July 13th, 2010
Flights were suspended, ferries cancelled, roads shut, hospital emergency departments overwhelmed with calls and people urged to stay indoors. A terrorist attack? No, a 2009 dust storm in Sydney, Australia. Dust storms at their worst can bring disease and disruption, which is why it’s useful to know if one’s about to start. A team of scientists from the University of Pittsburgh think they’ve developed a new method of doing just that. In fact, Dr Stephen Scheidt and colleagues were able to predict a dust storm two days before it struck.
For a dust storm to start, you need lots of sand or dust that’s dry enough to be picked up by the wind. But it’s expensive and difficult to keep track of dust wetness over large areas by sticking instruments into the soil or using rainfall gauges. This is especially the case in deserts, which are inhospitable places for scientists to work. So, instead, researchers use satellite instruments to take regular snapshots of the condition of dust and vegetation. But, traditionally, they’ve used instruments that produce quite low-resolution digital images. Each pixel on images produced by a satellite called the Advanced Microwave Scanning Radiometer (AMSR-E), for example, covers 25km of the Earth’s surface. Compare this to the iPhone 4, which has 326 pixels per inch!
These big pixels were the problem facing Dr Stephen Scheidt and colleagues, who were trying to predict dust storms coming off the White Sands Dune Field in southern New Mexico. White Sands is the world’s largest area of gypsum – the bright white mineral used in plasterboard – sand dunes. Gypsum doesn’t normally form sand because it’s soluble in water, but it gets trapped in White Sands area because of its unusual geology. The white sands cover the floor of an enormous bowl-shaped hollow – the Tularosa Basin, which is ringed by the San Andres and Sacramento mountains. Water dissolves gypsum out of the mountain rocks and runs into the basin, but the basin has no river running from it toward the sea. So the water gets stuck and eventually evaporates away, leaving the gypsum behind.
When the basin is very dry, it’s prone to dust storms that can travel for more than 200km and inundate the nearby city of Alamogordo. Unfortunately, the dry river basins and dunes are too small to be monitored by instruments like AMSR-E. So the University of Pittsburgh team used a different satellite instrument – the Advanced Spaceborne Thermal Emission and Reflection (ASTER) radiometer, which has pixels between 30 and 90m across. They compared images collected by ASTER with those from an instrument with bigger pixels – the Moderate Resolution Imaging Spectroradiometer (MODIS), which collects data more frequently. ASTER sweeps over White Sands every 16 days, but MODIS passes over the area every 12 hours.
The team used the ASTER and MODIS data to calculate the wetness of the sand and the wind speed needed to carry the sand away. From this, they could predict when the gypsum was susceptible to a dust storm. However, ASTER and MODIS don’t actually measure soil moisture directly. They work a little like cameras but, instead of detecting light, they detect infrared radiation – the electromagnetic radiation given off by hot objects like people’s bodies. The infrared radiation given off by sand can used to work out if it’s wet or dry because wet sand is harder to heat than dry sand. Once the team knew if the sand was wet or not, they could calculate how high the wind speed would need to be to move the sand.
They tested whether this worked using images collected by ASTER and MODIS on seven different dates between 2000 and 2008. And, this is where it gets interesting, the driest image was from March 10th, 2008. Four days later – there was a big dust storm. The team have still got some issues to iron out, for example, testing their results against measurements made on the ground and refining their equations, but they think this approach might be useful for monitoring dust storms in deserts elsewhere. In fact, they hope it could be part of a global dust storm monitoring system.
Categories: News from outside