In this issue:

Credit ratings, says Josh Lauer, are basically a way to quantify trust. Banks and other lenders giving out loans—subprime mortgages aside—want to be reassured that the money will be repaid. And given the importance of credit to modern economies, it's essential that lenders have a way to gauge the risk of their investments.

Lauer, an assistant professor of communication, has been studying the development of credit reporting in the United States. Two hundred years ago, when prospective borrowers were usually business owners, lenders tended to know them personally. But as lending became more impersonal, credit-reporting agencies began to spring up in the mid-19th century. Their methods would seem primitive today, Lauer says. Rows of ledgers detailed business owners' personal habits and prospects, based for the most part on reports written by local correspondents. Often the reports were little more than hearsay.

Soon, credit reporting expanded from business credit to consumer credit. From the start there was a close link between a person's behavior and their credit rating. If someone was said to spend money gambling, drinking or philandering, for example, that information was considered relevant to their risk as a borrower.

That connection is still made today. Credit scoring is now done with computer models and sophisticated algorithms, Lauer says, "but underneath all of that, somebody is using this information to try to figure out how honest you are."

But while methods of rating credit have improved, the risk of abuse is increasing as well. One problem, Lauer says, is the use of credit scores for unrelated purposes. He gives as an example employers who check the credit ratings of job applicants. Employers are assuming that a person's financial history reveals something about their ability to perform their duties, he says. But there could be many reasons for a less-than-stellar credit score, such as unexpected medical expenses or being laid off.

While mathematical models for portfolio risk analysis have been criticized lately, the computer models used for personal credit reporting have both advantages and disadvantages, says Lauer. Ratings have become more democratic—when a lender knows a borrower, personal prejudices can sway the decision, he notes. Prejudice was at the root of redlining, which occurred when banks refused to issue mortgages in minority neighborhoods. Redlining was made illegal in the '60s and '70s, and now, because a credit report is based on numbers produced by a computer, it is less likely that factors such as race or ethnicity will play a role, as long as the program is written without bias.

Still, by relying on computer models, an important personal dimension is lost. Where once a credit manager might be willing to accommodate a customer who got behind on bills due to a temporary job loss, for example, automated credit scoring makes no distinction. This is a dilemma now faced by thousands of hardworking Americans seeking relief in a deteriorating economy. "To lose your job or home is bad enough," Lauer says, "but a poor credit record adds the stigma of being labeled too risky or unreliable to trust. Your personal story becomes irrelevant."

Nothing ruins an otherwise perfect day at the beach like rotting shellfish and seaweed. Unfortunately, noisome odors are not the worst side effect of an invasive plant living along the coastline of the northeastern United States.

The invader is a subspecies of the common seaweed Codium fragile, and is native to the eastern Pacific Ocean. In the late 1950s it made its way to Long Island, possibly by hitching a ride on the hull of a ship. Today, it's thriving as far north as Prince Edward Island, and leaving both economic and ecological damage in its wake. "It's a menace," says Anita Klein, an associate professor of biological sciences who is studying the plant.

In its native habitat, the seaweed's population is controlled by natural predators, but local fish and and invertebrates don't seem to like its taste. As a result, the seaweed is gradually replacing kelps native to the northwestern Atlantic, changing the habitat of the area.

Shellfish populations have been particularly affected. Codium attaches to hard surfaces such as rocks or the carapace of shellfish. When waters rise—during a storm, for example—shellfish can get carried ashore along with Codium, which is how it earned its nickname "oyster thief."

Even worse, says Kelly Cullen, an associate professor of natural resources and the environment, is the impact on the shellfish industry. "This invasive species could devastate the entire industry," she says. It's not just those who catch shellfish for a living who could suffer: "The ripple effect could be significant." She's working to estimate the damage.

Klein and Arthur Mathieson, a professor of biological sciences, are trying to figure out how the oyster thief has been able to proliferate, and they hope to help prevent it from spreading any further.

One important question, says Klein, is whether Codium has been introduced to the northwestern Atlantic repeatedly or just once. By examining DNA at several locations in the seaweed's genome, Klein will be able to observe the amount of genetic variation. If samples from different areas along the East Coast show a lot of variation, that would mean that there must have been multiple introductions. "We don't think it's been around long enough to become genetically variable at the particular genes we're looking at," she explains.

Another explanation for the seaweed's move north is climate change. It had been thought, based on the plant's natural habitat, that it would be unable to live as far north as it does. It's possible, Klein says, that ocean waters have warmed. Or, as with the first arrival of Codium, the explanation could be that it tagged along with humans, attached either to their ships or to shellfish used in aquaculture.

It will be hard to exterminate the seaweed where it is already flourishing, but more can be done to prevent it from spreading, such as screening boats more carefully. Such efforts can be difficult, but, as beachgoers have discovered, the alternative stinks.

Graduate student Ram Ray has witnessed firsthand the damage that landslides can cause. In Nepal, where Ray grew up, more than 100 people die in a typical year as a result of landslides, not to mention the destruction of roads and buildings. His childhood experiences are one reason that he is now working with Jennifer Jacobs, an associate professor of civil engineering, to find ways to predict these natural disasters.

Sometimes landslides are triggered by earthquakes, but many slides occur when the soil on a slope becomes saturated with water. Water also reduces the shear strength of the soil, Ray adds, undermining the soil's ability to hold together. When a downpour hits a slope already inundated with water, tons of dirt and debris can end up burying the area below.

To determine the likelihood of a slide, Jacobs and Ray study three key variables: the steepness of the slope, and the amount of soil moisture and precipitation. For the past few years, they have used NASA satellite data to analyze these factors in landslide-prone areas of Nepal, the Philippines and California. These remote sensing tools, Jacobs says, can help them predict when "conditions are ripe for a hazard."

Using the data, Jacobs and Ray developed a slope stability model. Given a certain level of moisture on a particular grade, the model reveals how much rain it would take to make a landslide highly likely. They checked their model by comparing its calculations with actual landslides. In the Philippines, for example, a landslide in 2006 devastated the village of Guinsaugon, killing more than 1,000 people. Jacobs and Ray found that weeks of intense rain in February had nearly saturated the soil. Another week of heavy rain in the same month set off the slide.

Eventually, they would like to be able to use this model to predict future landslides. Jacobs cautions the model could not determine with certainty that a landslide will occur, just as weather forecasts can only show the probable path that a hurricane will take.

"Right now with landslides, everything is examined in retrospect," Jacobs says. By taking the model's prediction into consideration, governments and aid agencies would be able to encourage evacuations and have rescue operations ready before disaster strikes. The data could also be used in planning construction of buildings and roads. If a road is likely to require frequent construction to repair landslide damage, Ray says, it might make more sense to reroute the road.

The next step will be to gather soil samples from some of these areas. By comparing the soil moisture of these samples to the satellite data, Jacobs and Ray will be able to further verify the accuracy of their model and make any necessary adjustments. With a little more work, people living in Nepal and other mountainous regions may soon have an early warning system for landslides.