Earth has had a tumultuous past, so it's hard to get a clear idea of those wild early days. But a new study by researchers from the University of Adelaide raises the possibility that continents may have risen out of the sea much earlier than is currently believed, before being destroyed once again by tectonic activity.

It's generally thought that the bulk of the continental crust – the granite-like rocks we're all walking on every day – mostly formed about 2.5 to 3 billion years ago, with some coming together both before and after that window.

But the new study suggests that these rocky landmasses were preceded by much older continents that were later destroyed by tectonic movements or melted down. These older rocks could date as far back as 4 billion years, which isn't too long (geologically speaking) after the birth of the planet, around 4.5 billion years ago.

The team came to this conclusion after analyzing 75,800 samples of igneous rocks, collected from different continents. The researchers then used the estimated ages and current radioactivity of these rocks, and constructed a model of their average radioactivity from the past 4 billion years.

"All rocks contain natural radioactivity that produces heat and raises temperatures in the crust when it decays – the more radioactive a rock the more heat it produces," says Derrick Hasterok, an author of the study. "Rocks typically associated with the continental crust have higher radioactivity than oceanic rocks. A rock 4 billion years old would have about four times as much radioactivity when it was created compared with today."

But intriguingly, the team found that rocks older than about 2 billion years were found to have lower levels of radioactivity than expected. The model was altered to account for the higher heat production that would have occurred earlier in the Earth's violent past, and found that this statistical anomaly disappeared. That tells the researchers that continents may have arisen earlier than is conventionally thought – but these early landmasses would have been broken down and reshuffled, before arising again in a more accepted timeline.

"We think there would have been more granite-like – or continental-type – rocks around but because of the higher radioactivity, and therefore higher heat, they either melted or were easily destroyed by tectonic movement," explains Hasterok. "That's why these continental crusts don't show in the geological record. Our prevailing models suggest that continents eventually grew out of the oceans as the crust thickened. But we think there may have been significant amount of, albeit very unstable, continental crust much earlier."

Of course, a claim this huge requires more evidence to back it up. But still, it goes to show just how little we know about the early days of our home planet.

"We use this model to understand the evolving processes from early Earth to the present, and suggest that the survival of the early crust was dependent on the amount of radioactivity in the rocks – not random chance," says Hasterok. "If our model proves to be correct, it may require revision to many aspects of our understanding of the Earth's chemical and physical evolution, including the rate of growth of the continents and possibly even the onset of plate tectonics."

The research was published in two studies, appearing in the journals Precambrian Research and Lithos.