Understanding the ALFALFA Catalog

Link to Public ADS Library of ALFALFA Design, Survey, Catalog Papers

Why are the optical vs radio positions different?

When possible, optical counterparts to ALFALFA detections have been identified and published in the catalog. Due to the combination of Arecibo’s 3.5 arcminute beam size and the ALFALFA driftscan approach, ALFALFA positions are uncertain to 24 arcseconds (compared to roughly 2 arcseconds for optical positions). Thus, when an optical counterpart has been identified, using the optical position over the radio position is recommended.

Is this the radio counterpart to my optical galaxy that I’m looking for?

Great care is taken to link optical counterparts with ALFALFA detections. The assessment of identification between optical and HI sources is based on spatial proximity, redshift (if optical is available), morphology, color and evidence for optical emission lines (if known). For sources with no discernible optical counterpart and those for which such assignment is ambiguous, due to the presence of more than one equally possible optical counterpart, no optical position is listed. The latter set includes HVCs. For objects with more than one possible candidate as an optical counterpart but such that one of the candidates is significantly more preferable than the others, an optical identification is made; however, a comment on the possible ambiguity is added in the notes to the catalog.

I can’t find a radio counterpart for my source. Does that mean it’s gas-less?

If a position search of the catalog does not return an ALFALFA detection/source, this could be telling you: Your source location could be located outside of the ALFALFA footprint. See documentation on grid locations and coverage. Your source may not be detected by ALFALFA and thus ALFALFA provides an upper limit to its gas emission. The sensitivity of ALFALFA is a function of velocity width: a wide low S/N source will be harder to detect than a narrow low S/N source. See also the note below about frequencies/velocities with lower weights.

Are there frequencies/velocities with less/less reliable data?

There is a steep dropoff at each end of the spectral bandpass that is an instrumental effect associated with the edge of the spectrometer’s sensitivity. This leads to much lower weights for velocities ~< -1000 km/s and velocities ~> 17500 km/s.

Radio frequency interference from the San Juan Federal Aviation Administration radar operating at 1350 MHz leads to a significant decrease in available data at velocities of ~15,000 to 16,000 km/s. There are less significant dips in data availability/quality at the positions of modulations of the 1350 MHz radar at 8800 km/s, 3300 km/s, 2200 km/s. While the feature between 15,000 and 16,000 km/s cuts off access to a larger cosmic volume, the narrower feature at 8800 km/s ultimately affects more possible detections because it is located near the peak in the redshift distribution of the ALFALFA catalog.

See Figure 1 of Giovanelli et al. 2007 for examples of weight distributions for ALFALFA data. https://ui.adsabs.harvard.edu/abs/2007AJ....133.2569G/abstract

See Section 2.3 of Giovanelli et al. 2007 for more information on how these weights are determined from visual inspection of the data: https://ui.adsabs.harvard.edu/abs/2007AJ....133.2569G/abstract

What do the data flags/codes mean?

Code 1: Reliable, extragalactic HI source; down to a soft limit of S/N = 6.5

More detail: the signal should exhibit a good match between the two independent polarizations and a spatial extent consistent with expectations given the telescope beam characteristics. Thus, some candidate detections with S/N > 6.5 have been excluded on grounds of polarization mismatch, spectral vicinity to rfi features or peculiar spatial properties. Likewise, some features of S/N < 6.5 are included as reliable detections, due to optimal overall characteristics of the feature. The S/N threshold for acceptance of a reliable detection candidate is thus soft.

Code 2: Candidate HI source of lower S/N (~4.5 to 6.5) corroborated by the vicinity of a possible optical counterpart of the same redshift

More detail: sources of low S/N (< 6.5), which would ordinarily not be considered reliable detections by the criteria set for code 1. However, those HI candidate sources are matched with optical counterparts with known optical redshifts which match those measured in the HI line. These candidate sources, albeit “detected” by our signal finding algorithm, are accepted as likely counterparts only because of the existence of previously available, corroborating optical spectroscopy. We refer to these sources as “priors”. We include them in our catalog because they are very likely to be real.

Code 9: objects assumed to be High Velocity Clouds

More detail: no estimate of their distances is made.

Catalog Entry Descriptions

(original source: https://ui.adsabs.harvard.edu/abs/2007AJ....133.2569G/abstract)

• Col. 1: an entry number for this catalog.
• Col. 2: the source number in the Arecibo General Catalog, a private database of extragalactic objects maintained by M.P.H. and R.G. The AGC entry normally corresponds to the optical counterpart except in the cases of HI sources which cannot be associated with an optical object with any high degree of probability.
• Col. 3: center (J2000) of the HI source, after correction for systematic telescope pointing errors, which range between a few and about a dozen arcsec. The accuracy of the HI positions depends on source strength. On average, the positional accuracy is about 24 arcseconds.
• Col. 4: center (J2000) of the optical galaxy found to provide a reasonable optical counterpart. This position has been checked for each listed object and assessed using tools provided through the SDSS and Skyview websites. Quality of centroids is estimated to be 2 arcseconds or better. The assessment of identification between optical and HI sources is based on spatial proximity, redshift (if optical is available), morphology, color and evidence for optical emission lines (if known). For sources with no discernible optical counterpart and those for which such assignment is ambiguous, due to the presence of more than one equally possible optical counterpart, no optical position is listed. The latter set includes HVCs. For objects with more than one possible candidate as an optical counterpart but such that one of the candidates is significantly more preferable than the others, an optical identification is made; however, a comment on the possible ambiguity is added in the notes to this table, as alerted by an asterisk in Col. 12.
• Col. 5: heliocentric velocity of the HI source, cz, measured as the midpoint between the channels at which the flux density drops to 50% of each of the two peaks (or of one, if only one is present) at each side of the spectral feature. Units are km s −1 . The error on cz to be adopted is half the error on the width, tabulated in Col. 6.
• Col. 6: velocity width of the source line profile, W50, measured at the 50% level of each of the two peaks, as described for Col. 5. This value is corrected for instrumental broadening. No corrections due to turbulent motions, disk inclination or cosmological effects are applied. Between brackets we show the estimated error on the velocity width, w, in km s −1 . This error is the sum in quadrature of two components: the first is a statistical error, principally dependent on the S/N ratio of the feature measured; the second is a systematic error associated with the subjective guess with which the observer estimates the spectral boundaries of the feature: maximum and minimum guesses of the spectral extent of the feature are flagged and the ratio of those values is used to estimate systematic errors on the width, the velocity and the flux integral. In the majority of cases, the systematic error is significantly smaller than the statistical error; thus the former is ignored.
• Col. 7: integrated flux density of the source, Fc, in Jy km s −1 . This is measured on the integrated spectrum, obtained by spatially integrating the source image over a solid angle of at least 7 0×7 0 and dividing by the sum of the survey beam values over the same set of image pixels (see Shostak & Allen 1980). Estimates of integrated fluxes for very extended sources with significant angular asymmetries can be misestimated by our algorithm, which is optimized for measuring sources comparable with or smaller than the survey beam. A special catalog with parameters of extended sources will be produced after completion of the survey. The estimated uncertainty of the integrated flux density, in Jy km s −1 is given between brackets. Uncertainties associated with the quality of the baseline fitting are not included; an analysis of that contribution to the error will be presented elsewhere for the full survey. See description of Col. 6 for the contribution of a possible systematic measurement error.
• Col. 8: signal–to–noise ratio S/N of the detection, estimated as S/N = 1000Fc W50 w 1/2 smo σrms (1) where Fc is the integrated flux density in Jy km s −1 , as listed in col. 7, the ratio 1000Fc/W50 is the mean flux across the feature in mJy, wsmo is either W50/(2 × 10) for W50 < 400 km s −1 or 400/(2 × 10) = 20 for W50 ≥ 400 km s −1 [wsmo is a smoothing width expressed as the number of spectral resolution bins of 10 km s −1 bridging half of the signal width], and σrms is the r.m.s noise figure across the spectrum measured in mJy at 10 km s −1 resolution, as tabulated in Col. 9. In a similar analysis, in Giovanelli et al. 2005b (hereafter Paper II) we adopted a maximum smoothing width W50/20 = 10. See Figure 5 and related text below for details. The value of the smoothing width could be interpreted as an indication of the degree to which spectral smoothing aids in the visual detection of broad signals, against broad–band spectral instabilities. The ALFALFA data quality appears to warrant a more optimistic adoption of the smoothing width than previously anticipated.
• Col. 9: noise figure of the spatially integrated spectral profile, σrms, in mJy. The noise figure as tabulated is the r.m.s. and measured over the signal– and rfi-free portions of the spectrum, after Hanning smoothing to a spectral resolution of 10 km s −1 .
• Col. 10: adopted distance in Mpc, DMpc. For objects with cz_cmb > 3000 km/s, the distance is simply czcmb/H◦; cz_cmb is the recession velocity measured in the Cosmic Microwave Background reference frame and H◦ is the Hubble constant, for which we use a value of 70 km s −1 Mpc −1 . For objects of lower cz_cmb, we use a peculiar velocity model for the local Universe, as described in Paper II. Objects which are thought to be parts of clusters or groups are assigned the cz_cmb of the cluster or group. Cluster and group membership are assigned following the method described in Springob et al. 2006. A detailed analysis of group and cluster membership of ALFALFA objects will be presented elsewhere. Note that the Virgo cluster extends over much of the solid angle surveyed. This introduces unavoidable ambiguities in the distance assignment, as the peculiar flow model only corrects for large–scale perturbations in the velocity field and is unable to deal effectively with regions in the immediate vicinity of the cluster and along a section of a cone which contains the cluster, up to cz ∼ 2500 km s −1 . The distance to the Virgo cluster was assumed to be 16.7 Mpc.
• Col. 11: logarithm in base 10 of the HI mass, in solar units. That parameter is obtained by using the expression MHI = 2.356 × 10^5 D^2 MpcFc.
• Col. 12: object code, defined as follows: Code 1 refers to sources of S/N and general qualities that make it a reliable detection. By “general qualities” we mean that, in addition to an approximate S/N threshold of 6.5, the signal should exhibit a good match between the two independent polarizations and a spatial extent consistent with expectations given the telescope beam characteristics. Thus, some candidate detections with S/N > 6.5 have been excluded on grounds of polarization mismatch, spectral vicinity to rfi features or peculiar spatial properties. Likewise, some features of S/N < 6.5 are included as reliable detections, due to optimal overall characteristics of the feature. The S/N threshold for acceptance of a reliable detection candidate is thus soft. In a preliminary fashion, we estimate that detection candidates with S/N > 6.5 in Table 1 are reliable, i.e. they will be confirmed in follow–up observations in better than 95% of cases (Saintonge 2007). Follow-up observations planned for 2007 will set this estimate on stronger statistical grounds. Code 2 refers to sources of low S/N (< 6.5), which would ordinarily not be considered reliable detections by the criteria set for code 1. However, those HI candidate sources are matched with optical counterparts with known optical redshifts which match those measured in the HI line. These candidate sources, albeit “detected” by our signal finding algorithm, are accepted as likely counterparts only because of the existence of previously available, corroborating optical spectroscopy. We refer to these sources as “priors”. We include them in our catalog because they are very likely to be real. Code 9 refers to objects assumed to be HVCs; no estimate of their distances is made. Notes flag. An asterisk in this column indicates a comment is included for this source in the text below