From: hitchner@riacs.edu (Lewis Hitchner)
Subject: Re: Virtual Mars
Date: Wed Jan 16 19:37:30 1991 GMT
Organization: Research Institute for Advanced Computer Science



As several respondents have noted, NASA has already begun
implementation of a "Virtual Mars" system.  Mike McGreevy has publicized
the concept of virtual planetary exploration for several years in both
spoken (e.g., keynote speech at ACM's CHI '89) and published form.
We also have an on-going project (on-going in spite of recent federal
budget cutbacks in NASA's Mars-Lunar Initiative program) entitled
"Visualization for Planetary Exploration" that already implements many
of the concepts proposed by Shawn Carlson in his article.  However, our
project is merely a prototype and the "reality" is somewhat limited due
to the limited resolution data currently available (digital elevation
data derived from Viking orbiter satelite imagery) and due to the limitations
of current display technology and graphics hardware technology.  But,
even given these limitations, we have been successful (at least, in our
opinions) in demonstrating that virtual planetary exploration is feasible
-- in a limited way today, and in a much more realistic way in the not too
distant future.  The system we are developing is intended for use by
research scientists, such as the NASA, USGS, and university planetary geologists
who are funded by NASA to study geological features of the planets (and, I might
add, who have been doing so for a number of years using both remotely gathered
satellite data as well as first hand data collection at terrestial study sites
similar to Martian terrain -- Death Valley, and Antartica).  Some initial
presentations of our system to some of these people have indicated that they
feel such systems will be very useful in their research (however, they also
strongly support the need for human presence for examining micro-features).

A few extra comments:

- Carlson and one of the follow-up'ers referred to data resolution and the
  quantities needed, storage problems, etc.
  Our data is quite low resolution -- 1 km grid spacing, measured in meters
       16 bits per elevation value.  That's pretty crude when you want to see
       things smaller than continents.  BUT -- we have that data for the whole
       planet.  So, it amounts to nearly 1/2 GigaByte (we also have another
       9/10 GByte of higher resolution digital imagery or texture map data for
       about 35% of the planet).  By the way, this data is in the public domain
       and is available from NASA Goddard's NSSDC (Natl. Space Sciences Data
       Center).
   What we are working with now is orders of magnitude less data than what will
       be available real soon.  Venus is now being imaged by Magellan's SAR
       radar imager which is moderately high res. (I don't know figures, but
       I believe it is somewhat comparable to USGS DEM data which is 150 meter
       grid spacing, i.e., roughly 36 times the res. of our Viking Mars data).
   In 1992 or later (depending on funding) NASA will launch the Mars Observer
       satellite.  MO has a laser altimeter that can directly measure surface
       topography.  It will be capable at its highest resolution of measuring
       surface elevations to 1.5 meter resolution at similar spatial sampling
       as that of Magellan I believe (maybe higher).

- regarding who might use such systems
  Though I agree strongly with Carlson's feelings regarding virtual versus
  real exploration, I'd take exception to his suggestions about how they would
  be used and by whom.  True, there is a great deal of educational and 
  entertainment potential for virtual exploration systems.  However, our
  "prototype" system with its limited performance requires about 1/4 million
  dollars of computer equipment and highly trained technicians to operate it.
  Hopefully, some day system development and operation cost will be much, much
  lower.  But, for the near future I can forsee such systems primarily used by
  highly trained experts in planetary geology.  We can't really afford to devote
  use of such systems to non-specialists who spend "a week" being trained to
  use the system (by comparison Shuttle astronauts spend 1 to 2 years training
  for each mission, AFTER their years of general astronaut training, AFTER their
  years of aviation or science training and education).  Initially, at least,
  I think virtual exploration systems would be most productive when used by
  experts.  For now, I think it would be possible to provide useful educational
  or entertainment systems without the data accuracy requirements of a scientist
. 
  Current commercial systems can satisfy this audience by presenting good lookin
g,
  although physically inaccurate, scenes at a much better cost/benefit ratio
  than technically precise systems such as ours. Of course, as technology marche
s
  on, we all hope and expect that such systems will become available to a much
  less restricted audience -- school children, citizens, thrill seeking "virtual
  explorers" -- for everyone to use.