SAN FRANCISCO — For more than a decade, satellites have provided vivid images of hillside communities and roads destroyed by landslides. Increasingly, researchers are using space-based sensors to pinpoint the causes of landslides and determine where they will strike.

“The goal is not to identify whether a landslide will happen on a single hill at a given time, but to try to provide a general indication of the chance of a landslide occurring,” said Dalia Kirschbaum, research scientist in the Hydrological Sciences Laboratory at NASA’s Goddard Space Flight Center in Greenbelt, Md.

NASA Goddard researchers are modeling landslide activity using precipitation data drawn from the Tropical Rainfall Measurement Mission (TRMM), a joint NASA Japan Aerospace Exploration Agency mission launched in 1997, and observations of land cover provided by the Advanced Very High Resolution Radiometer onboard the U.S. National Oceanic and Atmospheric Administration’s Polar Operational Environmental Satellites and the Moderate Resolution Imaging Spectroradiometer on NASA’s Terra and Aqua Earth-observing satellites. That information is used in conjunction with detailed elevation data drawn from NASA’s 2000 Space Shuttle Radar Topography Mission to identify factors that contribute to rainfall-triggered landslides.

Eventually, these models could be used to forecast imminent landslides. Because there is currently no global warning or operational monitoring system for landslide conditions, the Goddard research group is interested in developing “robust models to alert governments and aide organizations to the likelihood of landslides so they can respond more quickly to these events,” Kirschbaum said.

Goddard researchers have focused their attention on rainfall-triggered landslides occurring worldwide since 2007. During that period, rainfall-triggered landslides killed more than 14,000 people, Kirschbaum said. With the expanding global population and an increasing number of buildings and roads constructed on unstable slopes, landslides are likely to become more frequent and to cause more damage, she added.

At the same time, landslide models are expected to improve as they begin incorporating data from new space-based instruments, including TRMM’s successor, the joint U.S.-Japan Global Precipitation Measurement (GPM) mission and NASA’s Soil Moisture Active Passive mission, which are both scheduled to launch in 2014. GPM is expected to improve rainfall measurement dramatically because its primary satellite, known as GPM Core, will operate in a constellation of satellites to produce a global picture of rainfall every three hours at a resolution of approximately 10 kilometers. That constellation includes the NASA and the National Oceanic and Atmospheric Administration’s Suomi National Polar-orbiting Partnership satellite, the NOAA-19 weather satellite, the first spacecraft in NOAA’s Joint Polar Satellite System, the U.S. military’s Defense Meteorological Satellite Program and Japan’s Global Change Observation Mission. In contrast, TRMM offers 25-kilometer resolution, Kirschbaum said.

In addition, GPM is designed to measure rainfall over a much larger area of the Earth’s surface than TRMM. TRMM’s precipitation radar and passive microwave sensor gather data on Earth’s tropical and subtropical regions, from latitudes of 35 degrees north to 35 degrees south. GPM will gather data from 65 degrees north to 65 degrees south, an area that extends from the Antarctic Circle to the Arctic Circle, said Kirschbaum, an application scientist and outreach coordinator for the GPM mission.

Engineers at Draper Laboratories also are developing statistical models to pinpoint the complex causes of landslides. Since 2011, Draper engineers working under a NASA contract, have gathered data from ground-based, airborne and satellite sensors on the precipitation patterns, soil types, soil structures and weather conditions in areas where landslides occur. “The idea is not only to provide landslide forecasting, but also to develop models that show which factors contribute to landslides,” said Natasha Markuzon, lead technical investigator for the landslide research project at Cambridge, Mass.-based Draper. “Sometimes one storm produces a lot of landslides and a very similar storm produces no landslides at all.”

That understanding will be particularly useful as the global climate changes, Draper officials said. Researchers at the Massachusetts Institute of Technology working with Draper on the NASA project are studying whether changing climate patterns threaten to produce more frequent landslides in certain areas. By using climate change projections, researchers will evaluate the risk of landslides occurring in the next 10 to 20 years in areas that have not previously experienced them, said John West, Draper program manager.

As scientists gain greater insight into the multiple factors that contribute to produce landslides, they also may be able to identify specific actions communities could take to prevent landslides. Researchers might determine, for example, that certain types of ground cover or changing road construction practices would help to prevent landslides or diminish their impact, West said.

In conjunction with the Draper program, MIT students are exploring the social and economic impact of landslides. Because landslides tend to be small events that occur more frequently than hurricanes and floods, they receive less attention, Kirschbaum said. Nevertheless, their consequences in terms of lives lost and property damage can be dramatic. In August 2010, a landslide in China’s Gansu province killed 1,700 people and caused an estimated $760 million in property damage, Kirschbaum added.