The development could also prove invaluable for climate models that predict the world's evolving environment, said researchers at the University of Pennsylvania.

"To predict the future of the ice sheet and to understand the past, we have to put the information into a computer," said Richard Alley, professor of geosciences and co-author of the study to be published in the journal Science.

"The models we have do not currently have any way to figure out where the big ice sheets end and where the ice calves off to form icebergs."

Ice sheets in Antarctica and Greenland can spread out for vast distances across the ocean before their outer edges calve off to create icebergs.

The massive Ross Ice Shelf in Antarctica, for example, which is roughly the size of France, can floats for as much as 500 miles (800 kilometers) over the ocean, while other sheets float for only one or two miles (kilometers), before their outer tips break up into icebergs, said researchers.

"The problem of when things break is a really hard problem because there is so much variability," said Alley.

"Anyone who has dropped a coffee cup knows this. Sometimes the coffee cup breaks and sometimes it bounces."

The most important variable for knowing when icebergs will be created is how fast the parent shelf is spreading, said the study.

If it expands slowly, the ice cracks consistently throughout the shelf -- but if it spreads quicker, there is more chance chunks will snap off.

Like a coffee cup falling, the event of a shelf breaking can depend on "what it hits and where it hits, but the most important variable is the distance the cup falls or is thrown," researchers said.

Using the variables of the shelf's expanding speed and its dimensions, researchers succeeded in establishing the formula that will help scientists understand the process.

The research team, however, emphasize their formula may not capture the "totality of variation in the ice calving process but does account for a large percentage of the variability."

The study is published in the November 28 edition of the US journal Science.

All rights reserved. copyright 2018 Agence France-Presse. Sections of the information displayed on this page (dispatches, photographs, logos) are protected by intellectual property rights owned by Agence France-Presse. As a consequence, you may not copy, reproduce, modify, transmit, publish, display or in any way commercially exploit any of the content of this section without the prior written consent of Agence France-Presse.