“The Origin of Comets” on pages 301– 335 explains why rocks, dirt, and water launched by the fountains of the great deep soon merged in space to form comets. Consider what we could learn if each comet’s fairly constant—almost clocklike—orbital period around the Sun never changed. We could project each comet back beyond its earliest recorded sighting and find the date when all comets passed simultaneously near the Earth. That would provide an astronomical fix for the date of the flood.

However, planets gravitationally perturb comets,^{1} changing their periods, usually slightly, although large perturbations can happen when a comet passes very near a planet. For example, forty-five consecutive orbits of comet Halley have been computed back to 15 October 1403 B.C., when its period was 69.86 years. The change in period, from one orbit to the next, was usually less than 1.5 years, and no change was greater than 2.8 years. Can we estimate past positions of some clocklike comets accurately enough to date the flood?

Computer Simulations: A Technique That Will Not Work.

Specialized computer programs can calculate past (or future) comet positions. The more sophisticated techniques give greater accuracy but require much greater computer capabilities. All programs work essentially the same way. From a comet’s known position at a given time, all gravitational forces acting on it are calculated, especially those of the Sun and planets. The computer then calculates from where those forces moved the comet during a brief time period, such as one day. Next, the net gravitational force acting on each planet is calculated, and those planets are moved back by that same time interval. This “marching” process, one step at a time, is repeated for as long as one attempts to project the comet’s position back in time—or into the future. Similar calculations locate past and future positions of asteroids and spacecraft.

As you might expect, each step in time requires hundreds of computer multiplications. Every number stored in a computer is limited to a fixed number of digits. Let’s call that number x. Because the product of two numbers with x significant digits is a number with 2x digits, the x least significant digits are discarded when the computer stores that product. Those lost numbers become an error. Also, the slightest error in a comet’s (or planet’s) starting position affects the next computed position. These errors grow exponentially as the marching solution advances, so after about 4,000 years, the projection becomes a blur. Therefore, this technique cannot determine if comets came from near the Earth more than 4,000 years ago. Secular history for several postflood cultures is well established for the last 4,000 years, so the flood was even earlier.

A Statistical Solution: A Technique That Does Work. Chinese astronomers made The oldest recorded observation of Halley’s comet in 239 B.C. It passed perihelion (the point on an orbit closest to the Sun) at 2:49 A.M. on 25 May 239 B.C., based on Greenwich Mean Time and the Julian calendar. ^{2} From that date, Halley’s orbit has been numerically integrated (marched back in very short time increments) to 1403.80 B.C., when its period was 69.86 years.^{3}

What was its previous period? The best guess would be 69.86 years, although it could be slightly more or less. The changes in the lengths of consecutive orbital periods vary with a standard deviation, s. A small s indicates a narrow range of variations; a large s indicates a wide range of possibilities. Based on all known periods of Halley’s comet, s = 1.56 years. That is, there is almost a 2 out of 3 chance the previous period was within 1s of 69.86—between 68.30 and 71.42 years. There is a 95% chance the previous period was within 2s of 69.86—that is, between 66.74 and 72.98 years.

When was its perihelion passage 2 periods (N=2) earlier? That best guess would be 139.72 (2 × 69.86) years earlier. However, with each backward step, the total error grows, but at a diminishing rate, because of compensating errors.

If, in the distant past, the most clocklike comets all passed close enough to Earth in a particular year, we could be confident that was not a statistical fluke. Therefore, that would be the year the flood began.

Selecting the Most Clocklike Comets. An ideal clocklike comet would have:

a. long orbital periods and high angles of inclination, so the comet spends almost all its time far from the planets, minimizing their gravitational perturbations,

b. at least 2,000 years of ancient observations, so fewer orbits are needed to project it back to reasonable dates for the flood (4,000–6,000 years ago), and

c. hundreds of recorded observations that have been smoothly integrated into one marching solution.

The most authoritative source of information for all known comets is the Catalogue of Cometary Orbits 2008 (17th Edition).^{4} It lists on page 157 two comets that easily surpass all other comets in meeting these criteria:** Comet Halley and Comet Swift-Tuttle.** They are unique in having extremely long periods, high angles of inclination, and hundreds of recorded, eye-witness observations going back to 239 B.C. and 68 B.C., respectively. Furthermore, powerful computer simulations, which took into account the perturbations of all planets, large moons, and large asteroids, have accurately projected these comets farther back to 1403.80 B.C. and 702.30 B.C., respectively.^{5} If any comets are sufficiently clocklike, it will be these two. If, as we project them back in time, we find the date and time when they passed perihelion almost simultaneously, we will show, with high statistical confidence, that the material that merged to become those comets was launched from Earth a few hours earlier, when the flood began. Bingo!

Comets Halley and Swift-Tuttle were projected back to the time interval 4,000–6,000 years ago—the window of time that includes dozens of proposed, biblically-based dates for the flood. The tightest clustering occurred **in the year 3290 B.C,**^{6} after exactly 27.000 orbits for comet Halley and 20.000 orbits for comet Swift-Tuttle.

Table 24 gives each comet’s expected 1s error in arriving at 3290 B.C. Assuming the 99% confidence level is high enough to conclude that the material in both comets was launched from Earth at the same time, that single time distribution has a 1s error of **± **100 years—smaller than each comet individually.^{6}

Notice that 3290 B.C. is the most likely year of tightest clustering of only their perihelions. These comets would have been nearest Earth’s orbit a few months before or after those perihelion passes—as they approached perihelion or after they left perihelion. Those errors amount to only a few months—an insignificant error in comparison with the **± **100-year uncertainty. Therefore, the most clocklike comets were near the Earth in **3290 ± 100 B.C, when the flood began.**