possible to build an orbiting astrometric telescope, associated with the manned 

 Space Station, which would be capable of positional accuracy of 10 micro- 

 arcsec on stars in limited areas of the sky for periods as long as 10 to 20 years. 

 This accuracy exceeds current ground-based performance by two orders of mag- 

 nitude. With such a capability it would be possible to detect, with some 

 assurance, planets like Uranus around the nearest few hundred stars, and Jupiter- 

 like systems would, if they exist, be easily found. A reasonable statistical sample 

 of other planetary systems could be developed with an astrometric telescope in 

 orbit, or very meaningful limits could be placed on their existence. 



As already stated, spectroscopic searches for extrasolar planetary systems are 

 little affected by the atmosphere. As long as the apparent brightness of the star 

 is great enough, this technique can provide exactly the same capability to find 

 planets orbiting distant stars as it can for nearby ones. Systems with the required 

 stability to recognize the periodic wavelength shifts induced by a Jupiter-mass 

 planet already exist; detection of the smaller shifts induced by terrestrial-type 

 planets will ultimately probably be limited by the atmosphere. Current ground- 

 based efforts are now limited by the lack of detailed information about the pres- 

 ence of any long-term periodic or quasi-periodic changes within the stellar 

 photospheres that could mimic or mask the planetary velocity signature. It 

 seems quite likely that this technique will begin to pay off consistently in the 

 near future. 



Photometric searches capable of monitoring 4000 stars continuously in two 

 colors with an absolute precision in the luminosity determination of 10" 6 might 

 detect about one planet per month. Most of the detections will be the short- 

 period planets, and planets as small as Earth may not be detectable during the 

 active phases of the stellar cycles. A number of technological obstacles must be 

 overcome before this technique can be demonstrated. This may eventually be 

 possible with instrumentation proposed for the Block II Space Station; work is 

 beginning on the required technology. Photometry could prove to be an espe- 

 cially critical technique for studying the distribution of mass and orbits in other 

 planetary systems, since it works best for the short-period inner planets, unlike 

 the other approaches. 



In contrast to photometry, the opportunity for making an early start on an 

 orbiting astrometric telescope has never been clearer or more promising. An 

 orbiting manned Space Station would ensure that the necessary orbiting super- 

 structures will be available to allow the development and operation of a dedi- 

 cated astrometric telescope at very low cost in a reasonable time. The manned 

 Space Station is a key element for a number of reasons. First is an ability to per- 

 form consistent on-orbit maintenance and service, which will ensure that the 

 system can operate consistently over the required time. Second, although the 

 telescope system should be designed for largely routine, automatic operation, 

 there would no longer be a requirement for it to be completely autonomous, 

 thus reducing the cost substantially. Third, on-orbit integration with the Space 



56 



