Earth accreted from star dust and planetesimals in the youthful solar system some 4.6 billion years ago. The number is based on dating by radioactive elements and their daughters, notably the heavy elements uranium and thorium and the elements arising from their decay, mainly different types ("isotopes") of lead. The possibility of dating the age of Earth using these methods was realized in the early 20th century; but good numbers only were obtained by the 1950s.
The fact that Earth's mantle is engaged in convection and that the planet's surface is tectonically active (as we have seen) greatly complicates an assessment of its origin and early history. Measures must be found that are independent of the age of rocks (the oldest of which are less than 4 billion years old). After many heroic attempts by a number of scientists, the American geochemist Claire Patterson (at Caltech) came up with the value of 4.6 billion years, based on the analysis of lead isotopes- daughters of decay of different types of uranium.
This number was the precious result of much effort and much clever thinking. Before the contents of the products of radioactive decay in Earth's rocks could be evaluated in terms of age, one had to account for the initial conditions, set by the rock and dust particles falling into Earth during the time of accretion Patterson was able to do so by analyzing meteorites and back-calculating the ratio of lead isotopes for the time when Earth originated. Assuming that present-day meteorites are a good sample of primordial matter in the solar system, the correction could be made and the age of 4.6 billion years emerged. Subsequent studies confirmed this value.
Impact craters cover the surface of the Moon. They provide a record of the violence in the early solar system. Because the Moon is "dead" in the sense that it has no geologic activity, it provides a pristine history book.
The oldest rocks found on the surface of Earth are some 700 million years younger than the apparent age of the Earth, as seen in the lead isotopes. Why this discrepancy?
We must assume that the birth of Earth was a rather violent affair. Looking up at the Moon with a pair of good binoculars (or studying the images obtained by the Apollo program) we see innumerable craters of various sizes The craters are witnesses of a period in the solar system when rocks of various sizes, from meteorites to asteroids, slammed into the Moon. Mercury has quite a similar face (although harder to see from here). Apparently, the bombardment ended some 3.8 billion years ago - so what we see on the Moon are the ancient memories of impact events. If Moon was bombarded, so was the Earth, which makes a much bigger and better target, with its much greater gravity.
The heavy bombardment that our growing planet experienced during its first several hundred million years of existence would have wiped out much of the earliest record of recognizable rock. Anything that might have formed at or near the surface would have been soon re-heated and re-worked by impacts. Furthermore, after the period of heavy bombardment, convection would have obliterated much that was still left as a witness to that earliest history. Only by getting incorporated into continental crust, safe from subduction into the mantle, could any record be preserved. There was as yet not much crust to serve as a Noah's Ark for endangered rocks. Building continents had just begun.
An interesting conundrum arises in this context, as pointed out by the Swedish-American geochemist Gustaf Arrhenius (born 1922) working at Scripps Institution of Oceanography. It is that some of the earliest rocks (3.8 billion years old) found on Earth show beautifully regular stratification, suggesting quiet water. How is this possible if bombardment was still proceeding on the Moon 3.8 billion years ago? Perhaps bombardment ceased rather suddenly, or perhaps the Moon was someplace else, and not so close to the Earth or perhaps the Late Heavy Bombardment ceased earlier.