Capture of post-disaster conditions

In the aftermath of an earthquake, hurricane or other disaster, the details of precisely how the built and natural environments reacted to the forces unleashed upon them lie fresh in the devastation. Clues to why a home collapsed under high winds may be found in its rubble, and an explanation of how a slope turned into a landslide as the earth shook may be spread across a four-lane highway.

This vital information is rarely captured before it has been altered or removed. Emergency managers tasked with recovery efforts are trained to move quickly to restore basic services and get the local economy back on its feet. Roads are cleared, damaged bridges are repaired and crumpled buildings are razed. Ironically, the data that could be most valuable in rebuilding the community is lost in the rush.

‘That data is perishable; if there isn’t someone to swoop in and capture the information, it is gone forever,’ said Joe Wartman, H.R. Berg Professor of Civil and Environmental Engineering at the University of Washington (UW). ‘When one of these events occurs, it is a full-scale living laboratory of consequences and effects of wind speed, ground shaking and destruction.’

UW has taken the lead to ensure baseline geospatial data can be collected, processed and analysed immediately after natural hazard events to inform recovery and guide rebuilding. With US$4.1 million in National Science Foundation funding and cooperation from Oregon State University, Virginia Tech and University of Florida, UW has established the Natural Hazards Reconnaissance Facility (RAPID Facility) on its Seattle campus.

‘Data collected by the RAPID Facility will directly support the development and validation of simulation for natural disasters,’ said Wartman, the facility’s director.

‘These models can help communities anticipate what will occur in a disaster and take action to mitigate risk.’

RAPID has spent much of the past year acquiring a variety of geospatial technologies and training personnel in their application under difficult conditions. The facility has pinpointed LiDAR technology in the form of long-distance laser scanners, along with 3D data analysis capabilities, as among the most important solutions that will be deployed.

Speed, accuracy, safety

The RAPID Facility’s purview is to study natural disasters related to earthquakes, tsunamis, hurricanes, tornadoes, and floods. While their causes are distinctly different, the results are usually the same – they inflict often massive change on the built and natural environments. Winds, tidal surges, flood waters and ground tremors can knock a building off its foundation and dramatically transform the surrounding landscape in the process.

‘Laser scanners are unique in their capacity to capture the details of changes in three dimensions and with centimetre-level accuracy,’ explained Jake Dafni, PhD, PE and RAPID Site Operations Manager.

Because of the challenging conditions in which RAPID teams will conduct surveys, mapping and data collection, the facility is rigorous in its choice of technologies. Hardware must be rugged, and software packages need to be easy to learn, given that response teams are composed of researchers from diverse technical backgrounds.

RAPID ultimately selected Maptek as its supplier of long-range laser scanners and spatial analysis software. Maptek XR3 and LR3 models were designed primarily for use at open pit mine sites. Ruggedised and equipped with integrated optics and GPS, the laser scanners are lightweight, dustproof and can be used in rain, snow and extreme temperatures.

The primary driver was the ability to capture highly accurate scans from a long distance. The XR3 has a 2.4 km range while the LR3 operates at 1.2 km. The laser capture enormous volumes of data in their fields of view, allowing the RAPID teams to collect data quickly while located a safe distance from a still-unstable disaster site.

‘There are damage scenes that we can’t otherwise get close to for physical reasons,’ said Dafni, noting that speed and safety are paramount to their missions. Additionally, RAPID operating procedures call for not interfering with ongoing site rescue efforts.

‘During a pre-RAPID exercise, we deployed to the site of a large earthquake in New Zealand and we were interested in studying some large landslides across a valley. We couldn’t get across a river to scan onsite, and there was a lot of emergency helicopter activity preventing our use of drones,’ said Wartman.

This incident led RAPID to add long-range scanners to its equipment acquisition list, which includes digital cameras, total stations, mobile LiDAR, seismic sensors, water level gauges and aerial and marine drones.

Informing better building

Laser scanners are crucial to the RAPID Facility’s core mission because they capture in three dimensions the details of how the natural environment and built infrastructure interacted during the disaster. In a tsunami or earthquake, for example, it might be a wave of sediment or falling rocks that wipes out a bridge, and not the shaking ground. Such information is critical in determining how to engineer structures to withstand future events.

‘We’re supporting the geologists who want to know why a landslide occurred, but we also focus on the engineering community, which is responsible for the built infrastructure,’ said Wartman.

Central to the analysis is 3D point cloud processing. UW’s Civil and Environmental Engineering Department has a long-standing relationship with Maptek as a user of its I-Site Studio geotechnical analysis software.

The easy-to-learn software differs from most LiDAR processing packages in that it performs registration of the laser scans, processing of the point cloud and then numerous complex geotechnical analyses.

For RAPID purposes, its most important modelling function is the ability to digitally compare and contrast two point clouds acquired in the same area at different dates to detect changes, specifically movement of the ground, rock faces and even building structures.

With change detection in mind, RAPID has agreements with international organisations that maintain large archives of historical LiDAR data and remotely sensed raster imagery. These files will serve as before-disaster data in the change modelling.

Analysing pre- and post-event point clouds with I-Site Studio software can detect subtle differences in infrastructure and landscape that help explain what happened during the disaster.

‘This information will be used to develop models to predict locations of landslides in future events, for example, so communities may take action to mitigate landslides that will impact critical infrastructure,’ said Wartman.

RAPID and Maptek are working together to find ways to further the RAPID Facility’s mission. They are now engaged in using Maptek software to create an immersive environment called The Cave on the UW campus where researchers will step into a 3D disaster scene, generated from LiDAR data and colour photos, to interact with the data from an in-person perspective.

Thanks to Joe Wartman
H.R. Berg Professor of Civil and Environmental Engineering
University of Washington