NEOs, Public Safety,

1
(No Transcript)
2
NEOs, Public Safety, Massive Databases

• Dr. Jim Heasley
• Institute for Astronomy
• University of Hawaii

3
Deutsche Museum Planetarium

• February 8, 2001

4
Talk Outline

• Early Asteroid Discoveries
• Basic Celestial Mechanics
• The Public Safety Hazard
• The New Look Solar System
• Asteroids Close Up

• Modern Asteroid Searches
• Observing Strategies
• The Data Analysis Problem
• Orbit Determinations
• Whats Next?

5
Early Asteroid Discoveries

• The Titus-Bode law predicted the presence of a
  planet between Mars Jupiter.
• In 1800, Baron von Zach organized a systematic
  search of the ecliptic to look for the missing
  planet.

6
Early Asteroid Discoveries

• March 1802, Oblers discovered Pallas. Juno was
  discovered in 1804, Vesta in 1807. No more minor
  planets were discovered until 1845.
• By 1890, more than 300 minor planets had been
  discovered by visual observations. Most of these
  are located in the asteroid belt between Mars and
  Jupiter.

• On January 1, 1801, Giuseppe Piazzi discovered
  Ceres. By the summer of 1801, it was lost.
• Gauss developed a new method of calculating
  orbits. Von Zach reaquired Ceres on December 31,
  1801.

7
Early Asteroid Discoveries

• In 1891 Max Wolf introduced the technique of
  astronomical photography. The motion of an
  asteroid causes a streak on the photograph.
• Photographic searches can cover a larger area and
  are more sensitive to faint objects than visual
  ones.

8
Basic Celestial Mechanics

• Kepler described the basic laws of planetary
  motion between 1609 and 1619. Later Newton showed
  these laws are the natural consequence of an
  object moving under the force of gravity.

• Kepler showed that planetary orbits are conic
  sections, in particular ellipses.
• For a planet on an elliptical orbit, it moves
  faster when near the Sun, and slower when it is
  farther away, sweeping equal areas in equal times.

9
(No Transcript)
10
(No Transcript)
11
Basic Celestial Mechanics

• Gravitational forces from the major planets cause
  asteroid orbits to change with time. Just because
  an asteroid is on a safe orbit today, sometime
  in the distant future it might become a threat.

• Asteroids that start on nearly circular orbits
  inside the asteroid belt can be perturbed over
  time into orbits that cross into the inner solar
  system.

12
The Public Safety Hazard

• The evidence for frequent celestial
  bombardment is clearly visible on the Moons
  surface and that of Mercury, and to a lesser
  degree on Venus and Mars. The realization that
  these craters are the result of impacts rather
  than volcanoes is rather recent.

13
The Public Safety Hazard

• On Earth, these scars have been erased by
  water wind erosion and plate tectonics.
  However, careful study has turned up a large
  number of features that appear to be impact
  craters.

14
(No Transcript)
15
The Public Safety Hazard

• Meteor crater in Arizona was formed about 50,000
  years ago. The impacting iron meteorite had a
  mass of 1 million tons and diameter of 50 meters.
  The impact blast was equivalent to a 20 megaton
  nuclear weapon and produced a crater over 1 km in
  diameter.

16
The Public Safety Hazard

• In 1908 a blast equivalent to a 15 megaton
  bomb occurred near the Tunguska river in Siberia.
  It occurred about 8 km over the Earths surface,
  flattening trees over a thousand square
  kilometers. The object was a stony body with a
  mass of 100,000 tons.

17
The Public Safety Hazard

• Even more recently, an impact took place over
  Siberia near Vladivostok in February 1947.
  Craters up to 28 m across were produced and more
  than 23 tons of iron meteorite fragments were
  recovered.

18
The Public Safety Hazard

• The impact that is thought to have killed off the
  dinosaurs -- the Chicxulub crater, which is off
  the Yuchatan peninsula in the Gulf of Mexico--was
  produced by an object at least 10 km in diameter,
  resulting in a crater 200 km in diameter.

19
The Public SafetyHazard

• The Chicxulub impact happened about 65 million
  years ago. It is important to remember that while
  this is a large number on human time scales, in
  relation to the age of the solar system this was
  just yesterday!

• Locally, a crater about 20X larger than the
  impacting object is produced.
• Pulverized rock is injected into the stratosphere
  blocking sunlight.
• Fires put large amounts of carbon dioxide into
  the atmosphere creating a greenhouse effect.

20
The Public Safety Hazard

• The injection of massive amounts of dust into the
  upper atmosphere will create an extended winter
  on Earth, killing off the plant life and
  eventually much of the animal life as well.

• The effect will be far worse than those we feared
  from a man-made nuclear winter, and those
  surving the immediate impact will suffer over an
  extended period as the food supply disappears.

21
(No Transcript)
22
(No Transcript)
23
The Public Safety Hazard

• Public awareness of the danger from celestial
  impacts was heightened by the break up of comet
  Shoemaker-Levy 9 after a close pass by Jupiter.
• On its next orbit, the fragments collided with
  Jupiter producing atmospheric explosions.

24
The Public Safety Hazard

• Awareness of these dangers was heightened by
  Hollywoods blockbuster movies Deep Impact and
  Armageddon.
• More money was spent on making either of these
  movies than is spent for research on these very
  real threats.

25
The New Look Solar System

• Most of the asteroids discovered are located in
  the main belt between Mars Jupiter.
• There are families of asteroids that occur
  elsewhere in the solar system, both inside and
  outside the main belt.

• There are currently over 20,000 asteroids with
  good known orbits (10 year predictability).
• There are another 50,000 asteroids with orbits
  good enough to predict positions over a short
  period ( 2 years).

26
The New Look Solar System

• Earth-approaching asteroids include at least 4
  types
• Amors Mars orbit crossing asteroids with
  perihelia between 1.017 and 1.4 AU
• Apollos Earth-orbit-crossing asteroids with
  semimajor axes 1 AU
• Atens Semimajor axes
• Apoheles proposed name for asteroids with orbits
  totally inside 1 AU.

27
(No Transcript)
28
The New Look Solar System

• In the outer solar system, things are not as
  empty as once thought
• Centaurs are asteroids with orbits that cross
  inside Jupiter and Neptune.
• Kuiper Belt Objects are located out beyond
  Neptune.

29
Asteroids Close Up

• With the Hubble Space Telescope and NASA
  missions to the outer solar system we have had a
  good looks at these rather large asteroids
• Vesta (from the HST), 325 km diameter
• Gaspra, 20 km X 12 km X 11 km
• Ida, 56 km X 24 km X 11 km
• Mathilde, 52 km diameter
• Eros, 20 km long

30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
Modern Asteroid Searches

• In 1991 the U.S. Congress directed NASA to
  conduct workshops on how potentially threatening
  asteroids could be detected, and how they could
  be deflected or destroyed.

• How many asteroids have orbits that cross that
  of the Earth? In the early 1990s, the thinking
  was
• 400 with diameters 2 km
• 2,000 1 km diameter objects
• 9,200 asteroids 0.5 km in diameter

36
(No Transcript)
37
Modern Asteroid Searches

• In 1994 the House Committee on Science and
  Technology directed To the extent practicable,
  NASA, in coordination with the Department of
  Defense and the space agencies of other
  countries, shall identify and catalogue within 10
  years the orbital characteristics of all comets
  and asteroids that are greater than 1 km in
  diameter and are in an orbit around the sun that
  crosses the orbit of the Earth.

38
Modern Asteroid Searches

• Among the active asteroid search programs in
  the U.S. are
• Spacewatch (Kitt Peak in Arizona)
• NEAT (JPL using USAF Maui telescopes)
• LINEAR (Lincoln Labs at White Sands, NM)
• LONEOS (Lowell Observatory)
• University of Hawaii Search (Mauna Kea)

39
Modern Asteroid Searches
40
(No Transcript)
41
Modern Asteroid Searches

• Whats Not Been Found?
• Any killer asteroids that are on a collision
  course with Earth. (This is a good thing, but
  that doesnt mean there arent any out there!)
• The number of detected NEOs appears to be about a
  factor of 2X lower than previously expected.

• Rabinowitz has argued the number of 1 km sized
  Earth crossers is about 750, not 2000.
• However, given we are not rediscovering the
  same objects at a high rate, others have argued
  there are probably more like 1000-1500 such
  objects.

42
Observing Strategies

• To conduct searches that will identify the
  potential killer asteroids we have two basic
  requirements
• Sky coverage
• Limiting magnitude.
• For a given telescope camera system, these are
  tradeoffs -- you cant have both.

• Other factors enter into observing strategy. An
  important one is the overhead in reading the
  digital camera.
• Another is the smear of a moving object -- its
  flux is spread out over several pixels in the
  image. The faster it moves, the more smear.

43
Observing Strategies

• Most searches hunt for asteroids when they are in
  opposition to the Sun relative to the Earth where
  they are illuminated like a full moon. There the
  NEOs move slowly and may be hard to distinguish
  from main belt objects.

• An alternative approach is to look for objects at
  small solar elongation. These objects are dimmed
  by phase effects, but could be entirely inside
  Earths orbit, and would be missed by the
  strategy of only observing objects in opposition
  to the Sun.

44
Observing Strategies

• One can fix the telescope and let the sky
  drift over the CCD detector, moving the charge
  along at the same rate as the sky. By moving the
  telescope and scanning the same piece of sky
  again one detects moving targets.

• The alternative to this drift scanning approach
  is to make pointed observations of a given field,
  revisiting it several times (after some interval)
  to allow asteroids to move across the field. This
  is called stare mode observing.

45
The Data Analysis Problem

• We are trying to cover as much sky with as
  possible with observations that go just faint
  enough to detect these relatively small bodies,
  so interesting objects in our images are often
  just brighter than the sky.

• How the asteroid appears on the digital images
  depends largely on its rate of motion.
• Objects moving rapidly appear as streaks, while
  those moving slowly are hard to distinguish from
  stars.

46
The Data Analysis Problem

• Shown here is an image of 1994XM1 as observed
  with the Spacewatch telescope. This objects
  motion is sufficiently large that it shows as a
  distinct streak. This object is also relatively
  bright, so that the trail is quite visible.

47
The Data Analysis Problem

• The next slide shows the Spacewatch discovery
  images of 1997XF11. Unlike the previous example,
  this asteroid is moving so slowly that in any one
  frame it appears to be star-like. It is only
  after comparing several frames of the same piece
  of sky taken at different times that we see this
  asteroid has moved relative to the background
  stars.

48
(No Transcript)
49
The Data Analysis Problem

• 1997XF11 is an object on an orbit that will bring
  it close to Earth. It was discovered near its
  aphelion and consequently it was moving slowly.
  We will discuss this object again a bit later.

• The data analysis should be done quickly, so
  newly discovered objects can be confirmed and
  their preliminary orbit refined so the object
  doesnt get lost. (The modern version of what
  happened to the discovery of the first asteroid!)

50
The Data Analysis Problem

• An alternate observing strategy developed by
  Tholen (IfA) is to take 2 images separated in
  time by a known gap. One then looks for 2 short
  streaks separated with a gap of known length.

• One can blink the images to look for the
  streaks with a gap between them. Another way to
  examine the data is to difference or ratio the
  images and look for positive and negative
  line pairs.

51
(No Transcript)
52
(No Transcript)
53
Orbit Determinations

• The initial orbit of an object in a gravitational
  field is completely determined by its initial
  position and velocity (6 vector components).
• To determine an orbit from observations, we
  therefore need at least 3 observed positions.
  Over time gravitational forces from other planets
  can change the orbit.

• The short arc problem -- when a new asteroid is
  discovered (or an old one recovered) one may get
  3 or more position measurements over a period of
  several nights. This will define an inital orbit,
  but generally much more data is needed to define
  a precise one.

54
Orbit Determinations

• The accuracy of the calculated orbit will improve
  as the object is followed over a long time.
  Sometimes it requires several years worth of data
  to produce an extremely well-defined orbit.

• For most new discoveries follow up is at best
  spotty, and the predicted orbit can be uncertain.
  Some of these get lost again, even after only a
  few nights from the initial discovery.

55
Orbit Determinations

• An interesting case is the aforementioned
  1997XF11. The initial orbit predicted a very
  close approach to Earth in 2028. Because to this
  possible encounter, this asteroid was reported
  widely in the press.

56
(No Transcript)
57
Orbit Determinations

• Even after 88 days of following the asteroid,
  the orbit still showed large uncertainties as to
  its location in 2028 and the error ellipse
  included a very close approach to Earth!

58
Orbit Determinations

• Checking back on old observations, the
  asteroid was found in images from 1990. Adding
  these data to the orbit calculation greatly
  reduced the error ellipse. It is now clear it
  will miss Earth by a wide margin.

59
SEAs

• Spacewatch has discovered population of very
  small (10 m diameter) sized objects that are
  very common in Earths neighborhood.
• These Small Earth Approachers may be the tail
  of the general NEO population or some other type
  of object altogether.

• Some analyses of the SEA population suggest that
  there is a SEA belt located very close to 1 AU.
  
• While the origin of SEAs is unclear, they are
  being continually swept up by the inner planets,
  so they must be replenished on a regular basis.

60
SEAs

• Estimates suggest that there are 100 of these
  objects within the sphere defined by the Moons
  orbit at any time. Observations by DoD and DOE
  spacecraft are consistent with this number, but
  suggest it could be a lower bound.

• SEAs disintegrate high in Earths atmosphere
  where they convert their kinetic energy into
  kiloton sized detonations. Empirical evidence
  suggests that very few of them ever make it to
  the ground.

61
(No Transcript)
62
(No Transcript)
63
SEAs

• The infrared sensors on the spacecraft are
  scanning devices and may miss as many as 90 of
  these impacts. The tail on the light curve
  suggests a high altitude cloud of hot dust. These
  might be visible in the IR observations from
  weather satellites.

64
SEAs

• Because the SEAs are so close to Earth, they move
  by quite quickly, crossing the Earth-Moon system
  in about a day. I have proposed a prototype
  system of small telescopes that could, in
  principle, detect all of these objects as they
  pass the Earth.

• Observing the same piece of sky with small
  telescopes from two locations, about 100 km
  apart, we would confirm the detection and at the
  same time be able to determine a distance to an
  SEA out to 3X the distance of the Moon.

65
(No Transcript)
66
Whats Next?

• Progress on finding all the potential NEO threats
  to Earth has been slower than the pace mandated
  by Congress. (There is alot of sky to cover!)
• There is great uncertainty as to just how many
  potential killer asteroids there are in the
  inner solar system.

• A number of scenarios for mapping the entire sky
  have been presented
• The Planetary Defense Telescope
• The Dark Matter Telescope (DMT)
• The Large Synoptic Survey Telescope (LSST)

67
Whats Next?

• Some of these proposals are really targeted
  toward science than other than NEOs. They do have
  the power to detect potentially threatening NEOs.
  
• The LSST has been ranked very high by the U.S.
  National Academy of Sciences decadal review
  committee.

• Projects designated as high priority in these
  decadal reviews generally turn out to become
  funding priorities for NSF and NASA.
• There is concern in these whole sky surveys that
  unless follow up observations are made the finds
  might get lost again.

68
Whats Next?

• A self-proclaimed front-runner for the LSST
  is the DMT which has been proposed by Roger Angel
  and Tony Tyson. This large monolithic telescope
  would have an effective collecting area of a
  6.9-m telescope and a field of view of 3 degrees.

69
Whats Next?
70
Whats Next?

• An alternative LSST design is being developed
  at the IfA. An array of smaller telescopes,
  operating together and/or independently, can do
  the same science 30 of these telescopes would
  have the same light gathering power as the DMT.

71
Whats Next?

• Whatever the design of the LSST, it will generate
  a huge volume of data. The project goal is to
  observe the entire sky each week! This will
  generate about 5 terabytes of new observations
  per week.

• The more successful we are in observing the sky
  for these time variable objects, the worse our
  data processing problem becomes, in terms of the
  requirements for raw computing speed, data
  storage, and network transport capacity.

72
What Next?

• Another priority of the NAS decadal review report
  is the National Virtual Observatory (NVO). It is
  intended to be a distributed database of
  astronomical observations.

• Given the huge volume of data that LSST will
  generate, it will certainly be one of the key
  nodes of the NVO.

73
Whats Next?

• An important aspect of the NVO will be detecting
  time variable objects, either those that move
  (asteroids, comets) or those whose brightness
  changes with time (e.g., distant supernovae)
  among the vast sea of objects that appear to be
  constant over time.

• A key challenge in the NVO project is how one
  mines the massive database to find these and
  other interesting objects. We must go beyond
  traditional user interfaces and incorporate AI
  for database self-mining.

74
Whats Next?

• As I have emphasized several times, follow up
  observations to new discoveries is critcial. This
  requires that we process the massive input data
  stream from an LSST or similar telescope, in an
  automated fashion, in near real time.

• We have already begun formulating an approach to
  deal with this onslaught of data. In the next
  talk my colleague, Herr Schloer, will tell you
  about Project Dark Expedition and how it will
  help us meet this immense challenge.

75
In Summary

• The threat from NEOs to Earth is real. An impact
  will happen in the future (unless we can detect
  the object in advance and deflect it).
• We have the technology to detect the potentially
  threatening objects. A viable Spaceguard system
  is not in place at this time.
• Data processing will be a critical component both
  in detecting threatening objects and tracking
  them over time.
• The volume of observations will be more than
  people can easily analzye!

76
(No Transcript)
77
(No Transcript)