# Pitfalls Iii Predicting A Temth Planet

by Dr. E. Myles Standish, Jr., Jet Propulsion Laboratory

The residuals of Uranus and Neptune do not demand the existence of a tenth planet, for there are other explanations that are at least as plausible. It is easy to ignore the fact that the residuals themselves need reexamination.

Do the residuals represent true deviations in the motion of the planets? Or can the residuals be more simply explained as errors and inaccuracies in the procedures of working with the observational measurements?

There is a succession of steps involved in determining the residuals for a planet:

1. Begin with the raw measurements. Classical planetary position observations are made by measuring the altitudes of both the planets and the stars and timing their transits across the meridian (the north-south line in the sky). This step can never be repeated for a specific observation. One cannot go back and reobserve the position of Mars in 1850.

2. Derive the observed positions, 0. Using the raw measurements and the most accurate catalog of star positions available derive the observed positions of the planets. These derived "observables" have been published and, more recently, have been put into computer-readable form Many of the raw measurements have also been published, while others still exist only in the original observing notebooks These raw measurements have not, how ever, been put into computer-readable form—a task that would be colossal.

3. Calculate the computed positions, C. Use an existing orbit (ephemeris) to predict positions of the planet for comparison with the observed positions.

4. Form the residuals, O-C. The residuals are the differences between the observed or bital positions, O, and the computed orbital positions, C. In the ideal world, with

0f 0.02 seconds of arc—about 20 times better than standard posi-^ gl measurements by telescopic photography in the optical wavelengths.

pushing beyond that accuracy are teams of astronomers at the Cali-■ institute of Technology's Jet Propulsion Laboratory and NASA's rd Space Flight Center. They are using very widely separated radio

I scopes to determine the positions of quasars to an accuracy of 0.001 second. Quasars are the most distant objects in the universe and therefore are the closest thing the cosmos offers to a fixed background of objects that do not appear to move up, down, or sideways. By relating cecraft ancj planets to the reference background of quasars, navigation of space probes and the predicted positions of planets should be improved by a further factor of 20.

Thinking back over his discovery of Pluto and his 14 years of planet searching, Clyde Tombaugh offered Ten Special Commandments for a Wbuld-Be Planet Hunter. The Final commandment decrees: "Thou shalt not engage in any dissipation, that thy years may be many, for thou shalt need them to finish the job."19

perfect observations and with perfect ephemerides, the residuals would be zero.

5. Adjust the computed orbit of the planet, C, to fit the observed positions, O, as nearly as possible, so as to minimize the residuals. (This adjustment is usually accomplished using the mathematical method of least squares, developed by Gauss.)

6. Examine the final residuals. What keeps them from being zero? Perhaps the catalog of star positions still has distortions. Perhaps the computed orbit can be further improved. When no more improvements are possible, one goes on to step 7.

7. Other explanations of the residuals. What can cause nonzero residuals to a best-fit orbital calculation? A defect in the law of gravity? Planet X?

Star catalogs produced since 1910 have been shown to contain significant distortions. Certainly earlier and cruder catalogs must also contain distortions of similar, if not greater, magnitude. However, ■>o one has ever gone back to step 2, properly reprocessing the raw measurements of the planets using a modern star catalog.

For the post-1910 data, a lesser alternative has been used: updating the published positions using general differences found by comparing the original catalog with a modern one. For data previous to 1910, not even this lesser alternative has been applied. The original catalog distortions remain embedded in the residuals.

At best, a few tenth-planet seekers start at step 5, cranking out new orbits in an attempt to fit the observations. Far worse, most begin with step 7: "May 1 get a copy of your residuals?"

The complete job must start back at step 2—taking the original measurements of the planets and recalculating their positions using modern star catalogs. Such an analysis would bring no funding or headlines. It would be time-consuming anc tedious. But until the basic observations (from which come the residuals) are processed properly there is no necessity to invoke Planet X.