Weather forecasters certainly have a difficult job right now, trying to predict when an unpredictable "frankenstorm" will make landfall.

They're trying to make predictions a day or two in advance. And while that's tough, the most challenging predictions about events like Hurricane Sandy might be the ones that had to happen months and months ago.

Before this storm showed up on a Doppler screen, property insurance companies were setting the prices they would charge businesses and homeowners for an insurance policy. Their actuaries went through reams and reams of historical data to figure out what damage they would be on the hook to cover.

Storms like Sandy are making the actuaries' jobs is increasingly difficult. A new report from Ceres, a non-profit that focuses on sustainability, shows a huge rise in weather-related claims paid out over the past three decades.

This makes the traditional practice of relying on historical data to predict future weather-related claims increasingly shaky. It also raises questions about how we'll pay for disasters in years to come, as property insurers aren't enthusiastic about a business where they pay out more in claims than they take in premiums.

"Over the past 30 years, the insurance sector has been paying more and more in losses related to catastrophes," says Cynthia McHale, director in the Insurance Program at Ceres, a non-profit that focuses on sustainability issues.

Actuarial forecasting is admittedly dull stuff, number crunching that happens way in the weeds. But if you're an insurer, getting a premium right is crucial to staying in business.

A health insurer has to collect enough money to cover the expensive hospital treatments accrued by a few of its members. Likewise, property insurers bank on premiums they collect from lots of customers to cover a smaller pool of property losses. 

Insurers traditionally look back at historical experience to price their premiums. That's how they justify any premium rate increases to the regulators that approve the prices they charge.

That data, however, are becoming less reliable for two reasons. First, there's the increased frequency of extreme weather events, which you can see charted here. 

Then, there's also growing population density: If more people live in a small area, and a weather event strikes, that's more businesses and homeowners likely to file property claims. 

Taken together, McHale makes the case that these factors are driving up the number of weather-related claims that property insurers pay out. AM Best, an insurance rating firm, echoes this view. In a recent report on the property insurance market, it notes that catastrophic activity is always a "wild card." Recently though, its getting wilder.

"Expectations are for continued above average storm activity, including increased frequency of non-hurricane storms," their analysts wrote.

You see that play out in this chart below, with insurers paying out some of the largest sums ever in recent years.

In 2011, the high number of claims meant that the property insurance industry operated at a loss: Property insurance carriers spent $34 billion more on paying for damage than they collected in premiums.

There are a few possible solutions here. One is having property insurers simply raise their premiums: If they take in more funds, they can cover more damage. And that's pretty much what's been happening lately. One report, back in January, estimated that premiums went up by about 10 to 20 percent in these types of policies.

The concern there is that property insurance becomes prohibitively expensive and a death spiral occurs: Only those who are nearly certain to need coverage (say, a house at sea level on the Gulf Coast) purchase a policy. Prices keep going up, enrollment keeps going down and the market eventually collapses.

Ceres advocates for insurers taking a more active role in preventing risks. They are, after all, the ones who bear the brunt of high levels of damage when they pay out claims.

"We want to see insurers think about this issue not just in terms of raising rates, or raising deductibles, but rather how can we make communities more resilient?" says McHale. "That should really be part of the dialogue. There's been a lot of work in the science of city infrastructure, about how we can make places more resilient to weather events."

Related: Is there a better way to do disaster relief?