There’s no mistaking it now. Even though we don’t have the final numbers, it is abundantly clear that the sea ice ringing the Antarctic continent has fallen precipitously — reaching a record low just a few short years after it reached a record high.

In 38 years of records dating back to 1979, the sea ice lows seen as of the end of February 2017 — a time of year when ice in the Antarctic is at its annual minimum — are unprecedented. The area of ocean covered by sea ice still appears to be shrinking, but as of Feb. 28, there were just 2.131 million square kilometers of floating ice surrounding Antarctica, according to near-real time data provided by the National Snow and Ice Data Center.

That’s much less than the prior low of 2.29 million square kilometers on Feb. 27, 1997. The difference — about 159,000 square kilometers, or 61,390 square miles — amounts to an area nearly as large as Florida.

The data here are reported as a five-day average and should not be considered final — there could still be adjustments. And the ice could go lower before it rebounds as colder temperatures begin to return to the Antarctic. Still, the margin is large enough that a record is unavoidable, says Mark Serreze, director of the National Snow and Ice Data Center, although he said his group will wait to call one formally until they’re sure an annual minimum has been reached and the ice is growing again.

“It’s going to be a record low minimum no matter what happens right now, it’s just a matter of, how low do we go,” Serreze said Monday. “It could be any day now.” (The ice is even lower now than it was when we spoke.)

Far more difficult than pinpointing the actual ice extent is the question of why this is happening. The simple fact that the previous record low was in 1997, 20 years ago, “really goes to show you how variable the system is down there,” Serreze said.

Indeed, the overall Antarctic sea ice trend, bucking climate change expectations, has been a slight increase over time, rather than a shrinkage.

All of that said, 2017 is a year that looks like what we ought to expect on a warming planet based on climate models, said Cecilia Bitz, an expert on sea ice at the University of Washington in Seattle. The confusing thing, she continued, is that other recent years have not matched the expectations created by those models at all. As mentioned earlier, before this most recent crash, Antarctic sea ice reached a record high in October 2014.

This is one reason that, until recently, climate change skeptics loved to point to Antarctic sea ice behavior to justify their rejection of mainstream climate science — something I highlighted two weeks ago when the current Antarctic record was already beginning to emerge.

The story for Antarctic ice this year, Bitz said, can’t be understood unless we look back to September and October, when ice levels were relatively normal but suddenly plunged. It was the loss then that exposed the ocean to 24-hour sunlight in the austral summer, helped it warm up and ultimately set in motion the record we’re now seeing.

“We went from really typical conditions to record low in a two-month period, so I think that’s when we should be focusing,” Bitz said of the months in late 2016.

“I think it’s a mystery,” Bitz said.

One leading idea to explain what’s happening with Antarctic sea ice, advanced in a paper by Bitz and Gerald Meehl of the National Center for Atmospheric Research (and several colleagues), targets changes in the Pacific Ocean to explain the unexpected expansion of Antarctic sea ice between 2000 and 2014 — a trend that ran contrary to climate model projections.

This study suggests the expansion was a fluke of natural variability in the world’s largest ocean, which changed the atmosphere in ways that affected winds all the way down in the Antarctica (if winds blow floating sea ice outward from the continent, more ice can then fill the empty space, and Antarctic ice as a whole expands).

But Meehl has cautioned that the cycle in question, the Interdecadal Pacific Oscillation, plays out over decades, making it hard to be sure yet whether a turn in this cycle will correspond to Antarctic ice shrinkage. Bitz, too, said she is cautious for now about seizing on this explanation, noting that the sharp loss in Antarctic ice lagged behind a positive peak in the IPO by six to nine months, raising doubts about the connection.

But you can bet that scientists are going to spend plenty of time trying to take this question apart — and that the analysis has just begun.