© ESO, 2011

1. Introduction

M 43 (NGC 1982) is an apparently spherical H ii region several arcmin to the northeast of the well-known Orion nebula. HD 37061 (NU Ori, Par 2074, Brun 747), an early B-type star with broad lines located at the center of the nebula, is the main ionizing source of M 43. The spectral classifications found in the literature for HD 37061 range between B0.5 V and B1 V, probably because of the difficulty in detecting the faint He ii 4541 line when using low resolution spectra and photographic plates. Up to three components have been identified within this stellar system (Abt et al. 1991; Preibisch et al. 1999). The primary star is known to be a spectroscopic binary (SB1) with an estimated stellar mass ratio M₂/M₁  ~  0.19 (Abt et al. 1991). Using bispectrum speckle interferometry, Preibisch et al. (1999) found a third companion at  ~470 mas with a flux ratio in the K-band of 0.03  ±  0.02, implying a stellar mass ratio M₃/M_1,2 ~ 0.07. Although the less massive components may affect the spectrum of the primary, they can be neglected in terms of ionization of the nebula.

M 43 belongs to the same molecular complex as the Orion nebula (Goudis 1982). In particular, M 43 is located at the northeast border of the extended Orion nebula (EON), a large elliptical structure surrounding the Huygens region (corresponding to the central, brighter part of the Orion nebula). Although the precise structure of the EON is still unknown, recent studies (O’Dell & Goss 2009; O’Dell & Harris 2010) have associated the diffuse¹ and extended emission arising from this region with scattered light, produced by dust reflecting the stellar continuum emission from the Trapezium cluster stars and the nebular line emission from the Huygens region. A dust lane defines the outer northern boundary of the EON and separates M 43 from the Huygens region. As inferred from the optical polarization map of the M 43 region obtained by Khallesse et al. (1980), this dust lane is not a foreground obscuration but a wall of dense material between M 42 and M 43. This wall of material ensures that it is quite unlikely that ionizing light from the Trapezium cluster stars reaches the nebular material ionized by HD 37061; however, owing to the proximity of M 43 to the assumed northern boundary of the EON, a scattered light contribution (similar to the one described above) may still affect affect the nebular emission received from M 43.

Most studies of M 43 have been performed at radio and far-infrared wavelengths. These studies were centered on the investigation of the global properties of the nebula (i.e. geometry, T_e and n_e, Mills & Shaver 1968; Thum et al. 1978; Thronson et al. 1986; Subrahmanyan 1992), and the characteristics of the dust re-radiation inside the nebula (Thum et al. 1978; Smith et al. 1987). As for optical images, radio continuum observations indicate that M 43 has a circular shape (Subrahmanyan 1992). In addition, images at 60 μm published by Thronson et al. (1986), and the kinematical study by Hänel (1987) allow one to associate the nebula with a shell-shaped cavity on the side of a dense molecular ridge.

Its proximity², relatively high surface brightness, simple geometry, and isolated ionizing source make the Mairan’s nebula an ideal object to investigate several topics of interest in the field of H ii regions and massive stars. To this aim, we collected a set of high quality observations comprising the optical spectrum of HD 37061, along with optical imaging and long-slit spatially resolved spectroscopy of M 43. These observations were used to perform a comprehensive study of M 43 and its ionizing star, which is presented in two parts. In this first paper, we (1) obtain the stellar parameters of HD 37061, (2) investigate the presence of the scattered light in the region of M 43 and its effect on the derived nebular properties, and (3) determine the physical conditions and nebular abundances in M 43 obtained at various distances to the center of the nebula. All this information is then used in a second paper (Simón-Díaz et al., Paper II, in prep.) to construct a specialized photoionization model of the nebula using as many observational constraints as possible. The main drivers of this second part of our study are to (1) test the reliability of the ionizing spectral energy distributions (SEDs) provided by the modern stellar atmosphere codes using nebular constraints, (2) investigate the effect of stellar pumping (Ferland 1999; Luridiana et al. 2009) on the nebular emission arising from the inner part of this H ii region, and (3) test our understanding of the nebular energy budget temperature distribution.

This paper is structured as follows. The observational data set is presented in Sect. 2. A quantitative spectroscopic analysis of the optical spectrum of HD 37061 is performed in Sect. 3. The morphological characteristics of the nebula, along with its physical conditions and nebular abundances are determined in Sects. 4 and 5 by analyzing of the Hα, Hβ, [O iii], and [S ii] narrow-band images, and the nebular spectra, respectively. Both nebular analyses also allow us to find several indications of an extended nebular emission, not directly related to M 43, which must be adequately subtracted for a correct interpretation of the nebular images and spectroscopy. Finally, we outline the main results of this study in Sect. 6.

2. The observational data set

2.1. Nebular imaging

The Wide Field Camera³ (WFC) attached to the Isaac Newton Telescope (INT) at the Roque de los Muchachos Observatory (La Palma, Spain) was used to obtain several narrow-band filter images of M 43 in 2004 February 16, and 2006 October 21. The WFC-INT consists of four thinned EEV 2k  ×  4k CCDs located at the primary focus of the telescope. Each CCD yields a field of view of 11.2′ × 22.4′, which results in a edge to edge limit of the mosaic (neglecting the  ~1 arcmin inter-chip spacing) of 34.2 arcmin. The pixel size is 13.5 μm corresponding to 0.33 arcsec/pixel.

We centered M 43 on chip#4 of the WFC, and obtained images of the region using the narrow-band filters Hβ, [O iii] λ5007, Hα, and [S ii] λλ6716+30. The log of the observations, along with the characteristics of the various filters can be found in Table 1. While the [O iii] and [S ii] images were only obtained to provide some qualitative information about the nebula, we wished to produce pure emission-line, flux-calibrated images for Hα and Hβ. We hence obtained these images during a photometric night, and included the Hα redshifted, and Hβ broad filters in our set of observations, to correct the Hα and Hβ images for the adjacent continuum. The exposure times considered for the images in the various filters were 4 × 60 s for Hα, 4 × 120 s for Hα continuum, 3 × 120 s for Hβ, 3 × 180 s for Hβ continuum, 5 × 12 s for [O iii], and 5 × 12 s for [S ii].

The reduction of the images was performed following the standard procedures (trimming, bias subtraction, flat-fielding, and alignment) with iraf⁴. Hα and Hβ images were flux calibrated using observations of the spectrophotometric standard star BD +28°4211 (Landolt & Uomoto 2007) at different airmasses. We obtained the continuum-subtracted Hα and Hβ images following the procedure described in López-Sánchez & Esteban (2008).

The large field of view of the WFC also allowed us to include M 42 in the images. A color-composite image of the M 43+M 42 region obtained from a combination of the [O iii] (blue), Hα (green), and [S ii] (red) images is presented in Fig. 1. Individual images in each one of these three filters, showing a square region of  ~6 arcmin centered in M 43, are presented in Fig. 2.

2.2. Stellar spectroscopy

The spectroscopic observations of HD 37061 were carried out with the INT on 2006 August 30 and September 5. The Intermediate Dispersion Spectrograph (IDS) was used with the 235 mm camera and two different gratings (see Table 1). We observed the spectral region between λλ4000 and 5050 Å using the H2400B grating, which resulted in an effective resolving power R    ~   7500 (equivalent to a 0.24 Å/pixel resolution and  ~2.6 pixel FWHM arc lines). For the Hα region (~λλ6090–6760 Å), the H1800V grating was used, resulting in a similar spectral resolution (0.31 Å/pixel, R    ~   8000). With these configurations, three exposures were needed to cover the whole range. We obtained spectra with exposure times of 180 s and 70 s for the blue (2) and red (1) configurations, respectively, and obtained two spectra per spectral range. A large number of flat fields and arcs for the data reduction process were obtained.

The reduction and normalization of the spectra was made following standard techniques, with iraf and our own software developed in IDL. The signal-to-noise ratio of the reduced spectra is about 300–400, depending on the spectral range. Figure 3 shows the global stellar spectrum, along with the H i and He i-ii lines used for the quantitative spectroscopic analysis.

2.3. Nebular spectroscopy

A long-slit, intermediate-resolution spectrum of M 43 was obtained on 2006 December 23 with the Intermediate dispersion Spectrograph and Imaging System (ISIS) spectrograph attached to the 4.2m William Herschel Telescope (WHT) at Roque de los Muchachos Observatory (La Palma, Spain). The 3.7′  ×  1.02′′ slit was located to the west of HD 37061, along the E-W direction (PA = 90°, see Fig. 2). Two different CCDs were used at the blue and red arms of the spectrograph: an EEV CCD with a configuration 4096  ×  2048 pixels at 13.5 μm, and a RedPlus CCD with 4096  ×  2048 pixels at 15.0 μm, respectively. The dichroic used to separate the blue and red beams was centered on 5400 Å. The gratings R600B and R600R were used for the blue and red observations, respectively (see log of the observations in Table 1). These gratings give a reciprocal dispersion of 33 Å mm^-1 in both cases, and effective spectral resolutions of 2.2 and 2.0 Å, respectively. The blue spectra cover from λλ3386 to 5102 Å and the red ones from λλ6064 to 7585 Å. The spatial scales are 0.20′′ pixel^-1 and 0.22′′ pixel^-1, respectively. The seeing during the observations was between  ~0.5′′ and  ~0.8′′. The exposure times were 3  ×  300 s, in both the blue and red observations (obtained at the same time).

The spectra were wavelength calibrated with a CuNe+CuAr lamp. The correction for atmospheric extinction was performed using the average curve for continuous atmospheric extinction at Roque de los Muchachos Observatory. The absolute flux calibration was achieved by observations of the standard stars HD 19445, H 600, and HD 84937. We used the standard iraf TWODSPEC reduction package to perform bias correction, flat-fielding, cosmic-ray rejection, wavelength, and flux calibration. Figure 4 shows an illustrative example of the global appearance of our nebular spectroscopic observations, also indicating the main nebular lines used in this study.

3. Quantitative spectroscopic analysis of HD 37061

The stellar parameters of the star were derived by visually comparing the observed optical H i, He i, and He ii line profiles with the synthetic ones resulting from the stellar atmosphere code FASTWIND (Santolaya-Rey et al. 1997; Puls et al. 2005), which is an established technique. Similar analyses, along with some notes on the methodology, can be found in Herrero et al. (2002), Repolust et al. (2004), Simón-Díaz et al. (2006), and references therein.

To this aim, we constructed a grid of FASTWIND models with T_eff, and log g, ranging from 28 000 to 33 000 K (500 K steps), and from 3.9 to 4.3 dex (0.1 dex steps), respectively. The microturbulence, the helium abundance, and the wind-strength parameter⁵ were fixed to characteristic values for an early B dwarf star (ζ_t = 5 km s^-1, ϵ = 0.09, and log Q =  −15, respectively) and metal abundances were assumed to be solar (following the set of abundances by Asplund et al. 2009).

This technique requires the projected rotational velocity (v sin i) of the star to be previously established. All metal lines in the spectrum of HD 37061 are blended or very shallow because of the large v sin i of the star; therefore, we decided to apply the Fourier method (cf. Gray1976; see also Simón-Díaz & Herrero 2007, for a recent application to OB-type stars) to the He i lines, and obtained a v sin i  ~  200 km s^-1.

The synthetic H i and He i-ii line profiles resulting from the models were then convolved with the corresponding instrumental and rotational profiles and compared with the observations. The best fit was found for T_eff = 31 000 K and log g = 4.2 dex (see Fig. 5). Given the quality of the spectrum and the sensitivity of the used lines to the variation in T_eff and log g for this range of stellar parameters, an accuracy of 500 K and 0.1 dex, respectively, could be achieved. As an example, in Fig. 5 we show the effect of a variation of  ±1000 K in the FASTWIND models. For this range of stellar parameters, the He ii lines are decisive in constraining the T_eff.

We note the poor quality of the fits in the red wings of the H i and He i lines (but not in the He ii lines). These lines are likely affected by the spectrum of the cooler secondary spectroscopic component (see Sect. 1).

Once the stellar parameters have been determined, the stellar radius can be derived from the absolute visual magnitude (M_V) and the synthetic spectral energy distribution provided by the stellar atmosphere model. By using m_V = 6.84 and A_V = 2.09 (Hillenbrand 1997), and adopting a distance to the star of 400  ±  50 pc, we obtained an M_V =  −3.3 ± 0.3, quite consistent with the spectral type of HD 37061. The resulting stellar parameters, along with the total number of ionizing photons are summarized in Table 2.

4. Qualitative and quantitative analysis of the nebular images

Figure 1 shows a color composite image of the M 42+M 43 region. The object under study, M 43, appears as a roundish H ii region centered around HD 37061, and well-separated from the Orion nebula (M 42) by a dust lane (known as the northeast dark lane). The nebula has a diameter of  ~4.5 arcmin (~0.65 pc at a distance of 400 pc), and is crossed by a dark lane oriented N-S in the eastern side (known as the M 43 dark lane). This dark cloud is located in front of the nebula, blocking the nebular light coming from behind.

Figure 2 shows a closer view of the nebula, where three images of M 43, taken through filters isolating Hα (+[N ii] λλ6548+84), [S ii] λλ6716+30, and [O iii] λ5007, are presented separately. The images show diffuse and extended emission beyond the limits of the nebula (mainly to the south and west). This emission can be more clearly seen in the [O iii] image. If the entire nebula were spherical with a density of 500 cm^-3, the size of the O²⁺ region would be  ~25% of the total size of the nebula⁶; however, Fig. 2 shows that the [O iii] emission originates in a more extended region (i.e. is unlikely to be related⁷ to nebular material ionized by HD 37061).

To more clearly illustrate this, we plot in Fig. 6 (three upper panels) the spatial distribution of Hα, [S ii], and [O iii] emission in a line passing by HD 37061 along the east-west direction, obtained from the corresponding WFC-INT images. Since we are only interested in the relative spatial changes, the three distributions are normalized to their maximum value. In all cases, a faint emission, not expected to be related to M 43, is clearly detected beyond the limits of the nebula. In the [O iii] panel, we also superimpose the spatial distribution derived from the Hα-continuum image to show that the extended emission observed in the [O iii] image is not associated with the nebular continuum.

4.1. Extinction

The flux-calibrated, continuum-subtracted Hα and Hβ images (i.e. pure emission) were used to obtain the Hα/Hβ flux ratio distribution along the diameter in the east-west direction passing by the ionizing star of M 43 (see bottom panel in Fig. 6). Some interesting information about the extinction in the studied region can be extracted by inspecting of this figure:

• a)

  there is a jump in the Hα/Hβ flux ratio 140 arcsec west of HD37061 (i.e. western boundary of M 43), indicating a clear difference in the type of emission arising from outside the limits of the nebula. We note that the lower Hα/Hβ ratio found in the “extended emission” region agrees with the hypothesis that this emission is associated with scattered light (see also Sect. 5.9), and is not necessarily an indication of a lower extinction in this region;

• b)

  the Hα/Hβ ratio in the western region of M 43 is fairly constant, though there is a small decrease close to the star. On the other hand, this ratio is larger in the eastern part;

• c)

  the M 43 dark lane location (labeled as a foreground dust lane in Fig. 6) is clearly represented in the Hα/Hβ distribution.

Smith et al. (1987) presented a general view of the spatial distribution of dust in the M 43 region. They found three regions where dust reradiation in 60 μm is concentrated (see a schematic map of dust features in their Fig. 1c). The first one is associated with the M 43 dark lane (they called it the foreground dust lane), the second one (M 43 east) is located to the east of HD 37061, between the ionizing star and the M 43 dark lane, and a third one (M 43 front + M 43 west) follows the border of the nebula from the north to the west. This third feature of dust emission is separated  ~ 60–70 arcsec from the star, leaving a “dust-empty” space in-between. We indicate in the bottom panel of Fig. 6 the location of these dust features using the same nomenclature as Smith et al.

The observed behavior of the Hα/Hβ flux ratio within M 43 along the west-east direction is perfectly correlated with the dust distribution presented by Smith et al. (1987). It is remarkable that, in contrast to expectations, the amount of extinction indicated by the Hα/Hβ flux ratio in the foreground dust lane is very low. The explanation is simple. This dark feature obscures the emission from M 43, hence the observed light originates in the material located in front. This emission follows the trend indicated by the region marked as “extended emission”.

4.2. Total nebular Hα luminosity and surface brightness

To derive the total nebular Hα luminosity we proceeded as follows. First, the Hα image was corrected for the effects of distance, extinction, and [N ii] contamination⁸. Denoting with F_Hα the nebular flux received at Earth, the corresponding extinction-corrected Hα luminosity can be derived from: ${\mathit{L}}_{\mathrm{H}\mathit{\alpha }}\mathrm{=}\mathrm{4}\mathit{\pi }{\mathit{d}}^{\mathrm{2}}{\mathit{F}}_{\mathrm{H}\mathit{\alpha }}{\mathrm{10}}^{{\mathit{c}}_{\mathrm{H}\mathit{\beta }}\hspace{0.17em}\mathit{f}\mathrm{\left(}\mathrm{H}\mathit{\alpha }\mathrm{\right)}}\mathit{,}$(1)where d is the distance to the nebula, c(Hβ) is the reddening coefficient, and f(Hα) is the value of the extinction function in Hα relative to Hβ.

The analysis of the Hα/Hβ image (Sect. 4.1) showed that the extinction is not constant across the nebula. The optimal strategy to follow would be to obtain c(Hβ) for every pixel from the Hα/Hβ images, correct the Hα image, and then integrate the whole nebula to obtain the total L(Hα). However, when we attempted to follow this strategy we found that it introduces many sources of uncertainty. In addition, the nebular region hidden by the dark lane cannot be corrected in this way. We therefore decided to follow a different approach, taking into account what we learned from the inspection of the nebular images, and making use of the information extracted from the spectroscopic observations (see Sect. 5). We decided to integrate only the west quadrant of the nebula, and multiply the resulting value by four (i.e. we assume the nebula is symmetric). Extinction in this quadrant is lower and more constant than in the other three. Furthermore, we have information about c(Hβ) (~0.76) and the nitrogen contribution to the Hα image ([N ii] λλ6548+84/Hα  ~  0.59) for this region from the ISIS-WHT spectroscopy (see Sect. 5).

Once the distance, extinction, and nitrogen corrections had been applied to the continuum-subtracted, flux-calibrated Hα image, we obtained the L(Hα) inside circles of increasing sizes⁹ (centered on the star), and the surface brightness distribution within concentric rings of increasing radii. Figure 7 shows the measured quantities as a function of distance to the star. Dotted lines show the initial determinations where the images were not corrected for the extended emission contribution. The Hα surface brightness distribution again shows that outside the limits of the nebula (θ    ≥    160 arcsec) there remains a non-negligible, more or less constant emission. This implies that L(Hα) increases continuously, even outside the nebula (top panel in Fig. 7). We therefore assumed a constant¹⁰ contribution of the extended emission to the surface brightness (~0.06 erg s^-1 cm^-2, see middle panel of Fig. 7) and subtracted it from the Hα image. The resulting L(Hα) and S(Hα) distributions, calculated from the corrected images in the same way as described above, are indicated as solid lines in the top and middle panels of Fig. 7. These quantities behave as expected, i.e., the surface brightness becomes zero, and L(Hα) remains constant for R > R_neb. The resulting total Hα luminosity is (3.3 ± 1.1) × 10³⁵ erg s^-1, where the associated uncertainty was calculated by taking into account uncertainties of 12.5%, 10%, and 20% in distance, F(Hα), and c(Hβ), respectively.

The bottom panel of Fig. 7 shows the c(Hβ) values obtained from the spectroscopic analysis of the 9 apertures extracted from the long-slit ISIS-WHT observations (see Sect. 5.2). As for the Hα/Hβ images (bottom panel in Fig. 6), we found that c(Hβ) is fairly constant for θ    >    50 arcsec, but is somewhat smaller in the inner region (close to the star). We also include some information obtained from the spectroscopy in the middle panel of Fig. 7 (surface brightness profile). Red triangles correspond to the calculated S(Hα) in each aperture assuming a constant c(Hβ) of 0.76. The derived values follow the S(Hα) profile obtained from the Hα image. The middle panel in Fig. 7 also shows the corresponding S(Hα) values computed from the actual c(Hβ) values derived from the spectroscopic analysis (black diamonds). These values deviate from the constant extinction-corrected values in the inner part of the nebula. We can conclude that the increase in S(Hα) in the inner part of the nebula is not necessarily caused by an increase in the emission measure or a deviation from the spherical geometry, but is more likely to be an effect of a non-constant extinction or stellar pumping effects on H recombination lines (see Ferland 1999; Luridiana et al. 2009). The total L(Hα) we derived above must hence be corrected for this effect, resulting in (3.0 ± 1.1) × 10³⁵ erg s^-1 (about 10% below the previously derived value).

In Sect. 3, we derived the total number of ionizing photons emitted by the central star (Q(H⁰) = 10^{47.2 ± 0.2} photons s^-1). This would imply a maximum nebular Hα luminosity of (2.5 ± 1.0) × 10³⁵ erg s^-1, a value that is in quite good agreement with the total Hα luminosity obtained from the analysis of the nebular images. This result is compatible to first order with M 43 being a (mostly) ionization-bounded nebula in which dust does not absorb any significant fraction of the Lyman continuum photons.

We note that the stellar and nebular results would be incompatible if we considered the total nebular Hα luminosity obtained from the images that are not corrected for the extended emission contribution. The total number of ionizing photons necessary to explain the derived nebular Hα luminosity would be at least double the value obtained from the analysis of the star.

5. Empirical analysis of the nebular spectra

5.1. More indications of extended nebular emission

Figure 8 shows the spatial distribution of the nebular emission in four selected lines (Hα, [O ii] λλ3720+30, He i λ6678, and [O iii] λ5007) and their respective adjacent nebular continua. These distributions were obtained from the ISIS@WHT spectroscopic observations. The colored lines represent the measured line fluxes (blue), the adjacent continuum emission (red), and the continuum-subtracted line fluxes (black). These figures display non-negligible emission beyond the limits of the nebula (θ    >    150 arcsec), that is not associated with the continuum emission. In most cases, this external emission remains fairly constant in the studied region (e.g. Hα, [O ii] λλ3720+30, He i λ6678, as well as [N ii], [S ii], and [S iii] lines). However, in some cases (e.g. [O iii]), this emission increases with increasing distance from the star.

5.2. Aperture selection and line flux measurements

We obtained one-dimensional (1D) spectra of regions of the nebula at different distances from the central star by dividing the long-slit used for the ISIS-WHT nebular observations into nine small apertures within the limits of the nebula (θ < 150 arcsec) with the aim of obtaining: a) the radial distribution of the physical properties of the nebula, b) information about nebular ionic and total abundances, and c) the radial distribution of critical nebular line ratios used as observational constraints on photoionization models of the nebula.

The size and position of the apertures are summarized in Table 3 and shown in Figs. 2 and 8. It is crucial to correct these spectra for the contribution of the extended emission component to obtain information about M 43 itself. To achieve this aim, spectra from two extra-apertures outside the nebula were extracted to correct the line flux measurements of the apertures for the extended emission component. The size of the extra-apertures was selected to be the same as the corresponding apertures inside the nebula (11 and 16.5 arcsec for the three innermost and the other outer apertures, respectively). Two different locations of these extra-apertures were considered depending on whether the considered nebular emission line had been emitted in the whole nebula (i.e. H i, [O ii], [N ii], [S ii] and [S iii] lines; apertures v1 and v2 in Fig. 8), or only in the inner region (i.e. He i, [O iii], and [Ar iii] lines; apertures v3 and v4 in Fig. 8). For the latter set of lines, it is impossible to use the same apertures (v1 and v2) as for the other lines, since the extended emission in these cases is not constant, but increases far away from the star. We hence considered two apertures as being close to the region where we were confident that the contribution of the nebular emission associated with M 43 had vanished. We expect the spectrum from these apertures to represent the properties of the extended emission close to the region where M 43 is emitting in the given nebular line.

The extraction of 1D spectra from each aperture was performed using the iraf task apall. The same zone and spatial coverage was considered in the blue and red spectroscopic ranges.

We detected H i and He i optical recombination lines, along with collisionally excited lines (CELs) of several ions, such as [O i], [O ii], [O iii], [N ii], [S ii], [S iii], [Ne iii], and [Ar iii]. Line fluxes were measured using the SPLOT routine of the iraf package by integrating all the flux included in the line profile between two given limits and over a local continuum estimated by eye. Although we detected two [Ne iii] lines, these were unsuitable for analysis because they were extremely faint (in the case of λ3869) or were severely blended with H iλ3967. In the case of [O ii] λ7330 Å, we corrected the flux for the overlapping telluric emission line OH 8–3 P2 λ7329.148 Å by estimating its intensity from the observed OH 8–3 P2 λ7329.148 Å/OH 8–3 P1 λ7340.885 Å line ratio in standard star exposures.

Each emission line in the spectra from the nine apertures was corrected for the extended emission component by subtracting the corresponding emission measured in the appropriate extra-aperture (v1 to v4). Line intensities were then normalized to a particular H i recombination line present in each wavelength interval (Hβ and Hα for the blue and red spectra, respectively). To produce a final homogeneous set of line intensity ratios, the red spectra were then rescaled to Hβ using the theoretical Hα/Hβ ratio I(Hα)/I(Hβ) = 2.92 obtained for the physical conditions of T_e = 7500 K and n_e = 500 cm^-3 (see Sect. 5.5).

5.3. Extinction correction

We assume the extinction law derived by Blagrave et al. (2007) for the Orion nebula. Both H ii regions have a deficiency in small particles that produces the relatively grey extinction observed in Orion (Magazzu et al. 1986; Baldwin et al. 1991). Furthermore, Rodríguez (1999, 2002) did not find large extinction variations in her optical spectroscopic studies of the M 42 and M 43 H ii regions.

The reddening coefficient, c(Hβ) was obtained by fitting the observed Hδ/Hβ and Hγ/Hβ line intensity ratios – the three lines lie in the same spectral range – to the theoretical ones computed by Storey & Hummer (1995) for T_e = 7500 K and n_e = 500 cm^-3. As pointed out by Mesa-Delgado et al. (2009a,b), the use of a Blagrave et al. (2007) law instead of the classical one by Costero & Peimbert (1970) produces slightly higher c(Hβ) values and also slightly different dereddened line fluxes depending on the spectral range.

The final set of line intensities corrected for both the extended emission component and the extinction are indicated in Table 3.

5.4. Uncertainties

Several sources of uncertainties must be taken into account to obtain the errors associated with the line intensity ratios. We estimated that the uncertainty¹¹ in the line intensity measurement due to the signal-to-noise ratio of the spectra and the placement of the local continuum is typically $\begin{array}{ccc}\mathrm{~}\mathrm{2}\mathrm{\%} & \mathrm{for}& \mathit{F}\mathrm{\left(}\mathit{\lambda }\mathrm{\right)}\mathit{/}\mathit{F}\mathrm{\left(}\mathrm{H}\mathit{\beta }\mathrm{\right)}\hspace{0.17em}\mathrm{\ge }\hspace{0.17em}\mathrm{0.5}\mathrm{;}\\ \mathrm{~}\mathrm{5}\mathrm{\%} & \mathrm{for}& \mathrm{0.1}\hspace{0.17em}\mathrm{\le }\hspace{0.17em}\mathit{F}\mathrm{\left(}\mathit{\lambda }\mathrm{\right)}\mathit{/}\mathit{F}\mathrm{\left(}\mathrm{H}\mathit{\beta }\mathrm{\right)}\hspace{0.17em}\mathrm{\le }\hspace{0.17em}\mathrm{0.5}\mathrm{;}\\ \mathrm{~}\mathrm{10}\mathrm{\%} & \mathrm{for}& \mathrm{0.05}\hspace{0.17em}\mathrm{\le }\hspace{0.17em}\mathit{F}\mathrm{\left(}\mathit{\lambda }\mathrm{\right)}\mathit{/}\mathit{F}\mathrm{\left(}\mathrm{H}\mathit{\beta }\mathrm{\right)}\hspace{0.17em}\mathrm{\le }\hspace{0.17em}\mathrm{0.1}\mathrm{;}\\ \mathrm{~}\mathrm{20}\mathrm{\%} & \mathrm{for}& \mathrm{0.01}\hspace{0.17em}\mathrm{\le }\hspace{0.17em}\mathit{F}\mathrm{\left(}\mathit{\lambda }\mathrm{\right)}\mathit{/}\mathit{F}\mathrm{\left(}\mathrm{H}\mathit{\beta }\mathrm{\right)}\hspace{0.17em}\mathrm{\le }\hspace{0.17em}\mathrm{0.05}\mathrm{;}\\ \mathrm{~}\mathrm{30}\mathrm{\%} & \mathrm{for}& \mathrm{0.005}\hspace{0.17em}\mathrm{\le }\hspace{0.17em}\mathit{F}\mathrm{\left(}\mathit{\lambda }\mathrm{\right)}\mathit{/}\mathit{F}\mathrm{\left(}\mathrm{H}\mathit{\beta }\mathrm{\right)}\hspace{0.17em}\mathrm{\le }\hspace{0.17em}\mathrm{0.01}\mathrm{;}\mathrm{and}\\ \mathrm{~}\mathrm{40}\mathrm{\%} & \mathrm{for}& \mathrm{0.001}\hspace{0.17em}\mathrm{\le }\hspace{0.17em}\mathit{F}\mathrm{\left(}\mathit{\lambda }\mathrm{\right)}\mathit{/}\mathit{F}\mathrm{\left(}\mathrm{H}\mathit{\beta }\mathrm{\right)}\hspace{0.17em}\mathrm{\le }\hspace{0.17em}\mathrm{0.005.}\end{array}$We did not consider any lines of weaker intensity than 0.001 × F(Hβ). We note that then uncertainties indicated in Table 3 refer inly to this type of errors.

By comparing the resulting flux-calibrated spectra of one of our standard stars with the corresponding tabulated flux, we could estimate that the line ratio uncertainties associated with the flux calibration is  ~3% when the wavelengths are separated by 500–1500 Å and  ~5% if they are separated by more than that. Where the corresponding lines are separated by less than 500 Å, the uncertainty in the line ratio due to uncertainties in the flux calibration is negligible.

The uncertainty associated with extinction correction was computed by error propagation. The contribution of this uncertainty in the total error is again negligible when line ratios of close-by lines are considered (e.g. [S ii] λ6716/[S ii] λ6730). The final errors in the line intensity ratios used to derive the physical properties of the nebula were computed by quadratically adding these three sources of uncertainty.

5.5. Physical conditions

The electron temperature (T_e) and density (n_e) of the ionized gas were derived from classical CEL ratios, using the iraf task temden of the package nebular (Shaw & Dufour 1995) with updated atomic data for several ions (see García-Rojas et al. 2009). We computed n_e from the [S ii] λ6717/λ6731 and [O ii] λ3729/λ3726 line ratios and T_e from the nebular to auroral [O ii] λλ(7320+30)/λλ(3726+29) line ratio¹².

We could not directly determine T_e([O iii]) because the [O iii] λ4363 auroral line was not detectedbecause of the low ionization degree of the nebula. As an alternative, we used a empirical relation between T_e([O iii]) and T_e([O ii]) obtained from the nebular information published by García-Rojas & Esteban (2007) for a sample of Galactic H ii regions¹³${\mathit{T}}_{\mathrm{e}}\mathrm{\left(}\mathrm{[}\mathrm{O}\hspace{0.17em}\mathrm{III}\mathrm{]}\mathrm{\right)}\mathrm{=}\mathrm{\left[}{\mathit{T}}_{\mathrm{e}}\mathrm{\right(}\mathrm{[}\mathrm{O}\hspace{0.17em}\mathrm{II}\mathrm{]}\mathrm{)}\mathrm{-}\mathrm{(}\mathrm{2640}\mathrm{\pm }\mathrm{1270}\mathrm{\left)}\mathrm{\right]}\mathit{/}\mathrm{(}\mathrm{0.7105}\mathrm{\pm }\mathrm{0.1524}\mathrm{)}\mathit{.}$(2)This fit is almost identical to that obtained by Pilyugin et al. (2006) and Garnett (1992) for integrated spectra of giant extragalactic H ii regions and photoionization models of giant extragalactic H ii regions, respectively. We only computed T_e([O iii]) for the three innermost apertures. For more distant apertures, the O²⁺ emission (if any) is not associated to M 43 itself, but to the residual emission of the diffuse component (see Sects. 4 and 5.1).

The methodology used to determine the physical conditions was as follows: we assumed a representative initial value of T_e of 10 000 K and computed the electron densities. The value of n_e was then used to compute T_e([O ii]) from the observed line ratios, and T_e([O iii]) from Eq. (2). We iterated until convergence to obtain the final values of n_e and T_e. Uncertainties were computed by error propagation. The final n_e([O ii]), n_e([S ii]), T_e([O ii]), T_e([O iii]) estimate, along with their uncertainties are indicated in Table 3.

In general, densities derived from the [O ii] line ratio are about 100 cm^-3 lower than those derived from [S ii] lines, but consistent within the errors.

5.6. Chemical abundances

To derive He⁺/H⁺, we used two observed lines of He i at λλ4471 and 6678. Case B emissivities were taken from the collision less (low-density limit) calculations by Bauman et al. (2005) using an on line available code¹⁴. The collisional-to-recombination contribution was estimated from Kingdon & Ferland (1995), using the interpolation formula provided by Porter et al. (2007). The effective recombination coefficients for H⁺ were taken from Storey & Hummer (1995).

Ionic abundances of N⁺, O⁺, O²⁺, S⁺, S²⁺ and Ar²⁺ were derived from CELs, using the iraf task ionic of the package nebular. We assumed a two-zone scheme, in which we adopted T_e([O ii]) for the low ionization potential ions (N⁺, O⁺, S⁺ and S²⁺) and the value of T_e([O iii]) derived from Eq. (2) for the high ionization potential ions (O²⁺ and Ar²⁺). In both cases, we assumed n_e([O ii]) instead of n_e([S ii]) because the first indicator is more representative of the whole nebula, while n_e([S ii]) provides insight mostly into the density of the material close to the ionization edge of the nebula. For O²⁺, Ar²⁺, and He⁺, we only have available lines in the three innermost apertures. Owing to the low T_eff of the ionizing star, these relatively high ionization species are only present in regions very close to the star.

We then derived the total abundances of O, N, and S for each aperture¹⁵. As a consequence of the low ionization degree of the nebula, M 43 has a very small amount of O²⁺ in comparison with the dominant species, O⁺, except in the innermost zones where a non-negligible fraction of O²⁺ is present. We note that for the intermediate apertures (apertures 4 to 6), the O abundance is given by O⁺/H⁺. In the case of S, the abundance of this element is given by the sum of S⁺ and S²⁺, because S⁺³ is not expected to be present in this nebula. Similarly to O²⁺, N²⁺ is only expected to be present in the innermost parts of the nebula. For the corresponding apertures (1–3), we applied the classical ionization correction factor (ICF) scheme of N (i.e. N⁺/O⁺ = N/O) to correct for the N²⁺ abundance (note that this correction is very small). As for O, for the intermediate apertures the N abundance is directly obtained from the N⁺ abundance.

The ionic abundances of He⁺ O⁺, O²⁺, N⁺, S⁺, S²⁺, and Ar²⁺, along with the total abundances of O, N, and S are shown in Table 3. We stress again that in this nebula the total abundances of O, N, and S can be obtained directly from observable ions (no ICFs are needed) from the intermediate apertures.

5.7. Comparison of O, N, and S abundances derived in M 43 and the Orion nebula

Given that the ionized gas from both nebulae, M 42 and M 43, comes from the same molecular cloud, it is reasonable to assume that both have very similar total abundances. In this section, we compare the O, S, and N abundances obtained from our study of M 43 with those determined by García-Rojas & Esteban (2007, GRE07; an update of Esteban et al. 2004) and Simón-Díaz & Stasińska (2011, SDS11) from the analysis of the nebular spectrum of the Orion nebula. Table 4 summarizes the final abundances indicated in the three studies. We note that GRE07, and SDS11 present analyses of the same spectroscopic dataset, but use different atomic data and ICFs. Our study of M 43 uses the same atomic data as that considered by GRE07.

5.7.1. Oxygen

For this element, where the total abundance is obtained directly from observable ions (O ⁺ and O ⁺+O ²⁺ in the case of M 43 and M 42, respectively), good agreement is found within the errors.

5.7.2. Nitrogen

For this element, GRE07 and SDS11 evaluate different values. This discrepancy is caused by the assumed ICF(N⁺). While GRE07 obtained this ICF from the widely used empirical relation N⁺/O⁺ = N/O, SDS11 determined the ICF from a model fitted to the data of M 42 (including a detailed description of the ionizing SED of Θ¹ Ori C predicted by the stellar atmosphere code WM-basic, Pauldrach et al. 2001). The GRE07 and SDS11 abundances differ by 0.17 dex. Interestingly, our determination for M 43, in which no ICF is needed produces an intermediate value, and is in perfect agreement with the B-type stars abundance derived by Nieva & Simón-Díaz (2011). This introduces a new important constraint on the nebular CEL/RL abundance conundrum (Simón-Díaz & Stasińska 2011).

5.7.3. Sulphur

GRE07 and SDS11 derive different values (0.17 dex difference) of the total S abundance in M 42. In this case, the cause of the difference is not, however, the assumed ICF, but the use of different sets of atomic data. Both computations considered the same data for the collision strength (Keenan et al. 1996; Tayal & Gupta 1999), but different transition probabilities: while GRE07 based their calculation in data from Mendoza & Zeippen (1982a,b) for [S ii] and [S iii], respectively, SDS11 used Froese Fischer (2004) and Froese Fischer & Tachiev (2004) values. For M 43, we obtained 12 + log(S/H) = 6.97  ±  0.04 using the same set of atomic data as GRE07. This value is slightly lower, but in agreement within the errors, than the determination by GRE07, and somewhat higher than the value provided by SDS11. The influence of the atomic data considered for this element is hence relevant.

5.8. Comparison with previous studies of M 43

The only nebular abundance studies found in the literature for M 43 are those of Rodríguez (1999, 2002), who analyzed long-slit optical spectroscopic data for five different slit positions along the nebula, deriving the physical conditions (T_e and n_e), and O, S, Cl, N, Ar, He, C, and Fe abundances of the nebular gas. O’Dell & Harris (2010) analyzed a set of optical spectrophotometric observations of M 43 as part of a more general study of the EON, and derived the physical conditions of the nebula. We note that none of these observations were corrected for the diffuse emission component as in our study.

For consistency, we reanalyzed the data of Rodríguez (1999) using the same atomic data as in our work. We obtained T_e([N ii]) between 7800 K and 8000 K for slit positions similar to ours. O’Dell & Harris (2010) obtained very similar results (T_e([N ii]) between 7780 K and 7940 K) from several aperture extractions of a slit that crosses the nebula in the east-west direction. Both authors derived n_e ([S ii])  ~  500–600 cm^-3, which is very similar to the value derived in this work.

We compared our abundances with those resulting from the reanalysis¹⁶ of the observational dataset by Rodríguez (1999). The derived O abundances are 0.14–0.23 dex lower than those resulting from the analysis of our spectra. As we indicate in Sect. 5.9, this effect is produced by the contamination of the M 43 emission by the diffuse component.

Finally, we compared our estimated T_e([O ii]) with previous determinations from radio continuum observations. Subrahmanyan (1992) derived T_e = 9000  ±  1700 K by combining 330 MHz and 10.7 GHz continuum observations and a model of an isothermal H ii region; Mills & Shaver (1968) estimated a value of T_e  ~  8000 K from 408 MHz continuum observations; finally, Thum et al. (1978) derived a much lower T_e  ~  6700 K from the HI 91α radio recombination line to continuum ratio. In general, the different estimates, except for Thum et al., agree within the uncertainties.

5.9. Scattered nebular light from the EON affecting the spectrum of M 43?

O’Dell & Goss (2009) and O’Dell & Harris (2010) demonstrated that scattered light from the bright Huygens Region of M 42 can affect the physical conditions derived from emission lines in the EON. O’Dell & Harris (2010) performed a detailed spectroscopic study of several zones of the EON, including the M 43 region. These authors could not find conclusive evidence of signatures of scattered light from the Huygens region affecting the nebular emission from M 43 (except for the case of an aperture located in the M 43 dark lane). Our study (based on more detailed observational dataset) allowed us to observationally prove that there is an extended diffuse emission not produced by ionization of the surrounding material by NU Ori affecting the whole nebula.

To investigate whether this extended emission relates to scattered light from the Huygens region, we compare in Fig. 9 (upper panel) the non-extinction-corrected Hγ/Hβ and Hδ/Hβ line ratios with (C) and without (NC) correction for diffuse emission for apertures from A1 to A9. The Hδ/Hβ line ratio is more affected than Hγ/Hβ. This is consistent with the diffuse extended emission being produced by dust scattered light, a well-know property of this type of emission is that blue lines are more affected than red lines. An additional argument supporting this hypothesis is presented in the bottom panel of Fig. 9. The value of the He i 7065/6678 ratio, obtained from the NC spectra of M 43 outside the He ⁺ Strömgen sphere associated with NU Ori agrees with the associated average value in the Huygens region taken from the data of O’Dell & Harris (2010)¹⁷.

We also investigated the effect that scattered light contaminating the spectrum of M 43 has on line ratios used to determine the physical conditions and chemical abundances (as well as constraints of the photoionization models presented in Paper II). Figure 10 shows several line ratios obtained from the C and NC spectra. Although the effect is negligible for the density constraints and the S ⁺ and O ⁺ abundance indicators, the influence of scattered light on the ionization structure constraints and the He ⁺, N ⁺, Ar ² + , S ² + , and O ²⁺ abundance indicators is not negligible (especially in the outer parts of the nebula). In addition, as already pointed out by O’Dell & Harris (2010), our comparison identifies an important effect on T_e([O ii]). If the scattered light contribution is not removed, the spectroscopic analysis overestimates the T_e and hence underestimates of the abundances.

To quantitatively illustrate this effect on the computed physical conditions and abundances, we present in Table 5 the results of two empirical analyses of the C and NC spectra from apertures 4 and 5. We also compare the result of the analysis of the NC integrated spectrum (extracting the global long slit spectra). As pointed out before, n_e is basically similar but the NC spectra result in a higher T_e and, consequently lower abundances. In particular, ionic abundances can be affected by  ~0.05–0.15 dex.

An adequate correction of the scattered light contribution in the spectra of M 43 is hence important for a reliable empirical analysis of the nebular spectra of this H ii region.

6. Summary and conclusions

M 43 is a close-by, bright H ii region of simple geometry, ionized by a single star. These characteristics makes M 43 an ideal object for investigating several topics of interest in the field of H ii regions and massive stars.

In a series of two papers, we present a combined, comprehensive study of the nebula and its ionizing star using as many observational constraints as possible. In this first part of the study, we have introduced the observational dataset, obtained the stellar parameters of HD 37061, obtained useful information from the nebular images, and analyzed the nebular spectra extracted from apertures located at various distances from the central star. All this information is then used in Paper II to construct a customized photoionization model of the nebula.

In this paper, we have found observational evidence of a diffuse and extended emission in the region where M 43 is located that is not associated to material ionized by HD 37061. Our nebular observations have allowed us to ascertain the most likely origin of this emission (e.g. O’Dell & Goss 2009; O’Dell & Harris 2010, and references therein), namely that light emitted in the Huygens region (the central, brightest part of the Orion nebula) is scattered by dust. We have also shown the importance of an adequate correction of this scattered light from the imagery and spectroscopic observations of M 43 to a proper determination of the total nebular Hα luminosity, the nebular physical conditions, and chemical abundances. In particular, we illustrated that an overestimate of T_e by  ~400–500 K, hence an underestimate of abundances by 0.05–0.10 dex, result from the empirical analysis of our spectroscopic dataset when the spectra are not corrected for the scattered light contribution. In addition, the derived total nebular Hα luminosity may be overestimated by a factor  ~1.5.

The quantitative analysis of the optical spectrum of HD 37061 with the stellar atmosphere code FASTWIND lead to T_eff = 31 000  ±  500 K, log g = 4.2  ±  0.1, and log Q(H⁰)=  47.2  ±  0.2 (assuming a distance of 400 pc).

The analysis of the Hα and Hβ images indicate a non-constant extinction distribution within the nebula that is well correlated with the dust features indicated by Smith et al. (1987). Once the Hα image is corrected for extinction and diffuse emission, a total nebular Hα luminosity of (3.0 ± 1.1) × 10³⁵ erg s^-1 is obtained. This value is compatible with the ionizing flux of the star, implying that the nebula is (mostly) ionization bounded.

We extracted nine apertures from a long-slit located to the west of HD 37061 (east-west direction) to obtain the spatial distribution of the nebular physical conditions (temperature and density) and ionic abundances (He ⁺, O ⁺, O ²⁺, N ⁺, S ⁺, S ²⁺, Ar ²⁺). Since it is important to correct these spectra for the contribution of scattered light from the Huygens region, we also consider two apertures outside the nebula that are used for this aim.

For sulfur, oxygen, and nitrogen, we have been able to determied total abundances directly from observable ions (no ICFs are needed). The derived abundances, 8.57  ±  0.05, 6.97  ±  0.03, and 7.80  ±  0.04, respectively, are compared with previous determinations in the Orion nebula. Although an overall agreement is found, our study illustrates the importance of the atomic data and, specially in the case of M 42, the considered ICFs.

---

Acknowledgments

S.S.D. acknowledges the funds by the Spanish Ministerio de Educación y Ciencia under the MEC/Fullbright postdoctoral fellowship program. S.S.D., J.G.R., C.E., and A.R.L.S. acknowledge finantial support by the Spanish MICINN under projects AYA2008-06166-C03-01 and AYA2007-63030. S.S.D. also acknowledges financial support from UNAM/DGAPA PAPIIT IN 112708 (IP: M. Peña) and CONACyT J49737 (IP: C. Morisset) and the members of the Instituto de Astronomía, UNAM for their warm hospitality. J.G.R. acknowledges the support of an UNAM postdoctoral grant. Part of this work was done while CM in sabbatical at IAC founded by CONACyT and PASPA-UNAM grants. This work has also been partially funded by the Spanish MICINN under the Consolider-Ingenio 2010 Program grant CSD2006-00070: First Science with the GTC.

References

All Tables

All Figures