Sergio Pascual And Bernabe Cedres

8.1. Introduction

The imaging method which uses narrow-band filters is the first and most direct way to obtain information about the emission lines. In contrast to broad-band photometry, in which several emission lines are integrated within one filter, the narrow-band filters are designed to transmit only one emission line (in ideal cases).

This method has considerable advantages over spectroscopic methods: we are able to obtain spatial resolution, i.e. several objects fit in only one exposure frame; the required time for an observation run is shortened; and lastly, the observation procedures and reduction are less complicated than those of a spectroscopic survey. Usually during a run information for only a few lines (three or four at best) can be obtained, in contrast to the full spectrum obtained in spectroscopy: see, for example, the works of Belley & Roy (1992), where hundreds of HII regions are presented with fluxes in three emission lines using direct imaging, and van Zee et al. (1998), with only about a dozen HII regions in ten different emission lines employing optical spectroscopy.

Narrow-band imaging is a powerful tool for characterizing the physical properties of the star-forming regions in nearby galaxies. With a few lines, we are able to determine, for example, the abundance, temperature, or initial mass function of the ionizing stars, thus obtaining clues to the physical processes that occur in the core of the H II regions; see, for example, Cedres & Cepa (2002). This can be seen in Figure 8.1, where the composite broad-band image (with the B, R, and K bands) and an Ha image of the galaxy UCM 2325+2318 can be compared. The star-forming regions are clearly defined in the narrow-band image.

Narrow-band imaging has successfully been used to build samples of diverse types of emission-line objects: see for example Rauch (1999) for galactic planetary nebulae, and Okamura et al. (2002), Castro-Rodriguez et al. (2003), and Arnaboldi et al. (2003) for extragalactic PNe.

Emission-line galaxies (hereafter ELGs) are an invaluable resource for our understanding of the evolution of galaxies in the Universe. Faint galaxies can be difficult to confirm spectroscopically, whereas the ELGs are generally easy to identify. Furthermore, the emission lines are produced within regions related either to star formation or to the phenomenon of active galactic nuclei (AGNs).

Narrow-band imaging reduces significantly the contribution of the sky brightness, since it is admitted in a small range of wavelengths. The sky background, which is the most-significant limitation in the detection of objects in broad-band images, is greatly reduced. A small wavelength range also increases the contrast (Thompson et al. 1995) between the emission line and the continuum. In order to use the narrow-passband filters with maximum efficiency, regions of the night-sky spectrum with a minimum background are selected. In the optical wavelengths, the windows in the Meinel OH bands around 8200 A and 9200 A have been used to center narrow-band filters.

Narrow-band imaging produces volume-limited samples, since the narrow observed bands correspond to small windows in redshift space, with approximately constant luminosity limits. In the case of detecting lines used as star-formation tracers (such as Ha),

The Emission-Line Universe, ed. J. Cepa. Published by Cambridge University Press. © Cambridge University Press 2009.

Figure 8.2. Wavelength of the emission lines as a function of the redshift of the galaxy. The emission lines are represented by lines increasing with wavelength. Ha is shown as a solid line, [O Ill] A5007 as a dashed line, and [O Il] A3727 as a dotted line. The airglow windows are shown as vertical shaded regions.

Wavelength (A)

Figure 8.2. Wavelength of the emission lines as a function of the redshift of the galaxy. The emission lines are represented by lines increasing with wavelength. Ha is shown as a solid line, [O Ill] A5007 as a dashed line, and [O Il] A3727 as a dotted line. The airglow windows are shown as vertical shaded regions.

the sample would be directly star-formation-rate (SFR)-selected, except for the AGN contribution.

The search for emission-line galaxies using narrow-band filters is open to different lines at different redshifts. If the redshift of the source and the rest wavelength of the emission line act together to put the line inside the narrow-band filter, the galaxy will be selected as long as it is bright enough. The lines we expect to detect at low redshifts are Ha, [O iii] AA4959, 5007 and [O II] A3727 (Tresse et al. 1999; Kennicutt 1992). At higher redshifts, Lya can be detected with 8-m-class telescopes. In Figure 8.2 we can see how the redshift

Brevundimonas Diminuta
Figure 8.3. Two narrow-band objects from Pascual et al. (2007) selected on the 8200-A airglow window. On the left is shown a z = 0.24 Ha emitter, on the right a z = 0.6 [O Ill] emitter. Images are composite U, g, r, 60 arcseconds wide, with north on top and east on the left.

of the source makes the different lines enter in the airglow windows. Also, some sample objects are shown in Figure 8.3.

Narrow-band imaging produces approximately volume-limited samples, since the narrow-observed bands correspond to small windows in redshift space. The objects are selected with a well-defined limit in equivalent width, and the line flux can be transformed into luminosity with some simple assumptions. Narrow-band imaging thus provides line luminosities for a volume-limited sample of emission-line galaxies. In the case of detecting lines used as star-formation tracers, the sample would be directly SFR-selected, except for the AGN contribution.

The problem with this approach is that stars contaminating the sample or contributions from different emission lines cannot be separated with narrow-band imaging alone. Additional assumptions (on the luminosity functions of ELGs, e.g. Jones & Bland-Hawthorn (2001)) or additional data (see, e.g., the color-color diagrams of Fujita et al. (2003)) are needed to complete the scenario.

The purpose of this chapter is to provide tools for handling narrow-band data. Although the notes are IRAF^-centric, they can easily be translated to other packages. Additionally, we assume that the images have been reduced up to the step of flat-fielding.* Section 8.2 deals with calibration and sky subtraction for narrow-band imaging of an extended source. Section 8.3 focuses on the problems that appear with narrow-band imaging in the airglow windows.

8.2. Extended-source imaging

In the case of extended-source images, such as, for example, nearby galaxies, several additional corrections are needed in order to wipe out the contributions from sources t IRAF is distributed by the National Optical Astronomy Observatories, which is operated by the Association of Universities for Research in Astronomy, Inc. (AURA) under a cooperative agreement with the National Science Foundation.

* Documentation about generic image reduction and IRAF usage can be found, for example, at http://iraf.net/docs.

Figure 8.4. Optimal regions in a frame for carrying out the sky subtraction in order to avoid contamination from a galaxy (marked as rectangles).

different from the ones studied. The ADUs must also be converted into physical units. This is done through flux calibration.

8.2.1 Sky subtraction

The sky counts are not as important in narrow-band as in broad-band images. Nevertheless, for long-exposure images, such as galaxy images and weak-line images (O II A3727, [S iii] A9069, etc.), it is necessary to subtract the sky contribution. The sky correction is done by collecting statistics in boxes that are not affected by the observed object (i.e. galaxy) near the corners of the image, as shown in Figure 8.4. The value for the sky is the mean of the mean values for each of the boxes. This is then subtracted from the image.

8.2.2 Flux calibration

In order to calibrate the images, fluxes from spectrophotometric stars are needed. We will use the stars proposed by Oke & Gunn (1983) and Oke (1990). The data can be

Was this article helpful?

0 0

Post a comment