Watching the Earth Move
PANGA's journey from theory to the international forefront of geophysics
By Rune Torgersen
Discovery Hall on Central Washington University’s Ellensburg campus is home to many wonders of science, including a giant telescope and one of the quietest places on Earth.
Hidden away on a rack of thousands of individual computer processors lies the capacity to monitor tectonic plate movement on a global scale, using a network of GPS receivers scattered across the world.
This is the heart of the Pacific Northwest Geodetic Array (PANGA), a multinational cooperative network of permanent GPS sensors comprising more than three dozen agencies and supported by the US Geological Survey, NASA, the US National Science Foundation, and the State of Washington through CWU. The network enables basic research into a host of Pacific Northwest geological phenomena—earthquakes, volcanoes, tsunamis, and landslides—as well as real-time analyses from anywhere on the planet that constitutes a basic pillar of earthquake and tsunami early warning systems. In Washington, Oregon and California, the PANGA lab in Discovery Hall is the primary GPS analysis system for ShakeAlert, the USGS earthquake early warning system funded by Congress.
CWU Professor of Geology and PANGA Director Dr. Tim Melbourne is at the center of this monumental feat of international cooperation, having helmed the construction of the initial few GPS receivers in 1998 to the several hundred operating today in the PANGA network, along with the development of the open-source software underpinning the analysis of its data.
Before his time at CWU, Melbourne’s fascination with geophysics led him to a job with the U.S. Geological Survey (USGS) as a software engineer, with particular focus on monitoring the activity of Mount St. Helens. In the days before the Global Navigation Satellite System (GNSS) that powers modern GPS applications, the process of keeping tabs on the active volcano was a very analog affair.
“You’d either hike in there or land in a helicopter, hammer in a reflector, and then set up a laser gun far away from where you might get killed by the volcano and take measurements off of that reflector,” Melbourne said. “Now, you could just go in there, set up a GPS receiver, and telemeter the data out via a radio network or something to get the data continuously.”
Once his career called him to start grad school, Melbourne found himself in Mexico, testing out the practical potential of using GPS receivers to measure deformations in the earth’s surface, as a supplement to more traditional seismological approaches. After planting the receivers in a small section of the Jalisco coast, his team struck scientific paydirt.
“There was this big old earthquake right beneath our network, and we were able to measure the deformation, how the land stretched, using GPS,” Melbourne said. “I published half a dozen really neat papers on that earthquake, so I was very lucky to have been involved in that project, and to have a magnitude 8 earthquake right under the instruments we had just set up.”
Knowing that this discovery was likely to represent a major leap forward in the characterization of geological events, Melbourne devoted a significant portion of his PhD thesis to exploring the algorithms necessary to turn that raw GPS feed into vital data about the movement of the earth’s surface.
“After my studies, I was absolutely aware of the power that this technology held, and I also knew that the only subduction zone in the continental U.S. was right here in Washington,” he said. “UW wasn’t interested, but there were two folks here at CWU who were, so I started my career in Ellensburg.”
With the help of former CWU Professors of Geology Charlie Rubin and Meghan Miller, Melbourne’s research had a way forward, and soon, a name.
“The name is sort of a joke,” he said. “A panga is a Mexican fishing boat, and those two liked to go to Baja and hang out on the beach where there are pangas all over. It just happened to lend itself nicely to the Pacific Northwest Geodetic Array.”
Growing Interest
GNSS-based geological measurements are taken from a receiver on the surface of the earth, as opposed to traditional vibration-based seismological measurements, which come from instruments embedded in the crust. That traditionally collected data can take a while to make sense of, particularly in the event of an earthquake.
“You cannot hide a serious earthquake from these stations, and we can characterize one much better than other seismic networks can,” Melbourne said. “There’s so much structure under the ground that the reverberations can make it hard to distinguish between the earthquake itself and everything shaking in response to it, which means that it can often take scientists 10 to 15 minutes to figure out the magnitude of the initial quake. Earthquake early warning is a matter of seconds, which is the kind of timeframe we operate in.”
This delay in processing can make a life-or-death difference in the case of underwater earthquakes, which are prone to causing tsunamis. Former NASA Solid Earth Program Director John LaBrecque started his career in oceanography and was already experimenting with GPS applications in the field in 1975.
“My entire research community began to get interested in earthquake prediction, and understanding the motion of the earth’s surface,” he said. “After a major tsunami in the Indian Ocean killed over 200,000 people in 2004, we discovered that if we had the GPS measurements from existing stations at the time, we could have predicted the magnitude of the earthquake that caused it, and better understood the wave before it hit, potentially saving hundreds of thousands of lives.”
LaBrecque had worked with Melbourne’s father, Bill, at NASA’s Jet Propulsion Lab, and found himself excited enough about the work being done at PANGA to throw the agency’s financial weight behind the project.
“It really was an attempt to build and strengthen the effort of monitoring ground motion, and using that data as a predictive capability,” LaBrecque said. “At NASA, it was great to see that others were attempting to build this out, and even more wonderful that I had access to the resources to help them do it.”
To help extend the reach of CWU’s work, LaBrecque and his wife, Shirley Winkler, recently established a $500,000 fund to expand earthquake and tsunami monitoring across Oceania in the central and southern Pacific. The gift will strengthen research and early warning systems throughout the region.
Dr. Tim Melbourne visits monitoring stations like the one at Ross Dam (pictured above) for maintenance and calibration.
A Unique Opportunity
As PANGA continued picking up steam in the early 2000s, Melbourne was presenting the group’s findings both domestically and abroad, including a guest lecture at a university in Costa Rica, attended by Ana Aguiar Moya (’07), a physics student looking for her next step.
"I was really excited about applying my physics experience to fieldwork, so Tim suggested that I apply to CWU for my master’s degree,” Aguiar Moya said. “I landed there and immediately found myself in the field learning about the local geology. I was the only physicist in a cohort of geologists, but I found my legs soon enough.”
Aguiar Moya’s fascination lies primarily in untangling and understanding messy data sets, a skillset that aligned perfectly with PANGA’s work. With her help, the group was able to identify an until-then poorly understood geological phenomenon known as the “slow-slip event,” which shows up in the data as a series of jagged, steady ridges indicating tiny, constant earthquakes.
“That sawtooth pattern is what paid the bills for a long time,” Melbourne said. “The Canadians beat us to it, but they published a paper that said the pattern is caused by deeper parts of the plate slipping a little bit, essentially little slow earthquakes happening all the time, far beneath the ground. This was the first time we had ever seen this happen, and for 10 years after that, I never had an NSF proposal get rejected.”
Because of the cleaner datasets provided by GNSS, these events could now be studied in fuller detail. Aguiar Moya’s work on it set her on the path to earning her PhD at Stanford, as well as her current focus on understanding explosions across the world by measuring telltale ground vibrations for the Lawrence Livermore National Laboratory in California.
“My experience with PANGA completely realigned my career and set me on my current trajectory, because that’s where I got interested in these very, very small seismic events,” she said. “I learned how to program tools to analyze that kind of data and extract even more knowledge from it. Once I became a postdoc here at Livermore, I was still studying microseismicity, so my decision to attend CWU was really a defining moment for me.”
As the work evolves, so do the opportunities available to students hoping to get involved with PANGA. Most recently, Isabella Haarer (’25), a master’s student from Germany who came to the U.S. to run track and field, helped develop the software that allows PANGA to visualize tectonic movement online, in real time. Her work on the project helped her secure a prestigious position with the Deutsches Zentrum für Luft- und Raumfahrt (DLR), Germany’s answer to NASA.
“I came into the project with almost no programming experience, which I had to learn,” Haarer said. “I did, and I took those skills with me to my current job at the DLR. Most of what I do now is software development and data analysis, which is also what I worked on for PANGA.”
Melbourne, right, discusses equipment usage with CWU Students.
A Global Model
These days, PANGA receives and processes data from over 2,000 GPS stations across the world with less than a second of latency, ensuring real-time alerts and monitoring for anyone who wants it.
The program receives funding from a whole host of agencies, including USGS, NASA, and the National Science Foundation. It provides the data-processing power for the West Coast’s critical ShakeAlert system, designed to automatically warn affected populations within seconds of detecting an earthquake. The technology also activates fail-safe mechanisms in water and gas lines to minimize damage.
Across the globe at Earth Science New Zealand, Science Operations Technical Lead and International GNSS Service Governing Board Vice-Chair Elisabetta D’Anastasio found herself tapping into Melbourne’s expertise in the field during the process of setting up a similar system in her country.
“Because I have this expertise in GNSS, but we weren’t able to process our data in real time like PANGA does, it was perfect to contact Tim for that knowledge sharing,” she said. “We were wanting to develop a similar capacity here, and since PANGA has been delivering those real-time GNSS waveforms for quite some time, that was our benchmark to test our systems.”
D’Anastasio has great respect for the work being done at PANGA.
“PANGA has been at the forefront of GNSS applications in geophysics since the start,” she said. “I believe it was one of the first examples of real-time processing of GNSS to produce those waveforms applicable to geophysics. And they’ll take data for processing from anyone willing to share it, helping to expand global understanding of these seismic events.”
For Melbourne, growing PANGA’s network of collaborators and continuing to improve on the system’s core data-processing capabilities is all part of his mission to save as many lives as possible, and understand our Earth as well as we can.
“None of us, when we get in our car to go to the grocery store, focus on the small numerical chance that we’ll get in a fatal accident on the way,” he said. “What we do is we put on seat belts, make sure our airbags are working, and follow the rules of the road.”