PAUL GRAD

A massive earthquake hit China’s Sichuan Province on 12 May. As aftershocks and landslides threatened further loss of life and property damage, a team from the University of New South Wales joined several other organisations to provide detailed satellite images of the affected area.
The earth tremor was of magnitude 8.0 on the Richter scale. At the time of writing, the estimate of the death toll was more than 80,000.

Radar satellite surveying of ground movements was used in the quake zone to supply information to the Chinese authorities. It detailed the level of damage and the dangers from subsequent landslips and other events that could affect the rescue and recovery efforts.

One aftershock caused the collapse of 420,000 dwellings and injured many people. It takes two to three days for pressure to build up before landslides, but a purpose-built network of satellites orbiting every two days could have predicted slope failure and anticipated aftershocks by detecting even small ground movements.

A team from the UNSW’s School of Surveying and Spatial Information Systems and the Co-operative Research Centre for Spatial Information (CRCSI) used synthetic aperture radar interferometry (InSAR) to generate a ground displacement map of the quake zone, showing upheavals in the Earth’s surface of up to five metres.

The team’s first results demonstrated the performance of L-band ALOS/ PALSAR for monitoring deformation over a large area. The seismic fault was estimated at more than 300 kilometres in length and preliminary results covered only a portion of the ruptured fault. Computing the ‘absolute’ displacement was also a longer process.

ALOS, the Japanese Advanced Land Observing Satellite, is used for cartography, regional observation, disaster monitoring and resource surveying. One of its main sensors is the Phased Array type L-band Synthetic Aperture Radar (PALSAR) – an active microwave sensor using L-band frequency to achieve cloud-free and day and night observation.

The SAR transmits waves at wavelengths from a few millimetres to tens of centimetres. The waves are reflected from the Earth’s surface, and part of their energy is reflected back to the SAR and recorded. Applying an image processing technique allows the intensity and phase of the reflected signal of each ground resolution element to be calculated.

Terrain slope, surface roughness and dielectric constants all have an effect on the amplitude or intensity of the SAR image. The phase is determined mainly by the distance between the satellite antenna and the ground targets; the effect of the atmosphere in slowing down the signal; and the interaction of the waves with the ground surface.

InSAR uses the interference of electromagnetic waves to measure distances accurately. It involves the use of two or more SAR images of the same area to extract landscape topography and its deformation patterns. InSAR is formed by interfering signals from two spatially or temporally separated antennas. The spatial separation of the two antennas is called the baseline. Two antennas may be situated on a single platform for simultaneous interferometry. InSAR can also be created by using a single antenna on an airborne or spaceborne platform in almost identical orbits for repeat-pass interferometry.

In the latter case, InSAR is capable of measuring ground surface deformation with sub-centimetre precision, at a spatial resolution of tens of metres over a large area.

An InSAR image is created by co-registering two SAR images and comparing their corresponding phase values on a pixel by pixel basis. The phase is mainly due to five effects: (1) differences in the satellite orbits when the images were acquired; (2) landscape topography; (3) ground deformation; (4) atmospheric propagation delays; and (5) systematic and environmental noises. The satellite’s position and altitude must be known to eliminate the effect caused by the differences in the two passes.

DInSAR (Differential InSAR) allows for the measurement of small deformations of the terrain, down to millimetre level, that have occurred between two different image acquisitions. In DInSAR analysis, the topographic contribution to the phase difference is carefully removed using DEMs (Digital Elevation Models). Atmospheric disturbances can either be accounted for using independent observations such as GPS, or neglected if the tropospheric delay is considered homogeneous at the time of the image acquisition.

InSAR has been used successfully to map ground surface deformation during volcanic eruptions and earthquakes. Early studies used images acquired before and after the event to record deformations that took place during such activity. These images have proven useful for understanding slip distribution and rupture dynamics during earthquakes. InSAR has also been used to map surface deformations immediately following an earthquake. It has also been successfully applied in studies of ground subsidence associated with the extraction of groundwater, and to map the slow movement of landslides.

The technology has also proved invaluable in monitoring short-term changes in the ice sheets over the Arctic, Antarctic and Greenland, improving our understanding of their impact on sea level change and global warming.

Associate Professor Linlin Ge, who leads the UNSW team, began studying DInSAR at UNSW in 2001, following research by Professor John Trinder into the use of InSAR for creating DEMs. Head of the university’s School of Surveying and Spatial Information Systems, Professor Chris Rizos, said: ‘We joined with other InSAR teams – from Oxford, Stanford and Harvard universities, and from the US Geological Survey – to generate the first ground displacement map for the China quake.’

He said that once the Japanese ALOS satellite successfully acquired imagery on 20 May, his team worked around the clock. They were able to supply early results, thanks to close collaboration with Japan’s Earth Remote Sensing Data Analysis Center (ERSDAC) and the Kochi Women’s University of Japan.

Ge said the imagery provided by UNSW could be overlaid onto satellite photos, allowing the China Earthquake Authority to assess damage around buildings or bridges. He said his team was among the first to generate a ground displacement map and was, at the time, the only team supplying analysis directly to the Chinese authorities.

At present, the ALOS satellite orbits Earth once every 46 days. More regular InSAR analysis could pick up ground movements before aftershocks occurred, said Ge. However, presently there are ‘less than a handful’ of satellites capable of providing the required radar images.

He added that once a comprehensive constellation of radar satellites is established, the technology would be able to predict aftershocks and landslides.

The Chinese space agency has plans to launch a new radar satellite, which will orbit Earth every four days. China plans to launch 100 satellites by 2020; about 20 per cent of these will be radar capable.

The project follows a research agreement, negotiated by Ge, between Chinese and Australian institutions. Under the arrangement, Australia’s CRC for Spatial Information will co-operate with the Centre for Earth Observation and Digital Earth at the Chinese Academy of Sciences. The initiative aims to strengthen China’s capacity for earth observation, mapping and monitoring of ground displacement.

Paul Grad is an engineering writer living in Sydney.
 
 
Issue 36; August – September 2008