nueva página del texto (beta)
Inglés (pdf)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
1. INTRODUCTION
Beginning in 1878 at the Chapultepec Hill in Mexico City, the Observatorio Astronómico Nacional (hereafter OAN) has been moved to several different sites, finally to be located at the San Pedro Mártir (SPM) Sierra in the 1970s. This final location possesses ideal physical and geographical conditions that have led to its classification as one of the five best sites in the world for optical astronomical observations. The mountains where the observatory is located have an approximate altitude of 2850 m and are surrounded by deserts, providing a dry atmosphere. The pine forest in the mountain gives a natural protection against wind erosion and dust from the deserts leaving a clear sky with low atmospheric extinction (Schuster & Parrao 2001; Parrao & Schuster 2003; Schuster, Parrao, & Guichard 2002), excellent seeing (Michel et al. 2003a; Tapia et al. 2007; Sanchez et al. 2012), and a small temperature variation throughout the nights (Tapia 2003; Michel et al. 2003b). These atmospheric conditions allow astronomers to obtain very high-quality photometric instrumental observations. Also, given its position on top of a ridge, it is possible to observe in any direction without obstruction and to follow objects in the sky for eight or more continuous hours (Schuster, Parrao, & Guichard 2002) with minimal interference from the surrounding pine trees. All these conditions make the OAN-SPM an excellent site for optical astronomical observations. Many characteristics of this site are discussed in Volumes 19 and 31 (October 2003 and October 2007) issues of the Revista Mexicana de Astronomía y Astrofísica (Conference Series).
In particular, the OAN-SPM has become an important site for photometric studies. The broadband UBVRI (Iriarte et al. 1965) and narrow-intermediate-band 13-color (Johnson & Mitchell 1975) systems were the first to be used at the observatory. Since 1983, with the arrival of the Danish spectrophotometer, astronomers have been able to carry out alternative intermediate-band photometric observations. The 6-channel Danish photometer possesses six non-refrigerated photomultipliers with four continuum filters (u,v,b, and y) and two line filters (Hβ, narrow (N) and wide (W)) in the Strömgren-Crawford system. This system was defined by Strömgren and Crawford in a series of papers (Strömgren 1951, 1954, 1955, 1958, 1963, 1966; Crawford & Mander 1966; Crawford et al. 1966; Crawford & Barnes 1970) that focused mainly on the study of spectral classification for F-type stars but were also applicable to B, A, G, and K stars. Strömgren's analysis consisted of the classification of his sample of 110 stars by temperature, luminosity class, metallicity, and stellar population using spectrophotometric features like the Balmer discontinuity and line blanketing. In this way, Stroomgren developed a quantitative photometric system for spectral classification closely related to the spectral features of stars and the filter band-widths. With these filters, Strömgren defined a set of three color indices: (b - y), measuring the continuum slope and therefore the stellar temperature; c1 which measures the Balmer discontinuity, giving information concerning the stellar surface gravities and absolute magnitudes, defined as
and m! related to the stellar metallicity via the ef fects of line blanketing, and defined as
Reproductions of the original photometers or spectrophotometers with very similar filters and de tectors were installed at various observatories, such as the La Silla Observatory in Chile, the Kitt Peak National Observatory in the USA, and the OAN-SPM in Mexico, among others. Since it is very hard to construct filters having exactly the same widths and central wavelengths, or photomultipliers (PMTs) with exactly the same photocathode response, it became very important to have an extensive list of standard stars observed by Stroomgren and coworkers. They compiled a set of ≈ 1200 stars homogeneously observed (Strömgren & Perry 1965) from which they selected a subset of 122 stars used as standard stars for the uvby and Hβ systems. The first set of standard stars for the uvby filters was published by Crawford & Barnes (1970). This list was extended by (Crawford et al. 1971a,b) to cover stars with V < 6.m5 leaving a set of 315 standard stars. Both sets of stars were joined to the data of 80 standard stars for the Hβ filters presented by Crawford & Mander (1966), becoming the original list of primary standard stars for the uvby -Hβ system. Later, this list was extended by Grønbech et al. (1976) to F-type stars, with a set of 27 late-F and early-G, and by Perry et al. (1987) with 366 bright stars. All these extensions conform a set of secondary standard stars. In this way, with a large set of primary and secondary standard stars, the Stroomgren system has become the very homogeneous, consistent, and easy to reproduce photometric system that is widely used today.
In 1973, Lindemann & Hauck (1973) collected the measurements for various photometric systems existing at the time, in particular, those made on the uvby-Hβ system were included. They compiled a set of 7603 stars at the Centre de Donnees Stellaires de Strasbourg (CDS). All data recorded by Lindemann & Hauck were homogenized using the lists of Crawford & Mander (1966); Crawford & Barnes (1970), and (Crawford et al. 1971a,b) as reference.
Unfortunately, not all standard stars are constant over time, introducing a source of error in the reproduction of the original system, and here lies the importance of compiling standard-star data for as long a time interval as possible. In 1983 and 1993, (Olsen 1983, 1993) published all of his standard star observations made for establishing his photometric catalogues, according to the suggestion of Lindemann & Hauck (1973). The documentation of the standard stars used for various projects and catalogues is necessary for the inclusion of any photometric data into the homogenized data sets, such as that of the CDS.
Since the 1980's uvby-Hβ observations have been carried out at the SPM observatory with the 1.5m telescope and the Danish six-channel spectrophotometer. These observations have provided the fundamental data sets for various astronomical programs like the study of high-velocity and metal-poor stars (hereafter HVMPS) (catalogues: Schuster & Nissen 1988; Schuster, Parrao, & Contreras 1993; Schuster et al. 2006), very-metal poor stars (VMPS) (catalogues: Schuster et al. 1996, 2004), stars at the North Galactic Pole (NGP) (Croswell et al. 1991), a Seyfert galaxy (Dultzin-Hacyan, D. et al. 1992), and extinction observations to determine an average atmospheric extinction curve for the OAN-SPM (Schuster & Parrao 2001; Schuster, Parrao, & Guichard 2002; Parrao & Schuster 2003). In all these programs, standard stars have been observed and used to obtain transformation coefficients from the instrumental system to the standard Strömgren system. In the case of the OAN-SPM, the Strömgren system is defined by the slots and filters described in Nissen (1984) and Gutiérrez et al. (2004). The original photomultipliers of this Danish photometer have been replaced occasionally, but all the corresponding photocathodes have been bialkaline with very similar response curves.
In this work, documentation of uvby-Hβ observations carried out over a time interval of approx imately 20 years, from September 1987 to November 2007, is presented. These data might be in cluded in the homogenized data bases compiled at the CDS, and information useful for this purpose is provided. The main documentation of the primary and secondary standard stars used in the HVMPS and VMPS projects is presented, such as their standard photometric values, taken from the literature at the beginning of these projects, as well as their transformed standard values, and in addition the errors in the instrumental system, the transformation coefficients obtained for the standard system, the transformation errors, and the methods used to obtain such observations and transformations. Further details concerning the methodologies have been given in the works of Schuster & Nissen (1988); Schuster, Parrao, & Contreras (1993); Schuster & Parrao (2001); Schuster, Parrao, & Guichard (2002); Parrao & Schuster (2003), and Schuster et al. (2006).
2. TECHNIQUES
Observations of HVMPS in the Strömgren-Crawford photometric system have been carried out at the OAN-SPM for more than 20 years, since March of 1984. In particular, photometric data ob servations of 33 observing runs from September 1987 to November 2007 are documented here. In the following sections the observing and reduction techniques used during all of these runs are described. It is worth emphasizing the consistency of the observing and reduction methods followed during all these runs, making this data set a very homogeneous set of photometric observations with reliable transformations to the original uvby-Hβ system. In general, the structuring of the observations and data reductions followed the precepts of Grønbech et al. (1976).
2.1. Observing techniques
In all 33 observing runs, the observing and reduction techniques as described in detail in a series of papers related to the study of HVMPS (Schuster & Nissen 1988; Schuster, Parrao, & Contreras 1993; Schuster et al. 1996, 2004, 2006), were followed. A brief description of such methods is presented here. Nearly all of the observations were obtained by one of us (W.J.S.) using the 1.5m telescope at the OAN-SPM. Whenever possible, identification charts were used at SPM due to the low pointing accuracy of the telescope. Photometric standard stars were carefully selected (see the following section for details). A six-channel "Danish" photometer was used. The four-channel uvby section is really a spectrograph-photometer which employs exit slots and optical interference filters to define the band-passes. At the beginning of each observing run the grating angle of the spectrograph was calibrated using a cadmium lamp, or using a supergiant star (only once), to po sition the spectra on the exit slots to within about ±1Ǻ (Nissen 1984; Gutiérrez et al. 2004; Schuster, Parrao, & Guichard 2002).
The centering of stars was done carefully and each observation consisted of at least three integra tions of the star and at least one sky integration. Integration times ranged from 3 x 10 seconds for the brighter standard stars to 2N x 20 seconds for the fainter program stars, with "N" being an integer. Sky measures were usually made 1.0 to 1.5 arcmin north or south of the stellar position. For the "extinction" and "drift" stars (see below for their definition) a full observation usually consisted of three "star" integrations (both uvby and Hβ), "sky" integrations, then again three "star" integrations. At the largest air masses, for atmospheric extinction determinations (for uvby only), 9-12 "star" integrations, two "sky" integrations, then again 9-12 "star" integrations were performed. For the other standard stars a full observation was usually three "star" integrations (both uvby and Hβ), and then one "sky" integration for each section.
In general, some standard stars ("drift" stars) with declinations similar to the observatory's latitude (20°
Approximately 15-25 standard stars (including the "extinction" and "drift" stars) were observed on all nights of an observing run to provide a consistent, connected instrumental photometric system for the data reductions of that run, while other standard stars were added progressively and observed on at least three nights each, to improve the photometric transformations; usually 35-50 standard stars were managed for each full observing run (see Tables 2, 3, and 4).
2.2. Reduction techniques
Reduction techniques for all these observing runs followed the precepts of Grønbech et al. (1976) using a computer program kindly loaned to us by T. Andersen. The reduction package contains two main parts: one which creates an instrumental photometric system using all nights of an observing run, and one which transforms this instrumental system to the standard system. Full description of the package and auxiliary codes as used in Mexico are described in Parrao, Schuster, & Arellano Ferro (1988). In every observing run we have assumed a linear time dependence of the night corrections and a constant atmospheric extinction throughout a night. Based on all our data reductions, these assumptions proved to be entirely adequate.
The equations for the transformation to the standard uvby-Hβ system are the linear ones of Crawford & Barnes (1970) and Crawford & Mander (1966) (Grønbech et al. 1976). In most of our observing runs (for the HVMPS and VMPS), the redder sub- giants and giants, (b-y) ≳ 0.m50 and c 1 ≳ 0.m35, had to be removed from the standard list to maintain good results with the linear transformations. All of the dwarf and subdwarf stars and the early-type giants and subgiants transformed well together. For wider ranging projects (such as the NGP and a few observations of VMPS), transformation equations were separated with different criteria: (b-y) ≲ 0.m40 and (b-y) ≳ 0.m40 to include bluer and redder stars, such as these red subgiants and giant stars, similar to Grønbech et al. (1976). Such a separation of the transformation equations was not needed for the HVMPS and (most of the) VMPS.
3. STANDARD STARS
Primary and secondary uvby standard stars were selected from the lists of (Olsen 1983, 1984, 1993) which include the Crawford & Barnes (1970) list; secondary standards had to be used due to the brightness limits of this photometer (see below). We have checked that those objects reported as probable variable stars were not included. Mostly our standard stars for the different observing runs were selected according to the distributions in their colors and indices, (b-y), m 1 , c 1 , and Hβ , as compared to the expected values for our F-, G-, and K-type high-velocity and metal-poor stars (see below). During the duration of this project four of the originally selected standard stars have appeared as definite variable stars in the SIMBAD data base, BS812 (δ Sct type variable star), HD80715 (BY Dra type), HD156026 (RS CVn type), and BS8799 (γ Dor type); these four have been removed from our photometric observations and data reductions. For the Hβ filters, standard stars were also taken from the list of Crawford & Mander (1966). The brightness limit for the uvby photometry was V ≳ 6.0, and V ≳ 4.5 for Hβ due to the differing sensitivities of these two sections in the 6-channel photometer. These limits also kept our count rates to less than about 300,000 c/s, and the dead-time corrections small, ≲ 0.m025. A few brighter stars with good Hβ photometry were selected from the above sources to extend the Hβ transformation to higher values, but finally those few with Hβ ≳ 2.82 were excluded from the observations and data reductions, since they introduced small non-linearities in the Hβ transformations. So, the Hβ standards: BS63, BS2857, BS4515, and BS8060 were not used in the final transformations to the standard Hβ system.
The list of 111 standard stars used in the HVMPS project is shown in Table 1. This table also contains the following two extensions: 16 stars used in the study of VMPS from the Schuster & Nissen (1988) catalogue, as well as 11 additional standard stars used in the NGP project (Croswell et al. 1991). Table 1 presents 2000.0 positions and the standard photometric values employed during our data reductions for all these stars, as well as comments concerning their use, or non-use, during these three projects. The variability types are given for the four stars, mentioned above, listed as definite variable stars in SIMBAD, plus comments for three additional variable stars identified more recently (see below).
The HVMPS standard stars (first section of Table 1) were selected from the (Olsen 1983, 1984, 1993) catalogues, where the precision of the data is reported. In the case of the Olsen (1983) cat alogue, the rms internal errors of one observation are ±0.0052, 0.0040, 0.0061, 0.0065 and 0.0065 in V, (b-y), m 1, c1 and Hβ, respectively. For the Olsen (1984) catalogue, weighted averages of the internal rms errors of one observation were ≤ ±0.005 in V, ≤ ±0.004 in (b-y), ≤ ±0.006 in m 1, and ≤ ±0.0065 in c1 . For those stars in the Olsen (1993) catalogue, internal rms errors of one observation were ±0.0047, 0.0029, 0.0041 and 0.0058, respectively, in V, (b-y), m 1 , and c1 . The standard Hβ values from Crawford & Mander (1966) have an rms deviation of ±0.0060.
The VMPS stars (first extension of Table 1) possess rms errors of ±0.006, 0.003, 0.005, 0.006, and 0.007 in V, (b-y), m 1, c 1, and Hβ, respectively. From these values the great precision and stability provided by this photometric system can be appreciated.
In order to show the spectral coverage of this set of standard stars, the V vs (b - y), m 1 vs (b - y), and c 1 vs (b - y) plots are given in Figure 1. The red points show the red subgiants which have been excluded from the HVMPS project, mostly excluded from the VMPS project, but used for the NGP project due to the greater range in colors and evolutionary states expected for these stars. The Crawford & Barnes (1970) stars are shown as black triangles, and blue squares represent the (Olsen 1983, 1984, 1993) stars. In general we can see that this set of primary and secondary standard stars covers well the ranges expected for the HVMPS and VMPS.
Fig. 1 Strömgren uvby standard values reported by Crawford & Barnes (1970) and (Olsen 1983, 1984, 1993) for those primary and secondary standard stars used in the HVMPS and VMPS projects, as given in Table 1. The red points show those red subgiants used mainly in the NGP project, those with c1 ≳ 0.35 and (b-y) ≳ 0.50. The Crawford & Barnes (1970) stars are shown as black triangles, while blue squares represent the (Olsen 1983, 1984, 1993) stars. The color figure can be viewed online.
For each observing run, histograms for (b-y), m 1, c 1, and Hβ were plotted for the standard stars to be used to check for adequate distributions as expected for F-, G-, and K-type HVMPS and VMPS, as shown in Figure 2 for the entire sample of Table 1. In this way we have covered adequately the following ranges: 5.8 ≲ V ≲ 9.0, 0.20 ≲ (b-y) ≲ 0.80, 0.00 ≲ m 1 ≲ 0.65, 0.10 ≲ c 1 ≲ 0.80, and 2.53 ≲ Hβ ≲ 2.82 during each observing run.
Fig. 2 Histograms showing the distribution of these standard stars for the different indices and colors: Hβ, c 1, m 1, and (b-y).
Figure 3 shows the dependence of the Strömgren indices, (b-y), m 1, and c 1, for our set of standard stars in Table 1,. as a. function of Hβ. The symbols are the same as in Figure 1. Several of the redder standard stars do not have standard Hβ values, only uvby.
4. TRANSFORMATION COEFFICIENTS AND ERRORS
4.1. Transformation coefficients
In order to achieve a good transformation between the instrumental and standard photometric systems, the standard stars must be selected with well as the transformation equations to be used, preferably linear ones when a close match of the filters and detectors to the original standard system has been achieved. In this case the optics, band-passes, filters, and detectors (photomultipliers) of the "Danish" 6-channel photometer in use on SPM have been closely matched to the original parameters (central wavelengths and band-passes) of the original Strömgren-Crawford, uvby-Hβ, photometric system. In the three main projects (HVMPS, VMPS and NGP) of the present work, the linear transformation equations of Grønbech et al. (1976), eqs. (6)-(9), have been used successfully, where the coefficients A, B, C, D, E, F, J, G, H, I, K and L are defined by the following:
where "std" refers to the standard photometric values, and "instr" to those values from the instrumental solution.
Fig. 3 Plots of these primary and secondary standard stars for c1, m1, (b-y), and V versus Hβ, with the symbols as in Figure 1. The color figure can be viewed online.
Table 2 shows the transformation coefficients for these equations obtained from different observing runs for the HVMPS. Zero point values (A, C, E, G, and K) are quite similar in all observing runs, showing the largest changes when the telescope mirrors have been cleaned. The first-order color coefficient values (D,F, and H) are approximately 1.0 indicating a good agreement between the original and present filter sets plus detectors. The L coefficient values show a slight deviation from 1.0 mainly due to a difference in the bandwidth of the narrow (N) filter of Hβ; Crawford changed this bandwidth early in his observations from ≈ 15Ǻ to ≈ 30Ǻ to avoid radial velocity effects in high-velocity stars for this Hβ index (Crawford & Mander 1966; Schmidt & Taylor 1979); our bandwidth for this narrow filter is very close to this latter value leading to values of ≈ 1.20-1.35 for L. Second-order color coefficients, B, J, and I, are small, indicating a reasonable, uncomplicated match to the standard photometric system. To first order these transformation coefficients agree well with those given by (Olsen 1983, 1993), Tables VIII and XII, and Tables 9 and 11, respectively. Exact agreement is not expected, since the filter-detector combinations are not identical and since Olsen divided his main-sequence stars into two groups, "BAF" and "GKV", while we have worked our main-sequence and subdwarf stars in a single group, early-F to late-K type stars.
A few observing runs, for the NGP project, were reduced using three different standard star group ings, one including the standard stars as "usual", as defined above, and a second and third reduction using only "blue" and "red" standard stars, respectively, to compensate for the wider range in color of the program stars found for this project, such as a "red" solution for the red subgiants and giants, and a "blue" solution for any blue stragglers or blue horizontal-branch stars in the sample. For wider ranging projects, such as the NGP, the red subgiants and the secondary standards in Extension 2 of Table 1 have been used to help obtain the "red" and "blue" transformation equations a la equations (10)-(15) of Grønbech et al. (1976). In Table 2 only these "usual" transformation equations, excluding the red subgiants and giants as well as those secondary standards in Extension 2 of Table 1, are documented.
Since there are no standard stars with V ≳ 8.9 in the first section of Table 1, the standard V magnitudes have been extrapolated for the fainter HVMPS, VMPS, and NGP stars in our catalogues. Figure 2 of Schuster et al. (1996) and Figure 1 of Schuster et al. (2004) show that our fainter V magnitudes (≳ 12.4 and ≳ 11.0, respectively) are quite linear and accurate, when compared to external sources.
4.2. Errors
The main error source for the final transformed uvby-Hβ photometry for the program stars comes from the color transformations (Manfroid & Sterken 1987, 1992). As can be seen in Table 3, the instrumental errors of our photometric observations (including the "extinction" and "drift" solutions) are quite small. To minimize the color transformation errors, an attempt was always made to observe all standard stars on at least three independent nights with approximately 15-25 standard stars observed on all nights, as mentioned in § 2.1.
Statistical standard errors of the transformation to the standard system for different observing runs are shown in Table 4. For most of the observing runs, standard deviations for the different color indices are of the order of a few thousandths of a magnitude, indicating very good transformations to the standard system. These errors are similar in size to the ones reported in Schuster & Nissen (1988) and Schuster et al. (2006).
5. DISCUSSION AND ANALYSIS
5.1. Instrumental system stability
One of the mainstays for supporting the instrumental stability of the six-channel "Danish" uvby-Hβ photometer at the OAN-SPM has been the calibration of the grating angle. This has been carried out frequently, usually with a cadmium lamp, three or four times a year, always before each one of our observing runs. In this way the position of the stellar spectrum on the mechanical exit slots has been maintained to within ±1Ǻ, and so to a high degree the instrumental photometric system has remained stable over many years. The filters might age, but the first-order band-passes have been maintained to high accuracy.
Another source of possible instability for the instrumental photometry derives from the phototube replacements. The original "Danish" photometer installed at the 1.5 m telescope at the OAN-SPM. possessed uncooled EMI 9789QA phototubes with bialkali cathodes. The uvby phototube set worked efficiently for about fifteen years, and then was replaced in October 1999 with uncooled Electron Tubes 9893Q/350A photomultipliers also with bialkali cathodes. (This replacement was required by the misuse of the photometer's neutral filters by some observers and the resulting deterioration of some of the photocathodes.) Since both sets of phototubes have nearly identical bialkali photocathodes, these tube replacements should theoretically not affect instrumental and standard observations. In practice, the only obvious observed change has been the dead-time corrections for the observing runs carried out with these new tubes in the uvby section. The Hβ section still retains the uncooled EMI 9789QA phototubes, as of November 2015.
5.2. Standard system stability
The high stability observed in the standard system is shown by the transformation coefficients D, F, H, and L which are very similar along the different observing runs, with average values of 0.9803 ± 0.0132, 1.0682 ± 0.0398, 1.0285 ± 0.0233, and 1.2851 ± 0.0682, respectively, as seen in Table 2, even though some variation in these coefficients is expected with changes in the mean nightly temperature affecting the transmission curves of the interference filters, with the aging of the filters, and with slightly differing standard-star selections from observing run to observing run. The same can be said for the coefficients B, J, and I: +0.0281 ± 0.0171, +0.0575±0.0467, and +0.1335±0.0419, respectively. As can be seen, the D, F, and H values are very close to 1.0 during the 20 years of observations with only a few isolated cases in which the values are somewhat different, and the values of B, J, and H are all small, agreeing with their second-order importance. The effect of the phototube replacement is reflected mainly in a change of the zero points of the transformation coefficients A, C, E, and G. Observational errors in the standard-star observations (transformation errors) have also remained on the order of a few thousandths, with the exception of a few isolated cases; see Table 4.
5.3. Transformed standard values for the uvby -H β standard stars
Table 5 shows the transformed standard values for the uvby-Hβ standard stars observed during various observing runs from 1987 through 2007 for the HVMPS and VMPS projects. As suggested by Lindemann & Hauck (1973), and as provided by (Olsen 1983, 1984, 1993) for his various uvby-Hβ catalogues, such values provide a means for homogenizing our uvby-Hβ data onto the CDS data bases. In this table the transformed standard values for each standard star during each observing run have been averaged (for the number of observing runs given in Column 15), and the total number of observations in V, in (b - y), m 1, c 1, and in Hβ are given in Columns 12, 13, and 14, respectively.
The average of the standard deviations given in Table 5 are ±0.m004, ±0.m003, ±0.m005, ±0.m005, and ±0.m004 for V, (b-y), m 1, c 1, and Hβ, respectively These small values show the good general constancy of the photometric standard stars used with the six-channel, uvby-Hβ photometer of the OAN-SPM over a period of about 20 years.
Such data also provide a means of identifying additional candidate variable stars within our standard-star set. Five of these standard stars (HD77354, HD126531, HD127029, HD130353, and HD137778B) have shown possible significant variations in the V magnitude, ∆V ≥ 0.m04, the dif ference between the maximum and minimum values throughout our data set, but only the first of these five is clearly confirmed as a variable-star candidate from the given standard deviations in Table 5; the other four may have been observed only very occasionally during nights of less photometric quality.
Other stars in this Table 5 suggesting possible photometric variability (variations ≳ 2.5σ) are HD16031 (in c 1), BS6467 (m 1 and c 1 ), HD108189 (V and c 1 ), HD108678 (c 1), HD125607 (c 1), HD154363A (c 1), HD161303 (m 1 and Hβ ), and HD198585 (Hβ). None of these stars are indicated as variable or candidate-variable stars in the SIMBAD or VSX data bases. HD2796 and HD132475 show possibly significant standard deviations in c 1 and m 1, respectively, but these two stars have fairly negative declinations (~ -17o and ~ -22o, respectively), and so are observed at larger air-masses at SPM; however, for example, the stars HD3621, HD128279, HD143131, HD144253, HD175384, HD193901, and HD196892 have similarly negative declinations (see Table 1) without showing larger standard deviations, indicating in general the high quality of the photometric sky at SPM and suggesting the possible photometric variability of HD2796 and HD132475. SIMBAD and the VSX do not indicate variability for HD2796 (and this star has been mainly used for V and Hβ values in the HVMPS project and only a few times to provide m 1 and c 1 in the VMPS and NGP), but SIMBAD and the VSX do list HD132475 as a suspected variable star with an uncertain range in V of 8.49:-8.66: (FitzGerald 1973); our data of Table 5 do not confirm any clear variability for this star in the V magnitude but do show a marginal dispersion of ±0.013 (≈ ±2.6σ) for possible variability in m 1.
In addition, the VSX and SIMBAD catalogues indicate that the following stars of Table 1 (first two sections) are suspected variable stars: HD19445, BS1430, HD32147, BS2601A, HD84937, HD94028, BS4550, HD128165, HD140283, BS5930, HD162503, HD191365, BS8086, BS8826, and BS9039. However, in Table 5 all of these stars used as standards for uvby photometry have σV ≤ ±0.006, and all stars used as Hβ standards have σ Hβ ≤ ±0.009, with dispersions and differences (as seen in Table 5 or by comparing Tables 1 and 5) significantly less than those suggested by the VSX or by SIMBAD, and thus providing little evidence for photometric variability.
However, more recent studies have shown that BS5930 and BS9039 are indeed "very-low-amplitude" (∆m ≤ 0.m02-0.m03) δ Scuti stars (Paunzen, et al. 2010; Le Contel, et al. 1974). In Table 5 all σ's for BS5930 are ≤ ±0.004; BS9039 has been little used for our observations, only for Hβ , and its σ Hβ is ±0.001. These values confirm the "very-low-amplitude" nature of these two δ Scuti stars.
In previous studies BS4550 and BS8086 have been tagged as possible flare stars (Beardsley et al. 1974; Blanco et al. 1972); but in Table 5 little evidence for flare activity can be detected: ±0.004 and ±0.003 for σ V , respectively; the largest dispersion for these two stars is σ ci = ±0.010 for BS4550. Either these stars are not in fact flare stars, or we have been lucky to always observe them outside of eruption; both have received a large number of uvby observations, especially BS4550.
HD191365 has been identified as an Algol-type eclipsing binary (Wraight et al. 2011), but again little evidence for this is seen in Table 5, with σv = ±0.004, and all dispersions less than or equal to ±0.008; it seems that all uvby-Hβ observations of HD191365 in Table 5 were taken outside of eclipse.
6. CONCLUSIONS
The main results of this paper are the following:
The secondary standards of Table 1 which have finally not been used for the observations and transformations of the HVMPS and VMPS projects are those four shown to be definite photometric variables in SIMBAD (BS812, HD80715, HD156026, and BS8799), and those which produce small non-linearities in the photometric transformations, such as the hotter Hβ standards (BS63, BS2857, BS4515, and BS8060), and the redder subgiants (b-y ≳ 0.5 and c 1 ≳ 0.35, such as HD2796, HD6268, HD17122, HD107550, BS5270, ...). These latter, non-linearity restrictions apply only to the uvby-Hβ photometer at the OAN-SPM.
The instrumental and transformation errors of Tables 3 and 4, as well as the transformation coefficients of Table 2, show that the uvby-Hβ photometry of our HVMPS and VMPS projects has remained consistent, precise, and accurate, to a high degree, over the extent of these projects, from 1987 to 2007. The transformation coefficients agree as expected with those obtained previously (Olsen 1983, 1993).
The constancy of our techniques and methods, of our photometer, and of the present standardstar list have allowed us to provide a large set of very homogeneous and consistent uvby-Hβ photometry for the HVMPS and VMPS projects at the OAN-SPM observatory. By implication, this is true for observing runs prior to 1987 and posterior to 2007, which have made use of the same grating-angle calibration, standardstar list and restrictions, and observing and reduction techniques.
The transformed standard values of Table 5 are necessary for putting our uvby-Hβ data into the CDS data base in a homogeneous way. These values have also provided us with a means to detect, or refute, variable-star candidates, especially long-period ones due to the long time extent of these projects. For example, a dozen "suspected" variable stars (from SIMBAD or the VSX) in our standard-star data base, Table 1, are not confirmed as variable stars, and three recently documented variable stars (two low-amplitude δ Scuti variables and one eclipsing binary) show very little effect upon the precision or accuracy of our photometric results.
