Loading ...
Sorry, an error occurred while loading the content.
 

GSC identifications

Expand Messages
  • geert hoogeveen
    (Prologue: This is more about the general questions that I have concerning investigation methods than about VSX.) While many technical details have already
    Message 1 of 2 , May 29 3:54 PM
      (Prologue: This is more about the general questions that I have concerning investigation methods than about VSX.)
      While many technical details have already been discussed in this discussion group, I have a more general problem that I would like to present.
       
      Basis for the VSX are the presently available databases of the GCVS/NSV, ASAS and NSVS (and some others that I do not know but that is not important for this story). Each of these databases has it's own way of naming variables and keeping records of the available data.
      When you are investigating a variable, you are always forced to look up every possible identification through various tables using coordinates, names etc. This happens because none of the robotic surveys have made an (serious) effort to relate their own observations to the variable star catalogue or standard star catalogues.
      The GCVS has now improved quite a lot on the point of coordinates and GSC-identifications, but they are not there yet, and propably quite some original finder charts have to be inspected as well.
      ASAS did not make the connection to a star catalogue, but does include GCVS identifications in their catalogue for several of their variables. But be aware that there are quite some mistakes in their id's. Always check them yourself (which can only be done when the GCVS has accurate coordinates, which then because a bit of a circular argument).
      The NSVS studies USV, Rotse1 and NSVS (Skydot) have only partly connected to variable star or star catalogues. The Skydot catalogue has GCVS identifications that are in general quite good, and also 2MASS identifications that I guess are good as well, although I haven't checked those. When you have accurate coordinates to begin with, it makes it all much easier.
      The USV and Rotse1 does not have either of those.
       
      Let me give an example to clarify this muddy way of working: The DO Her/V362 Her/ASAS 163035+2326.5 example that John gave recently. If you put it in words, the problem is like this:
      DO Her is at  163040.8 +232728 (J2000) according to the GCVS and V362 Her is at 163039+232642. In the ASAS catalogue, the variable 163035+2326.5 is identified as V362 Her. Is that correct? Put like this, it will sound very vague to most people.
      If you now use Guide8 or any other planetarium program to show the field of DO Her you will see the problem more clearly. There are 3 GSC stars in the immediate vicinity and both DO Her and V362 Her are swimming somewhere halfway between these GSC stars. It is mostly guesswork to decide which of the GSC stars is DO Her or V362 Her if you would only use Guide or coordinates.
      If I make a little table with the help of the GSC stars, it becomes even more apparent. There are 3 GSC stars in this field
       
      GSCnumber            GCVS identification            ASAS identification
      2044.0901                DO Her                                 163041+2327.5
      2044.0674                V362 Her                              163035+2326.5
      2044.0371                ------                                        V362 Her
       
      From this little list it is already very obvious that GCVS and ASAS disagree on the identification of V362 Her.
      To decide which star is which variable, we need data from real observations so here the data from NSVS and ASAS come in.
       
      For GSC 2044.0901 there are three NSVS records: 7887289/10699533/1070407, that have the following coordinates:
       
      NSVS number    coordinates
      7887289                163040.8 +232728.8
      10699533              163040.6 +232726.2
      1070407                163041.0 +232726.4
      All these three coordinates fit nicely on the GSC coordinates of GSC 2044.0901 which are 163040.9 +232728.3. The small differences are measurement errors which are propably normal (?) and for a star of this magnitude (around 11), you can safely say that this is the same star. If you are dealing with much fainter stars, I am not so sure if that still applies.
      If you take a look at the lightcurves for NSVS 7887289/10699533/1070407 you can see that it has a typical longperiodic Mira lightcurve.
      Although DO Her is not in the ASAS catalogue in their column of GCVS identifications, DO Her can be identified with ASAS 163041+2327.5. It would be even better if it could be identified with the ASAS identification chart but due to the hardware problems with ASAS that is not available at the moment.
      (Remark: the situation with the three NSVS records is not immediately clear: The average magnitude of the first one differs much from the two other ones. By inspecting the coordinates and lightcurve it becomes obvious that it is the same star in all three cases. The situation that there are multiple entries for one and the same star happens quite a lot in the NSVS).
      If we now add the data from the other variable databases, it becomes clear from the inspection of the Skydot and Rotse1 catalogues that DO Her is also Rotse1 records J163040.95+232725.5 and J163040.88+232730.1 and Skydot 1630408+232727.
      (As I said they all have their own specific way of naming variables which does not make it easier). Now that we have accurate coordinates from the GSC, the identification becomes much easier.
      So concluding we can say that the GCVS identification of DO Her with GSC 2044.0901 must be correct, and that:
      DO Her=GSC 2044.0901=NSVS 7887289=NSVS 10699533=NSVS 1070407=ASAS 163041+2327.5=J163040.95+232725.5
      =J163040.88+232730.1=1630408+232727=2Mass 16304081+2327281.
      (I don't know if Skydot has any special way of notation of their variables).
       
      Phew, one variable identified, two more to go.
       
      Taking the GSC coordinates for GSC 2044.0674 as 163039.6+232641.6, and the GCVS position for V362 Her as 163039.5+232642 then we can safely say that this must be the same star.
      Unfortunately, there are no data for this star in NSVS so we cannot check it from that side.
      Using the GSC coordinates we can identify  V362 Her with ASAS 163040+2326.6, which unfortunately hardly has any observations (due to their hardware problems) so from that side it is also not possible to check it.
      Identification of V362 Her with ASAS 163035+2326.5 is very unlikely because there is a difference in coordinates of more than one arcminute. Unfortunately neither Skydot nor Rotse1 have further data on this star.
      (That this is really an RR Lyr star can be inferred from other publications that present a lightcurve for V362 Her that really looks like an RR Lyr star, see Simbad for details).
      Looking at the coordinates of the other GSC star 2044.0371 with 163035.3+232629.9 it seems likely that this is the star that can be identified with ASAS 163035+2326.5. So let's take a look at the NSVS data for this star. There are three records:
       
      NSVS number    coordinates
      7887206                163035.3 +232630.7
      10699483              163035.2 +232629.8
      10707327              163035.4 +232629.8
      From the coordinates it is possible to say that this is GSC 2044.0371.
      From the NSVS records it can be said that this is a slowly varying semiregular star with an amplitude of perhaps 0.3 magnitude. This is certainly not the RR Lyr variable that V362 Her is supposed to be.
      But this was already to be suspected because in the ASAS catalogue V362 Her is designated as RR Lyr with a period of 43.26... days which is more typical for SR than for RR Lyr obviously. So the ASAS identification of 163035+2326.5 with V362 Her is wrong.
      Unfortunately neither Skydot nor Rotse1 have further data on this star.
       
      So although the data are not complete, it can be concluded that:
       
      GSCnumber            GCVS identification            ASAS identification
      2044.0901                DO Her                                 163041+2327.5
      2044.0674                V362 Her                              163040+2326.6
      2044.0371                No name yet                        163035+2326.5 (this is a new variable)
       
      Back to VSX
       
      What I wanted to show with this example is how complicated identifications from the different catalogues and databases can be.
      What is also very evident from this (for me) is that it is absolutely necessary to have accurate coordinates and identifications from a standard catalogue like GSC (or 2MASS or UCAC2).
      There need to be links between the records in ASAS, NSVS, GCVS and VSX, and I can't see this being done without a standard star catalogue functioning as informational crossbridge.
      Put in practical terms: GCVS records need to specify the GSC ids of the variable, as well as ASAS, NSVS and VSX need to specify the GSC ids, so that identifications e.g. between an NSVS record and a GCVS variable can be made without any doubts, or the identification between an ASAS variable and a GCVS variable etc etc etc
       
      The GCVS is on it's way to do this, but it will be quite some time before they will be finished (not to mention all the checks that have to be made). ASAS does not have any identifications with star catalogues, nor does the NSVS (website).
      VSX has some of the identifications copied from the constituing databases (like BD, HD etc) but the GSC ids have been left out of this. Needless to say that I think this is a bad decision. Now VSX is missing the link to other databases for most of the variables, and everytime you want to compare data of VSX with other sources of information you need guess (or find out in the way I described above) if it concerns the same star or not.
       
      So what happens if you try to identify ASAS variables with the GSC? Does it result in better identifications and less errors in GCVS ids?
      To get a better picture of this and to see what issues arise when you try to do that, I did the following:
      I took the first 150 entries of the ASAS catalogue, looked up the data on the ASAS website, took the ASAS identification charts to identify the variable with Guide8 to find the GSC id and the coordinates of this GSC star. After looking up the coordinates in Guide8, they were checked with the GSC database in Vizier to avoid the problems in Guide8.
      The results are as follows:
      Of the 150 records, only 120 could be identified, mainly because of the hardware problem in ASAS at the moment, and also because for one ASAS variable the ASAS chart did not match the GSC chart, and for another ASAS variable Guide8 systematically crashed when trying to look it up on the basis of it's coordinates.
      All the 120 variables could be identified by their GSC numbers, but 16 of the 120 ASAS variables consisted of blends of several GSC stars (i.e. 13 %). One ASAS variable consisted of 4 GSC stars, one of 3, and the other 14 consisted of two GSC stars forming the ASAS variable.
      This introduces an additional problem, besides making the coordinates less certain because they might be from one of the (non-varying) neighbouring stars.
      The distance in arcseconds between the original ASAS position (derived from the identifier) and the GSC position is 5.1" on average, which is a much better score that I expected, also because this list included the blends. The biggest distance was 17" in one case.
      In this sample of 120 variables there were 28 variables with GCVS/NSV designations. Taking the coordinates from their matching GSC ids and checking these coordinates with the online version of the GCVS resulted in a positive id for all 28 of them. The distance between the GSC positions and the GCVS positions was less than 2" in 27 of the 28 cases, where one remaining case had a difference of 29". It was also found that Guide8 is sometimes incomplete: one variable was completely missing in Guide8.
      This means that the GCVS and the GSC positions have a very good match, and that the ASAS positions are more accurate than is suggested by the way the identifier is shown (with ra seconds and tenths of an arcminute).
      To my surprise, all the GCVS identifications made by ASAS of these 28 variables were correct. I have seen bad examples of GCVS identifications in the ASAS catalogue (distances between the two positions up till 4 arcminutes), but apparently this is also a psychological effect : The exceptions stand out very clearly and are more easily remembered while they form a very small percentage of the total number.
      Of course it is still debatable how representative this little sample is.
      This is what I mean when I say that the GSC needs to function as an informational crossbridge to get this job done. Someone with a different planetarium program could just as easily do this kind of thing with the 2MASS catalogue I suppose.
       
      I have used the same "technique" (use GSC as a standard for identification) to tackle another problem.
      Many new variables have been reported in the IBVS in the last few years. Many of those have been announced by their GSC number, and unfortunately the IBVS people have failed to set a standard for notation of the GSC numbers.
      It can go from GSC 02044-00901 to 2044.901 or 2044 0901 etc etc  When you check all the different formats that have been used, you can see that there are 24 (!) ways of notation for GSC numbers. Which is a ridiculous big number when you are trying to find data in the IBVS through their search page ( or other databases).
      So when you are preparing a list of new variables, you want to be sure that they have not been published before.
      To make a list for myself which variables had been announced in the IBVS, I scanned the IBVS issues for reports on new variables, and used Guide8 to identify them by their GSC numbers if possible and enter their coordinates in the resulting database as well.
      By using a strict format for the GSC numbers, then it becomes possible to match other lists with this list and identify typo's and other identification problems.
      This is quite a timeconsuming job, and I only finished a part of the 5600-5700 IBVS series, but it has already resulted in the identification of a GSC-mistake in IBVS 5690.
       
      From all this, an agenda for variable star workers for the coming years is slowly emerging:
      - NSVS records need to be identified with GSC stars, which is mainly a programming job because there are no identification charts involved but only coordinates. In this way the multiples can be identified as well. (I think there are a few blends in the NSVS as well, as I have come across a NSVS variable (not in the catalogue) where the position was halfway two GSC stars).
      - ASAS records need to be identified with GSC stars, but this is propably only possible in the practical sense for the ones in the ASAS catalogue. Here the problem of blends remains, and can only be solved in part by more accurate individual measurements or with the help of NSVS records (or other databases). Here a complicating factor would be for variables outside the ASAS catalogue that ASAS variables do not have a unique identifier when they are not in the ASAS catalogue.
      - USV, Rotse1 and Skydot catalogues have to be prepared in the same sense to find the matching GSC stars
      - VSX : the same thing and I would strongly recommend against giving the VSX records their own identifications, because it would only mean more administration, more sources of error, more fields in a database, more checks to made etc etc
      My idea is that this is the only way to end the identification circus, although using GSC in this pivotal role is a matter of choice.
      And my guess would be that this kind of agenda is not going to be followed up by professional astronomers. It is the kind of astronomical work that only pays off in the long run and will not get you into citation indexes.
       
      What do you think about it?
       
      Geert Hoogeveen
    • arne
      By GSC, do you mean GSC1.2? If so, the problem is that many fainter stars have no GSC number, and GSC itself is missing many stars (or has blends of yet more
      Message 2 of 2 , May 29 4:26 PM
        By GSC, do you mean GSC1.2? If so, the problem is that many fainter
        stars have no GSC number, and GSC itself is missing many stars (or
        has blends of yet more stars). I agree that identifying stars
        with their accurate coordinates is the safest method, but I am
        uncertain about using *any* catalog as a bridge.

        The reason why GCVS has such a good match with GSC stars is because
        they are identifying variables against the GSC (or, another
        astrometric) catalog where possible. Certainly the new, improved GCVS
        is just that - improved. There will still be cases where they made
        the wrong identification, but my guess is that this will be a smaller
        portion of the current rework.

        Improving the coordinates of the GCVS and NSV are thankless tasks,
        taking up much of the available GCVS manpower resources. However,
        it is only one area of variable-star cataloging. Currently, GCVS
        is about 1/3 of VSX, and within a year or so, will be significantly
        less of the total as new surveys come on-line.

        Thanks for your investigation description, Geert! It shows how
        crowded field identifications have problems, and this will only
        get worse in the future.
        Arne

        geert hoogeveen wrote:
        > (Prologue: This is more about the general questions that I have
        > concerning investigation methods than about VSX.)
        > While many technical details have already been discussed in this
        > discussion group, I have a more general problem that I would like to
        > present.
        >
        > Basis for the VSX are the presently available databases of the GCVS/NSV,
        > ASAS and NSVS (and some others that I do not know but that is not
        > important for this story). Each of these databases has it's own way of
        > naming variables and keeping records of the available data.
        > When you are investigating a variable, you are always forced to look up
        > every possible identification through various tables using coordinates,
        > names etc. This happens because none of the robotic surveys have made an
        > (serious) effort to relate their own observations to the variable star
        > catalogue or standard star catalogues.
        > The GCVS has now improved quite a lot on the point of coordinates and
        > GSC-identifications, but they are not there yet, and propably quite some
        > original finder charts have to be inspected as well.
        > ASAS did not make the connection to a star catalogue, but does include
        > GCVS identifications in their catalogue for several of their variables.
        > But be aware that there are quite some mistakes in their id's. Always
        > check them yourself (which can only be done when the GCVS has accurate
        > coordinates, which then because a bit of a circular argument).
        > The NSVS studies USV, Rotse1 and NSVS (Skydot) have only partly
        > connected to variable star or star catalogues. The Skydot catalogue has
        > GCVS identifications that are in general quite good, and also 2MASS
        > identifications that I guess are good as well, although I haven't
        > checked those. When you have accurate coordinates to begin with, it
        > makes it all much easier.
        > The USV and Rotse1 does not have either of those.
        >
        > Let me give an example to clarify this muddy way of working: The DO
        > Her/V362 Her/ASAS 163035+2326.5 example that John gave recently. If you
        > put it in words, the problem is like this:
        > DO Her is at 163040.8 +232728 (J2000) according to the GCVS and V362
        > Her is at 163039+232642. In the ASAS catalogue, the variable
        > 163035+2326.5 is identified as V362 Her. Is that correct? Put like this,
        > it will sound very vague to most people.
        > If you now use Guide8 or any other planetarium program to show the field
        > of DO Her you will see the problem more clearly. There are 3 GSC stars
        > in the immediate vicinity and both DO Her and V362 Her are swimming
        > somewhere halfway between these GSC stars. It is mostly guesswork to
        > decide which of the GSC stars is DO Her or V362 Her if you would only
        > use Guide or coordinates.
        > If I make a little table with the help of the GSC stars, it becomes even
        > more apparent. There are 3 GSC stars in this field
        >
        > GSCnumber GCVS identification ASAS identification
        > 2044.0901 DO
        > Her 163041+2327.5
        > 2044.0674 V362 Her 163035+2326.5
        > 2044.0371 ------
        > V362 Her
        >
        > From this little list it is already very obvious that GCVS and ASAS
        > disagree on the identification of V362 Her.
        > To decide which star is which variable, we need data from real
        > observations so here the data from NSVS and ASAS come in.
        >
        > For GSC 2044.0901 there are three NSVS records:
        > 7887289/10699533/1070407, that have the following coordinates:
        >
        > NSVS number coordinates
        > 7887289 163040.8 +232728.8
        > 10699533 163040.6 +232726.2
        > 1070407 163041.0 +232726.4
        > All these three coordinates fit nicely on the GSC coordinates of GSC
        > 2044.0901 which are 163040.9 +232728.3. The small differences are
        > measurement errors which are propably normal (?) and for a star of this
        > magnitude (around 11), you can safely say that this is the same star. If
        > you are dealing with much fainter stars, I am not so sure if that still
        > applies.
        > If you take a look at the lightcurves for NSVS 7887289/10699533/1070407
        > you can see that it has a typical longperiodic Mira lightcurve.
        > Although DO Her is not in the ASAS catalogue in their column of GCVS
        > identifications, DO Her can be identified with ASAS 163041+2327.5. It
        > would be even better if it could be identified with the ASAS
        > identification chart but due to the hardware problems with ASAS that is
        > not available at the moment.
        > (Remark: the situation with the three NSVS records is not immediately
        > clear: The average magnitude of the first one differs much from the two
        > other ones. By inspecting the coordinates and lightcurve it becomes
        > obvious that it is the same star in all three cases. The situation that
        > there are multiple entries for one and the same star happens quite a lot
        > in the NSVS).
        > If we now add the data from the other variable databases, it becomes
        > clear from the inspection of the Skydot and Rotse1 catalogues that DO
        > Her is also Rotse1 records J163040.95+232725.5 and J163040.88+232730.1
        > and Skydot 1630408+232727.
        > (As I said they all have their own specific way of naming variables
        > which does not make it easier). Now that we have accurate coordinates
        > from the GSC, the identification becomes much easier.
        > So concluding we can say that the GCVS identification of DO Her with GSC
        > 2044.0901 must be correct, and that:
        > DO Her=GSC 2044.0901=NSVS 7887289=NSVS 10699533=NSVS 1070407=ASAS
        > 163041+2327.5=J163040.95+232725.5
        > =J163040.88+232730.1=1630408+232727=2Mass 16304081+2327281.
        > (I don't know if Skydot has any special way of notation of their variables).
        >
        > Phew, one variable identified, two more to go.
        >
        > Taking the GSC coordinates for GSC 2044.0674 as 163039.6+232641.6, and
        > the GCVS position for V362 Her as 163039.5+232642 then we can safely say
        > that this must be the same star.
        > Unfortunately, there are no data for this star in NSVS so we cannot
        > check it from that side.
        > Using the GSC coordinates we can identify V362 Her with ASAS
        > 163040+2326.6, which unfortunately hardly has any observations (due to
        > their hardware problems) so from that side it is also not possible to
        > check it.
        > Identification of V362 Her with ASAS 163035+2326.5 is very unlikely
        > because there is a difference in coordinates of more than one arcminute.
        > Unfortunately neither Skydot nor Rotse1 have further data on this star.
        > (That this is really an RR Lyr star can be inferred from other
        > publications that present a lightcurve for V362 Her that really looks
        > like an RR Lyr star, see Simbad for details).
        > Looking at the coordinates of the other GSC star 2044.0371 with
        > 163035.3+232629.9 it seems likely that this is the star that can be
        > identified with ASAS 163035+2326.5. So let's take a look at the NSVS
        > data for this star. There are three records:
        >
        > NSVS number coordinates
        > 7887206 163035.3 +232630.7
        > 10699483 163035.2 +232629.8
        > 10707327 163035.4 +232629.8
        > From the coordinates it is possible to say that this is GSC 2044.0371.
        > From the NSVS records it can be said that this is a slowly varying
        > semiregular star with an amplitude of perhaps 0.3 magnitude. This is
        > certainly not the RR Lyr variable that V362 Her is supposed to be.
        > But this was already to be suspected because in the ASAS catalogue V362
        > Her is designated as RR Lyr with a period of 43.26... days which is more
        > typical for SR than for RR Lyr obviously. So the ASAS identification of
        > 163035+2326.5 with V362 Her is wrong.
        > Unfortunately neither Skydot nor Rotse1 have further data on this star.
        >
        > So although the data are not complete, it can be concluded that:
        >
        > GSCnumber GCVS identification ASAS identification
        > 2044.0901 DO
        > Her 163041+2327.5
        > 2044.0674 V362 Her 163040+2326.6
        > 2044.0371 No name yet
        > 163035+2326.5 (this is a new variable)
        >
        > Back to VSX
        >
        > What I wanted to show with this example is how complicated
        > identifications from the different catalogues and databases can be.
        > What is also very evident from this (for me) is that it is absolutely
        > necessary to have accurate coordinates and identifications from a
        > standard catalogue like GSC (or 2MASS or UCAC2).
        > There need to be links between the records in ASAS, NSVS, GCVS and VSX,
        > and I can't see this being done without a standard star catalogue
        > functioning as informational crossbridge.
        > Put in practical terms: GCVS records need to specify the GSC ids of the
        > variable, as well as ASAS, NSVS and VSX need to specify the GSC ids, so
        > that identifications e.g. between an NSVS record and a GCVS variable can
        > be made without any doubts, or the identification between an ASAS
        > variable and a GCVS variable etc etc etc
        >
        > The GCVS is on it's way to do this, but it will be quite some time
        > before they will be finished (not to mention all the checks that have to
        > be made). ASAS does not have any identifications with star catalogues,
        > nor does the NSVS (website).
        > VSX has some of the identifications copied from the constituing
        > databases (like BD, HD etc) but the GSC ids have been left out of this.
        > Needless to say that I think this is a bad decision. Now VSX is missing
        > the link to other databases for most of the variables, and everytime you
        > want to compare data of VSX with other sources of information you need
        > guess (or find out in the way I described above) if it concerns the same
        > star or not.
        >
        > So what happens if you try to identify ASAS variables with the GSC? Does
        > it result in better identifications and less errors in GCVS ids?
        > To get a better picture of this and to see what issues arise when you
        > try to do that, I did the following:
        > I took the first 150 entries of the ASAS catalogue, looked up the data
        > on the ASAS website, took the ASAS identification charts to identify the
        > variable with Guide8 to find the GSC id and the coordinates of this GSC
        > star. After looking up the coordinates in Guide8, they were checked with
        > the GSC database in Vizier to avoid the problems in Guide8.
        > The results are as follows:
        > Of the 150 records, only 120 could be identified, mainly because of the
        > hardware problem in ASAS at the moment, and also because for one ASAS
        > variable the ASAS chart did not match the GSC chart, and for another
        > ASAS variable Guide8 systematically crashed when trying to look it up on
        > the basis of it's coordinates.
        > All the 120 variables could be identified by their GSC numbers, but 16
        > of the 120 ASAS variables consisted of blends of several GSC stars (i.e.
        > 13 %). One ASAS variable consisted of 4 GSC stars, one of 3, and the
        > other 14 consisted of two GSC stars forming the ASAS variable.
        > This introduces an additional problem, besides making the coordinates
        > less certain because they might be from one of the (non-varying)
        > neighbouring stars.
        > The distance in arcseconds between the original ASAS position
        > (derived from the identifier) and the GSC position is 5.1" on average,
        > which is a much better score that I expected, also because this list
        > included the blends. The biggest distance was 17" in one case.
        > In this sample of 120 variables there were 28 variables with GCVS/NSV
        > designations. Taking the coordinates from their matching GSC ids and
        > checking these coordinates with the online version of the GCVS resulted
        > in a positive id for all 28 of them. The distance between the GSC
        > positions and the GCVS positions was less than 2" in 27 of the 28 cases,
        > where one remaining case had a difference of 29". It was also found that
        > Guide8 is sometimes incomplete: one variable was completely missing in
        > Guide8.
        > This means that the GCVS and the GSC positions have a very good match,
        > and that the ASAS positions are more accurate than is suggested by the
        > way the identifier is shown (with ra seconds and tenths of an arcminute).
        > To my surprise, all the GCVS identifications made by ASAS of these 28
        > variables were correct. I have seen bad examples of GCVS identifications
        > in the ASAS catalogue (distances between the two positions up till 4
        > arcminutes), but apparently this is also a psychological effect : The
        > exceptions stand out very clearly and are more easily remembered while
        > they form a very small percentage of the total number.
        > Of course it is still debatable how representative this little sample is.
        > This is what I mean when I say that the GSC needs to function as an
        > informational crossbridge to get this job done. Someone with a different
        > planetarium program could just as easily do this kind of thing with the
        > 2MASS catalogue I suppose.
        >
        > I have used the same "technique" (use GSC as a standard for
        > identification) to tackle another problem.
        > Many new variables have been reported in the IBVS in the last few years.
        > Many of those have been announced by their GSC number, and unfortunately
        > the IBVS people have failed to set a standard for notation of the GSC
        > numbers.
        > It can go from GSC 02044-00901 to 2044.901 or 2044 0901 etc etc When
        > you check all the different formats that have been used, you can see
        > that there are 24 (!) ways of notation for GSC numbers. Which is a
        > ridiculous big number when you are trying to find data in the IBVS
        > through their search page ( or other databases).
        > So when you are preparing a list of new variables, you want to be sure
        > that they have not been published before.
        > To make a list for myself which variables had been announced in the
        > IBVS, I scanned the IBVS issues for reports on new variables, and used
        > Guide8 to identify them by their GSC numbers if possible and enter their
        > coordinates in the resulting database as well.
        > By using a strict format for the GSC numbers, then it becomes possible
        > to match other lists with this list and identify typo's and other
        > identification problems.
        > This is quite a timeconsuming job, and I only finished a part of the
        > 5600-5700 IBVS series, but it has already resulted in the identification
        > of a GSC-mistake in IBVS 5690.
        >
        > From all this, an agenda for variable star workers for the coming years
        > is slowly emerging:
        > - NSVS records need to be identified with GSC stars, which is mainly a
        > programming job because there are no identification charts involved but
        > only coordinates. In this way the multiples can be identified as well.
        > (I think there are a few blends in the NSVS as well, as I have come
        > across a NSVS variable (not in the catalogue) where the position was
        > halfway two GSC stars).
        > - ASAS records need to be identified with GSC stars, but this is
        > propably only possible in the practical sense for the ones in the ASAS
        > catalogue. Here the problem of blends remains, and can only be solved in
        > part by more accurate individual measurements or with the help of NSVS
        > records (or other databases). Here a complicating factor would be for
        > variables outside the ASAS catalogue that ASAS variables do not have a
        > unique identifier when they are not in the ASAS catalogue.
        > - USV, Rotse1 and Skydot catalogues have to be prepared in the same
        > sense to find the matching GSC stars
        > - VSX : the same thing and I would strongly recommend against giving the
        > VSX records their own identifications, because it would only mean more
        > administration, more sources of error, more fields in a database, more
        > checks to made etc etc
        > My idea is that this is the only way to end the identification circus,
        > although using GSC in this pivotal role is a matter of choice.
        > And my guess would be that this kind of agenda is not going to be
        > followed up by professional astronomers. It is the kind of astronomical
        > work that only pays off in the long run and will not get you into
        > citation indexes.
        >
        > What do you think about it?
        >
      Your message has been successfully submitted and would be delivered to recipients shortly.