Asteroid (29075) 1950 DA was discovered on 23 February 1950. It was observed for 17 days and then faded from view for half a century. Then, an object discovered on 31 December 2000 was recognized as being the long-lost 1950 DA. (Note this was New Century’s Eve and exactly 200 years to the night after the discovery of the first asteroid, Ceres.)
Radar observations were made at Goldstone and Arecibo on 3-7 March 2001, during the asteroid’s 7.8 million km approach to the Earth (a distance 21 times larger than that separating the Earth and Moon). Radar echoes revealed a slightly asymmetrical spheroid with a mean diameter of 1.1 km. Optical observations showed the asteroid rotated once every 2.1 hours, the second fastest spin rate ever observed for an asteroid its size.
When high-precision radar meaurements were included in a new orbit solution, a potentially very close approach to the Earth on March 16, 2880 was discovered to exist. Analysis performed by Giorgini et al. and reported in the April 5, 2002 edition of the journal Science (“Asteroid 1950 DA’s Encounter With Earth in 2880: Physical Limits of Collision Probability Prediction”) determined the impact probability as being at most 1 in 300 and probably even more remote, based on what is known about the asteroid so far. At its greatest, this could represent a risk 50% greater than that of the average background hazard due to all other asteroids from the present era through 2880, as defined by the Palermo Technical Scale (PTS value = +0.17). 1950 DA is the only known asteroid whose hazard could be above the background level.
However, these are maximum values. The study indicates the collision probability for 1950 DA is best described as being in the range 0 to 0.33%. The upper limit could increase or decrease as we learn more about the asteroid in the years ahead.
Expressing the risk as an interval is necessary because not enough is known about the physical properties of the asteroid. For example, radar data suggests two possible directions for the asteroid’s spin pole. If one pole is correct, solar radiation acceleration could significantly cancel thermal emission acceleration. Collision probability would then be close to the maximum 0.33%. If the spin pole is instead near the other possible solution, there would be little chance of collision. There are other factors also.
The situation is similar to knowing you have a coin that is biased so one side will land up 80% of the time – but you don’t know which side. You can only say that when you flip the coin, the chance of heads is 80% or 20%.
Whether or not the impact hazard of 1950 DA is excluded at some later date, results of the case have significance beyond the impact issue:
Physical knowledge of asteroids is required for long-term predictions, especially for objects gravitationally encountering planets. Regardless of how accurate the position and velocity measurements of an asteroid, it’s properties and environment affect the trajectory.
Asteroid deflection can be made easy and low-tech by modifying the surface properties of asteroids, given enough warning time. The required warning time for the method may vary from years to centuries, depending on the gravitational encounters along the way, which can amplify the effect.
Repetitive patterns of gravitational interactions (called “resonances”) can help preserve our ability to predict orbits into the future by constraining the growth of orbit statistical uncertainties.
Radar measurements allow us to predict trajectories 5-10 times further into the future than with optical telescopes only,
The paper explored the physical factors limiting such long-term predictions. It was found the most significant factor affecting its future long-term motion was the way heat radiates off the asteroid into space. Others factors discussed in the paper include: solar radiation pressure, uncertainties in the masses of the planets, gravitational tugging by thousands of other asteroids, the shape of the Sun, galactic tides due to other stars, solar particle wind and computer hardware imprecision.
The case of 1950 DA differs from previous hazard predictions. For past cases, a risk was detected based on a few days or weeks of data for a newly discovered object.
The uncertainty region that surrounds an object then is large, sometimes spanning a significant part of the inner solar system. Additional measurements made a few days or weeks later shrink the region such that the Earth falls out of it and the risk goes to zero.
Although other currently unknown asteroids may pose a risk before 2880, the situation with 1950 DA is unique. It is based on observations spanning 51 years, has high-precision radar data, and has a favorable orbit geometery. All these factors together allow us to predict far into the future and explore the physical limits of such collision probability predictions.
Predictions so far in the future require knowledge of the physical nature of the asteroid. How it spins in space, what it is made of, its mass, and the variations in the way it reflects light affect the way it moves though space over time. Such detailed knowledge of 1950 DA does not exist at present and may not be available for years, decades or longer.
However, because of the long-time span involved (878 years – 35 generations!), there is plenty of time to improve our knowledge. If it is eventually decided 1950 DA needs to be diverted, the hundreds of years of warning could allow a method as simple as dusting the surface of the asteroid with chalk or charcoal, or perhaps white glass beads, or sending a solar sail spacecraft that ends by collapsing its reflective sail around the asteroid. These things would change the asteroids reflectivity and allow sunlight to do the work of pushing the asteroid out of the way.
There is no reason for concern over 1950 DA. The most likely result will be that St. Patrick’s Day parades in 2880 will be a little more festive than usual as 1950 DA recedes into the distance, having passed Earth by.
The team reporting in Science about 1950 DA was led by Jon Giorgini and includes, Dr. Steven Ostro, Dr. Lance Benner, Dr. Paul Chodas, Dr. Steven Chesley, Dr. Myles Standish, Dr. Ray Jurgens, Randy Rose, Dr. Alan Chamberlin, all of JPL; Dr. Scott Hudson of Washington State University, Pullman; Dr. Michael Nolan of Arecibo Observatory; Dr. Arnold Klemola of Lick Observatory; and Dr. Jean-Luc Margot of the California Institute of Technology, Pasadena.
Arecibo Observatory is operated by the National Astronomy and Ionosphere Center at Cornell University, Ithaca, N.Y., under an agreement with the National Science Foundation. The radar observations were supported by NASA’s Office of Space Science, Washington, D.C. JPL is managed for NASA by the California Institute of Technology.
Parameter Relative Along-track Effect ----------------------------------------------- ----------------------------------- Solar particle wind 0.001 Galilean satellites -0.333 Galactic tide -0.833 Numerical integration error (128-bit vs. 64-bit) -1.000 (9900 km, 12 min) Solar mass loss +1.333 Poynting-Robertson drag -2.333 Solar oblateness [ +4.08, +1.75] Sun-barycenter relativistic shift +81.0 (inc. in nominal) 61 most perturbing "other" asteroids -144 Planetary mass uncertainty [ +132, -156] Solar radiation pressure -1092 Yarkovsky effect [+1152, -6924]
Numbers in brackets indicate a range of possible values due to poorly known physical parameters. These factors together reduce prediction window extent from 2880 to 2860 (-20 years, or -2.3%)