Lidar Reveals Hidden Aftershock Hazard in Haiti

Marie Freebody, Contributing Editor, marie.freebody@photonics.com

Rescue agencies the world over have rallied together in a bid to bring relief to the thousands of people left injured, homeless and without food or sanitation in the wake of the Haiti earthquake, which devastated the region in January this year.

One of most troublesome questions looming in the minds of many Haitians and aid workers alike is the possibility of further tremors. The fear of severe aftershocks is so real that, months after the quake, many people are living in makeshift tents set up in open areas away from the danger of falling debris.

This displays the point cloud of lidar from the Rochester Institute of Technology airborne system.

Crucial lidar-based mapping of Port-au-Prince and the surrounding area is providing survivors and rescue workers with invaluable information about the ground on which they find themselves trying to pick up the pieces.

According to an assessment of aftershock hazards carried out by the US Geological Survey (USGS) at the request of the United Nations, the frequency of aftershocks following this magnitude 7.0 earthquake will diminish with time, but damaging earthquakes remain a possibility. There is also a small chance of subsequent earthquakes larger than the initial shock.

An oblique three-dimensional view of the Idrija fault zone derived from the lidar digital terrain model (DTM) is shown. Arrows point to clearly defined fault segments. Courtesy of Stephen Grebby.

The geologic fault that caused the Port-au-Prince earthquake is part of a seismically active zone between the North American and Caribbean tectonic plates, known as the Enriquillo fault.

“The Enriquillo fault represents a real threat; it has a long-term slip rate of about seven millimeters per year and has, in its history, experienced large and damaging earthquakes as well as the 2010 magnitude 7.0 earthquake,” said Dr. Ken Hudnut, a geophysicist at the USGS.

The images above and below show an old stream channel that appears to have been offset about 163 m left-laterally by the Enriquillo fault in Haiti. This feature will be checked by geologists in the field and sampled to see whether material in the channel bottom sediment may be datable by radiocarbon methods to derive a long-term slip rate for the fault system. The current understanding of a slip rate of 7 mm per year is derived from a model based on global position system station velocities. Using data such as this, the USGS hopes to augment this information by obtaining a long-term geologically based slip rate. USGS images by Ken Hudnut.

Airborne lidar surveys of Port-au-Prince potentially may reveal the distribution of slip along the active fault in enough detail to enable the USGS to establish whether the slip was caused by early historic or prehistoric earthquakes.

“High-resolution airborne and satellite imagery revealed an absence of surface faulting,” Hudnut said. “The sections of the fault to the east and west of the portion that broke in this earthquake seem to have not broken since early historical events. This gives rise to a near- to intermediate-term concern of a larger event on this fault system to the east or west of the section that broke in the January 2010 event.”

This information is vital for the reconstruction effort currently under way in the area, and the USGS is already planning follow-up assessments. Accurate and up-to-date forecasts will help ensure that expectations are accurately set and that any new buildings can be properly engineered to minimize damage caused by possible future tremors.

Similar lidar projects are being carried out by the Leicester Lidar Research Unit in the UK, in which airborne lidar is employed to map seismogenic faults in densely forested terrain in the Julian Alps of Slovenia. In this area, the forest is so thick that it obscures subtle surface traces of faults captured on aerial photographs and also makes fieldwork logistically difficult.

Because laser pulses emitted by lidar systems can penetrate through gaps in the forest canopy to reach the ground, scientists can build up a “bare earth” digital elevation model (also referred to as a digital terrain model, or DTM) by eliminating laser reflections deemed to originate from the canopy.

“This ‘virtual deforestation’ of the terrain helps to reveal surface features and tectonic landforms of the fault systems in unprecedented detail, which ultimately contributes to a regional seismic hazard assessment program,” said Stephen Grebby, a research student in the geology department of the University of Leicester.

(a) Digital elevation model of the Idrija fault zone created using lidar data points originating from both the ground and the forest canopy. (b) Lidar DTM created through “virtual deforestation” reveals in unprecedented detail the surface trace of the Idrija fault, along with many other surface features. Courtesy of Stephen Grebby.

Both terrestrial and airborne lidar can provide geologists with accurate and high-resolution information about the surface morphology over large areas and in areas that would otherwise be inaccessible in the field. According to Grebby, lidar data allows scientists to study and try to understand surface phenomena in greater detail than was previously possible using more traditional approaches such as field surveys, aerial photography or satellite imagery.

Grebby is working on another study, carried out in conjunction with the British Geological Survey, in which lidar is applied to mapping in areas where the spectral identification of different rock types is hindered by the presence of vegetation. Rocks in the northern foothills of the Troodos ophiolite in Cyprus are found to respond differently to surface processes, such as weathering and erosion, according to their type and therefore typically have relatively distinct topographic characteristics.

Airborne remote sensing data used for lithological mapping of the Troodos ophiolite, Cyprus. (a) Airborne thematic mapper multispectral imagery, (b) Airborne lidar digital elevation model and (c) Lidar DTM after “virtual deforestation.” Courtesy of Stephen Grebby.

“I’ve been developing a methodology that describes the topographic characteristics of the main rock types using morphometric variables, such as topographic slope, roughness and curvature, derived from a lidar DTM. I can then generate a lithological map of the study area by classifying according to these topographic characteristics,” Grebby explained. “I’m now investigating whether lidar-derived topographic information can be combined with multispectral information to further enhance the lithological mapping performance.”

The business of lidar

Lidar is important not only to seismologists and geologists, but it also is becoming big business for numerous market areas, ranging from insurance services to mineral exploration. Over the past 12 years, mapping specialist Infoterra Ltd., also based in Leicester, UK, has been using lidar to provide 2-D mapping and 3-D visualizations for a wide range of commercial companies.

“Typical commercial applications of lidar technology in the UK include landscape modeling for flood risk analysis – this can either be broad-area surveys at river catchment scale or very precise site-specific surveys, for instance, modeling flood risk associated with electricity substations or similar infrastructure,” said Dr. Anthony Denniss, chief operating officer at Infoterra.

Other applications include measurements around electricity power networks to calculate the risk of vegetation encroachment upon power lines, as well as to measure the curvature or “sag” of the power lines between each pylon.

“This information allows power companies to calculate whether individual lines are able to carry more power or not,” Denniss said.

But Infoterra’s latest application of lidar reflects an emerging trend that is now beginning to embrace the technology – mapping cities in unprecedented detail.

Infoterra is currently mapping the major UK city centers to provide highly detailed models that are further enhanced by high-resolution oblique images that are “wrapped” onto the sides of the buildings. The end result is an accurate representation of the city that is useful for city planning, surveying and architecture applications.

“The sensor that we currently operate is one of the latest generation sensors, which means we can record a maximum of 167,000 measurements per second as we fly,” Denniss said. “Depending on a number of variables, such as flying speed, height of the aircraft above the ground and ground elevation, this scanning frequency translates into a ‘points per square meter’ (ppm) measure on the ground. Typically, for rural areas we collect data at 4 ppm, but for urban centers we increase this to 12 ppm, allowing detailed building structure to be defined.”

Coming back down to Earth

At ground level, both terrestrial and mobile lidar tend to be used where even more detail or precision is required, as both systems can produce mapping data accurate to millimeters. For example, Infoterra opts for static terrestrial lidar for precise engineering surveys, with a recent project involving the survey of more than 300 UK motorway bridges ahead of road widening schemes.

Mobile lidar technology introduces a new generation of sensor developed within the past few years specifically to operate on the roof of a moving vehicle. According to Denniss, numerous mobile lidar systems are now being launched but are best employed in applications such as road corridor mapping. And while these systems are slightly less than a static system, they do offer other advantages, such as being far more cost-effective, as long stretches of road can be examined in a short period of time.

Collecting data on the move also eliminates the need for lanes on highways to be closed while the survey is in progress, reducing traffic disruption. In urban areas, mobile systems are allowing town planners to map the “as built” environment in a way never before possible because of high costs previously associated with data collection.

Lidar technology has enjoyed rapid development in recent years. For example, when Infoterra purchased its first airborne sensor 12 years ago, it was able to collect 5000 measurements per second (mps). Today, airborne systems can achieve 400,000 mps, whereas terrestrial scanners routinely operate at 500,000 mps, with some close to 1 million mps.

This trend is expected to continue, while at the same time the sensors themselves will become smaller and more power-efficient. And as more sensors have appeared on the market, the cost of these systems has decreased in real terms, making the technology more accessible, which in turn results in its being used more routinely.

“Simple, low-cost sensors that were initially developed for factory applications, such as for the control of assembly line robots, are now making their way into the world of navigation,” Denniss said. “Research projects to date have successfully demonstrated how data from these simple laser sensors can be processed in real time and integrated into navigation systems, allowing autonomous vehicles to be navigated around complex terrains.”

Coupling this with the growing trend to map everything in ever-increasing detail could herald a new era of navigation and visualization services. “If you think about it, most mapping of complex environments tends to be two-dimensional, but now it is possible to cost-effectively map these same environments in high-definition 3-D,” Denniss said. “Once these high-definition 3-D mapping environments become commonplace, it will open up a completely new era of navigation and visualization services.”