NEW ORLEANS - Just how much oil is spewing from the ruptured well in the Gulf of Mexico and how important is it to know that? Experts can't agree on either question.
Some scientists who have studied a new video of the gusher estimate the leak could be 14 times worse than the government says. The feds say stopping the spill, not measuring it, is key.
But does the size matter now anyway?
Many veteran spill experts say trying to figure out how much oil is spewing is like trying to assess the damage of a house fire while firefighters are still trying to put it out. The size will matter later when the damage is tallied.
Others - especially lawyers, environmental advocates and the media - say numbers are needed to get a handle on how big the problem is. But a National Academy of Sciences report seven years ago said the size of a spill doesn't directly correlate to how bad the environmental damage is.
When BP, which owns the well, released a video Wednesday of the spewing oil, the whole magnitude of the problem seemed to change.
Soon after the explosion three weeks ago, the government said oil and gas were flowing from the seabed at a rate of 210,000 gallons - or 5,000 barrels - a day. Now, after viewing the video, some scientists calculate it at 2 million gallons a day or even higher.
But those estimates come with giant-sized asterisks. Because the video doesn't give experts a proper look, they could be off by millions of gallons of oil a day.
Still, most experts agree it is probably worse than the government acknowledges.
President Barack Obama addressed the issue Friday. "I know there have been varying reports over the last few days about how large the leak is," he said, "but since no one can get down there in person, we know there is a level of uncertainty."
In any case, the government says it has enough equipment to deal with a spill many times bigger than even the highest estimates.
"If you've already amassed all of the resources, boons and skimmers that you can, knowing the amount doesn't make much of a difference," said Wes Tunnell, a 35-year veteran of oil spills at Texas A&M University Corpus Christi.
That's the tack the government is taking.
"We're attacking it as if it were a much larger spill, anyway," Coast Guard Admiral Thad Allen said Friday.
Others see it differently, with a prominent House committee chairman demanding better numbers from BP.
"The amount of oil leaking is a big, big deal; it's not to be taken lightly if it really is off by the factors that we're hearing. Ten times as much is 10 times as bad - at a minimum," said Darryl Felder, a University of Louisiana at Lafayette biology professor.
The National Oceanic and Atmospheric Administration declined repeated efforts by The Associated Press to explain in detail its 210,000 gallon-a-day estimate, except to say "in the absence of direct measurements, NOAA made an estimate of flow rate based on aerial observations."
Initially, BP and the government put the leak at 42,000 gallons a day. That changed a couple of weeks ago to 210,000, which BP and government officials say they aren't changing.
"We've used satellite imagery, overflights, visual observation," said BP spokesman John Crabtree. "All the methods of calculating how much oil that's leaking are inexact."
Once the video came out, scientists looked at the flow and challenged each other to calculate how big it was.
Both Eugene Chiang, a professor of astronomy at University of California, Berkeley, and Timothy Crone of Columbia University said the spill falls somewhere between 840,000 gallons and 4.2 millions gallons a day, but couldn't pinpoint much beyond that.
"It's the same feeling you get when someone asks you how many jelly beans are in a jar," Chiang said. "I'm confident enough in it to say it is well above 5,000 barrels a day" - equal to more than 210,000 gallons daily.
Crone agreed the flow was almost certainly much higher than the government's estimate. "Whether it's twice, four times or 20 times I don't know," he said.
To make their estimates, the scientists essentially tracked particles or billows of oil across the video screen, then used the size of the pipes, particles and speed of the video to come up with a rate.
"It's an established method that's been around 25 years," said Carl Meinhart, an engineering professor at University of California Santa Barbara.
Meinhart did not make a calculation of his own, but was contacted by The Associated Press to assess the researchers' methodology. Under controlled lab conditions, the results are accurate within a couple percentage points, but that's not the case here, he said.
There are a lot of variables that can't be calculated because of the poor quality of the video and the lack of a sense of scale, he said.
Just eyeballing the video for the first time, Meinhart said he would guess a margin of error of plus or minus 20 percent on estimates.
One scientist who believes the government and BP have the right number is Paul Fischbeck at Carnegie Mellon, an engineering professor who also studies risk and regulation.
Just judging by the size of the slick - about 3,500 square miles - and the three weeks of the spill, 210,000 "looks about right as far as the slick goes," Fischbeck said.
If the higher estimates were right, there would be a much bigger slick, he said.
But some of the oil could be lingering under the water's surface. In previous cases, it's been as much as half of the spill, said University of California Berkeley engineering professor Bob Bea.
The environmental impact will be vastly different depending on the size of the spill, some experts agreed.
If the flow turns out to be closer to the higher end estimates, "that's a major catastrophe," Bea said. If it's closer to the government's estimate, it is something that can be handled.