\times$107)	1.658 $\pm$ 0.232	2.990 $\pm$ 0.537
S++/H+($\times$107)	7.212 $\pm$ 4.560	13.490 $\pm$ 10.660
ICF(S)	1.709	1.357
log(S/O)	-1.55 $\pm$ 0.23	-1.58 $\pm$ 0.29
Ar++/H+($\times$107)	1.500 $\pm$ 0.327	--
Ar+++/H+($\times$107)	1.957 $\pm$ 1.023	--
ICF(Ar)	1.026	--
log(Ar/O)	-2.18 $\pm$ 0.15	--

   

Table 4: Main parameters of studied galaxies.

Parameter	UM 151	UM 408	A 1228+12
$\alpha_{\rm 2000}$	01 57 38.87	02 11 23.55	12 30 48.52
$\delta_{\rm 2000}$	+02 25 23.9	+02 20 31.0	+12 02 42.1
ABN	0.12	0.15	0.12
$B_{\rm tot}$	16.21(1)	17.74(1)	17.15(3)
$V_{\rm Hel}$ (km s$^{\rm -1}$)	4851(2)	3507(4)	1263(5)
Dist (Mpc)	64.7	46.8	17.0V
MB0 (6)	-17.96	-15.76	-14.10
Opt. size ($\arcsec$)(7)	35$\times$15.3(1)	15.6$\times$10.6(4)	12$\times$9(3)
Opt. size (kpc)	11.0$\times$2.4(2)	3.5$\times$2.4(2)	1.0$\times$0.75(2)
12+log(O/H)	8.5(2)	7.93(2)	7.73(2)
H I flux(8)	<1.2(9)	1.5(4)	1.4(5)
$W_{\rm 50}$ (km s$^{\rm -1}$)	--	77(4)	84(5)
M(H I) (10 $^{(8)} M_{\odot}$)	<11.9(2)	7.8(2,4)	0.95(2,5)
M(H I)/L$_{\rm B}$(10)	<2.0(2)	2.5(2)	1.5(2)

(1) - Salzer et al. (1989b)(V₂₅-isophote, b/a - minor-to-major axis ratio, from LEDA).
(2) - parameters derived in this paper.
(3) - Binggeli & Cameron (1993).
(4) - Smoker et al. (2000).
(5) - Staveley-Smith et al. (1992).
(6) - corrected for the Galaxy extinction.
(7) - $a \times b$ at $\mu_{B} =$ 25 mag/sq. arcsec. See note (1).
(8) - in units of (Jy$\cdot$km s$^{\rm -1}$).
(9) - Thuan et al. 1999; upper limit is estimated for $W_{\rm 50}=100$ km s$^{\rm -1}$.
(10) - in units of (M/L_B)$_{\odot}$.
(V) - accepted for the Virgo cluster (Tikhonov et al. 2000).
(N) - data from NED, Schlegel et al. (1998).