Of course. In fact, ground based astrometry will get considerably better, including retroactively, thanks to Gaia, as the background reference frame will become much better. This will especially useful with asteroids. Modern ground based astrometry for unresolved asteroids (ones smaller than the telescope's resolution) are good to ~ 10 milli arcseconds (mas) (yes), except that errors in the background stars and the timing push their errors up to > 100 mas. (In other words, asteroid data can fit to 10 mas in one night, but when many nights are added together, and especially when the asteroids move so far the reference stars change, the residual fit goes up by a factor of 10 or more.) Gaia will improve than considerably. Gaia will also not be better than 1 mas for many asteroids, as it is most accurate for ~12th magnitude, and an asteroid that bright will be resolved. If ground-based asteroid observations can be brought down to 10 mas, then Gaia is not beyond competition (as 100's of less accurate observations could be acquired in the time between Gaia scans). When you add in that many asteroids are fainter than 20th magnitude (the Gaia limit) much or all of the time, that small Near Earth asteroids will have perceptible Yarkovsky accelerations, and that Gaia will be puzzled about some objects (e.g., fast moving new asteroids or long-period binary stars), there will still be lots for ground based observers to do.

Hi Marshall, thanks for the answer, yes asteroid astrometry will definitely require ground-based astrometry, you have summed it up very well. We have just put in a proposal to obtain an existing 0.75 m telescope (being replaced by a 'modern' 1.0 m) at SAAO Sutherland, which I would like to equip for astrometry. We are trying to densify the ICRF in the south (have just added a 22 Ghz receiver to the 26 m radio telescope), and perhaps there will be interesting opportunities to do some work on optical counterparts for new southern ICRF sources.

There will always be a future in ground based astronomy unless space based telescopes become more cost efficient. TMT (Thirty Meter Telescope) being built in Hawaii for example is of similar price and can do similar work with its AO system. If we could achieve better cost efficiency's with the space based telescopes is the only way to totally eradicate ground based observing.

Hi Derek, it's amazing what can be done with AO, I see that some cameras have go t built in AO now at relatively good prices.

It is amazing. Being a student, at the base of the best (the Chileans will argue this) site for observational astronomy, has given myself opportunities to work with some of these telescopes, and some of the image resolution that can be achieved is purely amazing in and of itself for ground based telescopes

Marshall and Ludwig already answered in part: I will say that very likely ground-based observations of all the small bodies (not only asteroids) in our solar system will indeed continue; also, as there will always be some new radiation sources detected - even after the GAIA mission ends - and these sources will have to be pinpointed. Besides, if the technology and techniques improve, we might continue to increase precision of the ground-based instruments beyond what we now consider to be the limit. There is always possibility for a breakthrough, so I wouldn't write off ground-based astrometry anytime soon. Not to mention a check of some measurements. Perhaps we will see development of a real astrometric network in the future?

To complete a bit the excellent answers given before, once Gaia is finished, we will need to go on observing Solar System objects, even if the precision of the observations we obtain are lower than those of Gaia.

Gaia observations will not have an infinite precision, and will be biaised, like any instrument. Because of these errors and uncertainties, the accuracy of the ephemeris obtained will be limited in time. The positions obtained for the observation time-span will be very accurate, and will begin to degrade with time once we do not observe the bodies anymore. This is how some irregular satellites of Jupiter and some asteroids are now considered as lost: we haven't observed tham for a long time, and their real position is now far from the one given by the dynamical models used. In fact, for long-term accuracy, it is better to have a long time span with less precise observations, than a very short time span of precise observations.

So, for Solar System objects, Gaia will definitely not be the end of observations, but a much needed step toward mas/submas accurate ephemeris.

Dear Colleagues: I think that a 3 component "collision" of participants is important in the future in search of Horseshoe orbiting satellites of the Earth: where ground based, satellite based observations and orbital calculations in celestial mechanics should cooperate

Hi Szaniszlo, these Horseshoe satellites....are you referring to asteroids? Can you elaborate a bit...I don't quite understand...

The satellites with horseshoe orbit belong (are attached) to a planet, but their orbit has a coverage like a horseshoe, in the reference system coorbiting with the planet. They need very good resolution obseravtion. Here I attache a paper about them. Wajer, P. (2010): Dynamical evolution of Earth’s quasi-satellites: 2004 GU9 and 2006 FV35. Icarus, 209, 488–493. or: Taylor, P. A. et al (2004): Properties of Horseshoe Object 2000 PH5 From Radar Observations. American Astronomical Society, DPS meeting #36, #46.06; Bulletin of the American Astronomical Society, Vol. 36, p.1181. or: Taylor, Patrick A.; et al. (2007): Spin Rate of Asteroid (54509) 2000 PH5 Increasing Due to the YORP Effect. Science, Volume 316, Issue 5822, pp. 274- (2007).

Connors, M.; Innanen, K. (2004): Resonant and Co-orbital Asteroids Associated with Earth. American Geophysical Union, Spring Meeting 2004, abstract #P11A-04.

From Gaia, we are expecting a further revolution in the observation of occultations of stars by asteroids, similar to what happened for the larger main-belt asteroids from HIPPARCOS's results. With Gaia results, we should be able to routinely be able to predict and observe occultations by smaller angular-size objects of interest, including many more TNO's and the small near-Earth asteroids. These efforts will need continuing support of accurate ground-based astrometry after the Gaia mission ends.

Oliver Manuel - off topic, take it somewhere else SVP.

The case of Halton Arp is interesting and sad at the same time. GAIA will detect thousands of additional quasars, of which about 1% (going on current statistics) will be radio sources. We are working on the ICRF by adding more quasars to the Southern Hemisphere, so the questions raised by Halton Arp can be laid to rest at some stage as the density and geometry is sorted out.

I do not want to mention about ground based astrometry but want to mention about space astrometry via infrared wavelength, JASMINE.

JASMINE seems pretty cool - http://www.jasmine-galaxy.org/pub/Gouda-proc.pdf

Is this still on schedule for the early part of the 2020s ?

Shogo, what has happened to the nano and small Jasmines which would have been precursors? Are these going ahead?

Parameters of the experiment GAIA help clearly delineate the astrometric applications of ground observations. Please note that not all of them are yet mentioned during this discussion. Billion objects brighter 20m will be observed of about 100 times, coordinate precision weakest of them will be worse than 100 microarcseconds. Two obvious areas were discussed in detail here: objects in the Solar system moving along complex trajectories and weak quasars. We should also mention the star triples and greater multiplicity; increase the density of reference frame for monitoring objects fainter than 20th magnitude in narrow fields of view of large telescopes, when the accuracy of the optical realization of ICRF using only objects of GAIA will be insufficient; ensuring astrometric observations of variable objects of large amplitude of brightness, in particular the systematic search of optical transients. I believe that the inherent limitations of GAIA, will likely serve to expand astrometric work on a new level in an era "after GAIA" than to their disappearance.

Hi Valerian, I think you are quite right that more astrometric work at a higher level will be required. There should be a great demand for follow-up and maintenance, particularly concerning proper motion and celestial reference frame issues, and in addition, much work will be created concerning the ties between optical and radio reference frames.

@Valerian - I would agree. Clearly, optical astrometry of stars will also be improved by Gaia, and will also need to continue. As just one example, there are many multiple star systems with periods longer than the duration of the Gaia mission (some known, some unknown). The observations of such systems will obviously have to continue if their orbits are going to be determined.

Concerning JASMINEs, see www.scholarpedia.org/article/JASMINE.

My understanding from a past GAIA conference is that many star clusters will not be well-studied with GAIA because of crowding issues; the same would therefore apply to the dwarf galaxies of the Local Group, even if the originally designed precision were available. And of course, every space mission has a limited lifetime, so limited numbers of visits per targets. Therefore, it seems to me, there will be ample work for ground-based astrometry in the future just as there is for photometry and spectroscopy.

Ground-based astrometry will have to adapt to a new environment after a successful Gaia mission, but it will survive in its transformed shape. The epoch of dedicated small and medium astrometric telescopes with CCD cameras will be over. But the large etendue, global surveys such as Pan-STARRS will become even more important in the world of astrometry, because they go a magnitude or two fainter than Gaia, and there are lots of interesting sources there, as usual. Pan_STARRS and later, LSST, will have to be aligned with Gaia. Resolving tight binaries and improving astrometric orbits (e.g., with speckle) is not going away. It is not clear how well Gaia will perform on the brightest stars (

Hi Valeri, yes it is going to get quite complicated. Apart from the rotations, weak and micro-lensing will be the next thing to worry about. Marshall mentioned that the solar system motion is to be addressed by the IAU sometime soon as far as it affects the ICRF.

averages out. 

Dear Valeri;

The galactic acceleration imposes a first order poloidal (i.e., curl free) vector spherical harmonic (VSH) proper motion on extragalactic sources, which is orthognal to rotations of the ICRF (technically,  first order toroidal VSHs). Of course, nothing is perfect, and in the real world there will be some correlation between poloidal and toroidal VSH, but in principle they could be separated. Now, the expansion of the cosmic microwave background radiation indicates that the net rotation of the CMB is very, very small, but I know of no way to tie the ICRF to the CMB, so the best we can do is take the average over all sources (with some weighting) and hope whatever motions the sources have

Dear Marshall,

You are absolutely right, that the effects of the Galactic acceleration and rigid spin are well separated (or, orthogonal) in the space of spherical harmonics. The problem is that if you constrain extragalactic sources' proper motions to zero already in the astrometric global solution (which was our plan A for the Space Astrometry Mission, SIM), you will fix the rotation but introduce a systematic error from the acceleration dipole.

Perhaps the right way to do this is to solve for all 5 astrometric parameters for extragalactic sources too, and then take the harmonic decomposition of the resulting proper motion field, and subtract only the dipole terms. Perhaps, this is the plan for Gaia, but it is risky, because such unconstrained solutions can be poorly conditioned (too many correlated unknowns are involved in the adjustment). But if it works, you can get the acceleration vector, and, who knows, may be the relic gravitational waves in the higher-degree terms you have been publishing about.

Should read "subtract only the rotation terms" in my previous mail, of course.

In the Soviet Union about 30-40 years ago during competitions to Americans in the space program, there was such joke: "All police went to Space". If all astronomers after police depart to space, is valid on Earth police officers, also astronomers doesn't remain, there are only crying widows and their children on Earth.

Alexander :)

Dear Ludwig,

Quite right, strong and weak gravitational lensing will be a challenge for the next generation of astrometric instruments. The strong lensing, when the lense crosses the Einstein disk, is intrinsically rare, but the astrometric signal can be detectable at the milliarcsecond level. The low rate of such events would require a large etendue astrometric telescope. Unfortanutely, the Pan-STARRS and the coming LSST are not precise enough for this work (I believe). The rate of weak lensing events is critically dependent on the mass function of the smallest stars and solivagant (free-floating) planets. As we are getting a better understanding of the former, the latter remains practically unknown. It can be shown that if the power of the mass function for solivagant planets remains the same as for late M dwarfs, or steeper, such objects should dominate the weak lensing stochastic scintillation despite their low mass. It is an exciting possibility, but not for the present day instrumentation. The weak lensing is measurable at the microarcsecond level of precision. We have to go up 1-2 orders of magnitude in precision with respect to Gaia, think long-base or formation flying interferometer in space.

I can think of two specific examples in my own research field where ground-based observatories will still be needed.

To get a full picture of the chemical and dynamical properties (and history) of our Galaxy, you need chemical abundances! The capabilities of Gaia to do this are quite limited, in the sense that the spectroscopy from the mission will not really be able to measure detailed abundances except in very bright (and usually nearby) stars. In addition, the measurement of radial velocities to any reasonable precision will not be possible with Gaia below some magnitude which is far brighter than the limits where it will obtain good parallaxes and proper motions.

Modern 8-m class telescopes can obtain spectroscopy , do chemical abundance analyses and find radial velocities (to

Hi all. One thing for sure that will give us plenty of work on ground astrometry after GAIA is the astrometric calibration of all the ongoing/upcoming  large very deep surveys like LSST, PanSTARR and VISTA. It would be impossible to do it without GAIA but it will take some good detailed work to have it done. The future I guess is to extend the optical reference frame deeper in magnitude and also tie it to the ICRF in radio, sooner than later.

It's hard to avoid giving examples of the continuing need for ground based astroNOMY, but keeping it on astroMETRY, my global answer is: GAIA will observe for only 5-6 years, but the sky will keep changing. E.g., Anomalous motions of binary and multiple stars have orbital time scales longer than that. As David Dunham pointed out, astrometry of solar system objects will require continued monitoring to further refine ephemerides and keep track of existing discovered objects. And back to astronomy, for occultation studies, there will ALWAYS be a need for more ground-based telescopes, where larger apertures will be able to observe more events involving fainter target stars

Dear Sirs,

 I the first time answered this question by means of a joke, but this time I will give two serious forecasts and in advance I ask astrophysicists not to appropriate two of my deep theoretical scientific results.

1. It is necessary to leave telescopes and field-glasses on Earth since in space there was a lot of space debris which strongly scratches glasses of space telescopes, sharply lowering their resolution therefore in them it is visible nothing.

2. Industrial activity of the person causes greenhouse effect and destruction of the Terrestrial atmosphere. Soon on our planet Earth the atmosphere will become same transparent as on the planet Mars both all stars and black holes can be observed without hindrances through magnifying glass of an average aperture.

Sincerely, Alexander

Greetings to all.

The microarcsecond astrometry after Gaia in reality rather will be on the level of 0.05 milliarccecond because of the correlation between the parameters of the movement of technical systems (space craft vibrations, orbital and rotational movement, parameters of the instrument and its registration system, timing and pure astrographyc fits and probably some others sources, that I cannot foreseen). We must accept the opinion of David Dunham about observations of the Solar system bodies. Including, surely, the ground-based astrometric observations.

The real problem does exist in the fitting (rather recreation) of the Celestial coordinate frame, the movements of the celestial bodies in which the real theories of baricentric  movements are traditionally been constructing on the base of the laws of the mechanical energy conservation. There are plenty of the work in that area for which the ground-based astrometry will be required. 

It is reasonable to wait the appearance of the other space projects for detalization of the hardly solving question in Gaia project and independent other questions. For instance, involving or prolongation of the Gaia's results into diapason of the more faint stars. See, for instance, Baltic Astronomy, 24, 84–91, 2015.   

Yes, there is future. We got future.