A scientist at the U.S. Geological Survey is questioning the soundness of an alarming study authored by a NASA scientist.

The study -- published last month in the American Geophysical Union's Earth and Space Science journal -- claims there's a 99.9 percent chance a massive earthquake hits Los Angeles within the next three years.

But Robert Graves, a USGS seismologist who assisted with some of the statistical analysis involved in the research, isn't buying the paper's conclusions.

"I have serious doubts that the conclusions of the paper are supported by the analysis that's presented there," Graves, a coordinator for earthquake hazards in Southern California, told the Los Angeles Times earlier this week.

"The 99.9 percent number -- I don't know the method that was used to derive that. But basically, that's saying that's going to happen," Graves continued. "And that level of certainty, to my knowledge, is just not attainable. We can never be that certain."

The central focus of the recent paper was not to predict the Southern California's next major earthquake, but show how new satellite data can be used to map the energy stored in the complex systems of faults in San Gabriel Valley and Chino Hills.

By studying ground deformation seen in satellite imagery before and after the 2010 5.1-magnitude La Habra earthquake, researchers calculated how much energy was released and how much energy was bottled up along newly deformed fault lines. They also attempted to determine what these changes revealed about energy stored along buried and hidden fault lines.

"The earthquake faults in this region are part of a system of faults," said lead study author Andrea Donnellan, a principal research scientist at NASA's Jet Propulsion Laboratory. "They can move together in an earthquake and produce measurable surface deformation, even during moderate magnitude earthquakes. This fault system accommodates the ongoing shortening of Earth's crust in the northern Los Angeles region."

But Donnellan and her colleagues also looked at historical earthquake data as a way to contextualize their analysis and offer a benchmark for gauging its accuracy. In looking at the 32 earthquakes of magnitude 5 or greater that have shaken the Los Angeles region over the last 81 years, Donnellan's research team calculated a 99.9 percent chance of another happening within the next three years -- all but a sure thing.

Now, other geologists and earthquake experts are questioning that analysis.

Donnellan says the probability number is an aside, and that the real conclusion is that La Habra failed to unleash a majority of deep-lying seismic strain -- the earthquake's release was, instead, dominated by shaking along shallower fault lines.

"We never said in this paper we were predicting an earthquake. And we said that's the probability of an event," Donnellan said. "There is still a 0.1 percent chance it won't happen. So we need to test it. And that's what we are doing as scientists."

But discussion of the 99.9 percent prediction have dominated headlines. Graves says it's unclear how Donnellan and her colleagues arrive at their number.

"While the earthquake forecast presented in this paper has been published in the online journal Earth and Space Sciences, it has not yet been examined by the long-established committees that evaluate earthquake forecasts and predictions made by scientists," the USGS wrote in a Facebook update.

USGS scientists are constantly monitoring the potential for earthquakes. Their analysis differs from Donnellan's.

"The accepted random chance of a M5 or greater in this area in 3 years is 85 percent, independent of the analysis in this paper," they wrote.

The legitimacy of Donnellan's prediction is something to be worked out by scientists. The real world, however, will either prove both predictions right or wrong. Southern Californians hope they're wrong.