How do we know?
Science!
We know about continental drift and where the landmasses were millions of years ago. Of course, lots of ark believers think the earth is 5000 years old.
You ark folk are bending over backwards with unprovable theories, a bunch of "perhaps" and "maybe" and other mental gymnastics because you are beginning realize that the ark story was a fable, but you still want to believe, like a child trying to rationalize Santa Claus.
"How do we know?
Science!
We know about continental drift and where the landmasses were millions of years ago."
Remember, though, that the science of paleontological history, is itself based on an assumption, taken on faith.
That assumption is uniformitarianism - that things in the past occurred at the same rate and in the same ways that they do today. To use a concrete example, look at the Ideal Gas Law. PV=nRT. That "R" in there is a constant, like "c", the speed of light constant so crucial to Einstein. The uniformitarian assumption is that R and c 500 years ago, or 10,000 years ago, or 1 billion years ago, were the SAME, and that therefore we can know a great deal about the past by studying the traces the past has left us in the present - fossils and rocks and the like - and applying our constants and observations about the rate things happen TODAY, simply extrapolating that backwards as far as is necessary to produce the thing seen.
All efforts at chronological history in natural science repose squarely on the uniformitarian assumption, but that assumption is not itself science. It cannot be falsified. We have no time machine. We have absolutely no way at all to look back into the past, particularly the pre-historical or pre-human past, to see whether or not things occurred at the same rate as now...that the constants of today had the same value back then. We don't know, and we can't know. So we assume it. We HAVE to assume it, because otherwise the past is opaque to us. What else do we have to go on?
It is the assumption that allows us to then proceed scientifically. But the assumption itself is not science at all. It's faith. It's like a postulate in geometry. Once you accept the postulates, geometry flows logically and systematically. But the postulates themselves are unprovable, and unchallengeable. They are simply ASSERTED and ACCEPTED, and based on those assertions and acceptations, geometry is then possible.
The same is true of the science of the past. It IS science (at least if it is properly conducted it is), but it is not based on science. Science is, above all, empirical observation. We CANNOT empirically observe the past on planet earth. We can look at rocks and sediment layers and fossils, and rates of radioactive decay, and we can ASSUME that decay rates are constant over time in order to date things. We can do lots of science based on that uniformitarian assumption. But the assumption itself is a postulate, an assertion. IT is not science. It is faith.
This element doesn't exist in modern empirical science studying the here and now. But it becomes more and more important the farther back one moves in time. If the constants themselves have altered, perhaps through a process of punctuated equilibrium, we don't know and CAN'T know.
So, our scientific descriptions of the pre-Cambrian world are logical and scientific, but they do NOT ultimately REPOSE on logic. They repose on faith, in uniformitarianism, which is a necessary postulate, but an unprovable one (at least until somebody builds a time machine).
So, yes, we think we know where the continents were a billion years ago, and 10,000 years ago. And if our uniformitarian assumption is true, we are probably very close. But if the constants have changed (and there is astronomical evidence of some variation in the values of constants), then we don't know at all, because the tool we have used as our yardstick - the behavior of things NOW, TODAY - would fail us utterly.
We can posit the findings of our paleontology and call it science, and be reasonably confident in it. But we must not become so overmighty in our belief that we are operating from PURE science and PURE logic that our position is unassailable. All of our science of the elder days reposes on the uniformitarian assumption. It's a good assumption (what else could we possibly use) and a necessary assumption (if we don't assume it, we can't do paleontology at all). But it's not empirical science, and it's not itself based on rational deduction (we don't have a particularly good reason to believe that the constants have been constant forever; and we do have some astronomical data that some of them may not be).
It's faith.
We shouldn't forget that, because the knowledge of that truth ought to make us a little less proud when dealing with others who take their view of the path on faith.