Leadership Talks – Signaling a Shakeup

Ken Hudnut (KH), U.S. Geological Survey geophysicist and leader of the Southern San Andreas Fault Evaluation (SoSAFE) Project, spoke with Managing Editor Tracy Cozzens (TC) on June 28.

Ken Hudnut uses GPS to locate a boulder offset 560 meters by the San Andreas fault

TC: Can you give us an update on the L1C finalization and approval process?

KH: L1C has been specified in a document called the IS-GPS-800, and the initial public release of that was in April 2006. So we’re now over one year through the public review process called the Interface Control Working Group (ICWG). That’s a formal process whereby the GPS Wing, through a Federal Register announcement, asks for public input from the international GPS user community. And that formal process of accepting and handling comments is nearing conclusion. So all public comments that have come in have been addressed in the current version of this IS-GPS-800 document. Interested people can find the document online at the GPS Wing website.

The IS-GPS-800 specifies all details of L1C signal design, and that’s in common with the earlier interface control documents for other GPS signals. That makes it so that people can begin designing receivers today, well in advance of launch. That’s the whole idea, to make it so that signal designs are entirely open and entirely specified for manufacturers and other users and stakeholders so that they can see every detail of the signal design.

This “ShakeMap” of estimated seismic ground-shaking intensity shows the San Andreas Fault during a Big One scenario, a magnitude 7.8 earthquake. California is heavily ‘wired’ with GPS stations that measure plate motion.

That review process is basically wrapping up and the intent of the GPS Wing is to move forward. They would like to get everything pinned down on GPS Block III before going out for the next stages of their acquisition and procurement process.

TC: When do you think we can expect the first Block III launch?

KH: People have been saying for some time now that initial launch of Block III would occur in August 2013.

TC: How long thereafter until L1C is operationally useful?

KH: It’s a gradational thing, and I think the anticipated rate of satellite launch would be at three per year, optimistically. So IOC (initial operational capability) would be expected about five years into Block III, around September 2018. IOC to FOC (full operational capability) is also gradational, but FOC would be in the neighborhood of 2021.

TC: At either stage, you’d still be able to make use of the signal.

KH: Yes, so as soon as a new signal is introduced, people are going to start jumping on it and using it. One example of that is L2C, which is currently being broadcast and received, and people are starting to make use of it. As more and more satellite signals are introduced, GPS becomes more and more useful. L2C is significant because it gives civilians a signal on L2 that they can legitimately use, so that’s a plus for civil users. Having that L1/L2 combination is very powerful for civil users. In the future with the introduction of L5, it will allow us to do tri-laning, which we are also looking forward to.

This map represents the Plate Boundary Observatory, funded by the NationalScience Foundation and built by UNAVCO, which increases the accuracy of real-time earthquake, volcano, and tsunami alerts.

TC: What other new opportunities exist for further enhancing GPS in later segments of Block III?

KH: Overall, the mindset is that people want to see to GPS sustainment with Block III, so the initial Block III satellites are going to have that as their priority. But there have been several interesting proposals for things that could be added to later Block IIIs. The hope is that there will be an opportunity to test those new ideas while at the same time, very importantly, minimizing risk. People would like to test and insert new technologies as early as possible within Block IIIs, but not at the risk of threatening constellation sustainment. Whatever is done will be done cautiously. There’s a lot of interest in several things that might be done to make GPS even better.

TC: Can you give me some examples?

KH: It’s probably a little too early. There are several really interesting concepts being floated right now. But they’re really not ready for prime time.

TC: How do you see the addition of the L1C signal improving early-warning systems?

KH: L1C itself is significant in that regard because we hope to phase it in along with new Galileo signals, so we’re talking about having many more satellites in the overall GNSS in that timeframe. Overall, the improvement for real-time uses — that we’re needing and looking forward to — have to do with improved accuracy and robustness in real time. L1C will help on robustness. The accuracy and precision, we’re looking at seeing those improvements even before L1C gets introduced. But basically, some of the techniques being used today — real-time precise point positioning (PPP), for example — are allowing decimeter , or better, level positioning in real time. We are starting to make better use of that in our early-warning systems on a test basis.

We’re doing work now that allows us to make use of GPS as it exists today. One limitation is simply the number satellites, but other limitations result in the ways in which people are doing dual-frequency carrier-phase positioning today, and doing carrier-aided pseudoranging through precise point positioning. All of these techniques are doing what you can with the existing GPS signal. Some of the future enhancements for GPS, and also other GNSS improvements, will definitely be helping us with real-time ambiguity resolution and therefore also with real-time precise point positioning that’s more robust, that’s more precise. Those are the areas where we really need improvement, possibly even beyond what currently is planned for GPS.

TC: As the leader of the Southern San Andreas Fault Evaluation (SoSAFE) Project, can you describe how GPS helps predict earthquakes?

KH: The overall project goal is to get a better idea of past earthquake behavior on the San Andreas Fault. That’s primarily studying evidence from trenches and using radiocarbon dating. So we’re trying to piece together the past history of earthquakes over the last 2,000 on the fault system as a way of better anticipating what may happen in the future. That’s the main purpose of the main SoSAFE project itself, which is part of a broader activity called the USGS Multi-Hazards Initiative.

As part of that, we’re also going to conduct a large-scale exercise on November 13, 2008, called ShakeOut. We’re going to try to get everyone in Southern California, actually statewide, engaged in an enactment of the Big One on the San Andreas Fault. My involvement has been to specify all the earth science details of that earthquake, using all the scientific information we have available, and taking into account the input of all the experts. From that, we’ve created an earthquake, and we’re doing modeling to simulate the effects of that earthquake on the socioeconomic fabric of Southern California. We’re going to do a full-scale simulation exercise and try to have everybody involved — news media, schools. This is the kind of thing emergency responders do all the time. It’s something people do to plan and be ready for future actual events. In our case, we’re hoping to do this on a scale that’s never been done before, at least in Southern California.

TC: So your GPS data will be used to help create this hypothetical earthquake?

KH: At this point, we have more than 410 continuously operating GPS stations in Southern California. Those are an integral part of earthquake monitoring. Right after the Northridge earthquake in 1994 is when we got our funding. Between 1996 and 2001, we built the first 250 stations and called that the Southern California Integrated GPS Network (SCIGN). We then, along with many other colleagues throughout the earth sciences, were able to get funding from the National Science Foundation to build EarthScope (www.earthscope.org), part of which is the Plate Boundary Observatory, which includes 875 additional new key GPS sites throughout the entire Western U.S.

So, since about 2002 that array has been constructed. The Plate Boundary Observatory is nearing completion. Now Southern California is really wired with continuous GPS receivers. What we’re doing is tracking the accumulation of strain on the fault system with that array. As the plate tectonic movements build up across Southern California, we can see that happening. Every day we measure the positions of these sites to within a couple millimeters. In the event of a big earthquake, of course, the ground is going to shake like crazy, and then each GPS receiver ends up in a different position. Afterward we can actually image the fault slip at depth by using the GPS data along with seismic data. With the SoSAFE project in particular and the GPS array we’ve built, we aren’t out to predict earthquakes in the usual non-scientific sense. What we’re doing is trying to understand them as scientists, and we call this forecasting. We are studying the physics of the earthquake source, which is chaotic because of friction, and trying to predict aspects of earthquake behavior at the system level.

Another thing I’m excited about is the use of GPS technology to do what we call earthquake early warning. It’s different than earthquake prediction or forecasting. You see the earthquake has started, and then much faster than the shaking has arrived, you send the word down the wire literally. You outrace the earthquake. Typically that’s been done with seismic instruments — accelerometers, basically. If you have an array of accelerometers along the fault system and distributed throughout the whole region, then as soon as the shaking starts you can detect that and triangulate the position of where the earthquake has begun, and then you can send word ahead that an earthquake is coming.

What we can do with GPS is actually build what I call a zipper array along the fault. In California, the source of future Big Ones will be the San Andreas Fault System and one of the branches of it, the San Jacinto fault. Either one could be the source of a future Big One. Over two-thirds of the plate motion happens on these two faults. Since it’s conceivable that a large, damaging earthquake could happen on another fault, we have to hedge our bets. It’s not like waiting and wondering which storm track some hurricane’s going to take, where you don’t know exactly where landfall’s going to occur. In our case, we can put a pretty sure bet on it. These faults are the big players. So if we put out instrumentation along these faults, it can help us with early warning for the Big One.

GPS can actually “see” the fault slip right when it happens. Seismic instrumentation is also good at that, but if you picture that the fault actually starts slipping, and you have a road that crosses the fault perpendicularly, the road will actually be laterally offset by the fault. So imagine that you have GPS receivers next to the road on either side of the fault, and as soon as it slips, you can see that happen. GPS is going to be crucial to future development of improved early warning systems.

The big problem with earthquake early warning has been false alarms. The seismic systems that don’t have an independent measurement technique built in are prone to false alarms because it’s a single technology. If you have a glitch that affects many instruments at once, it’ll set off a false alarm. Whereas if you have a system that includes other independent technologies along with the seismic, such as GPS, then you’re building in robustness to the system and it will be far less prone to false alarms.

That’s the future vision that I have; GPS will become embedded in our earthquake early-warning systems. In the next decade, we hope to fully embed GPS alongside inertial sensors for our future earthquake monitoring systems.

TC: In the past decade, you’ve come a long way in embedding GPS along with seismic monitoring. But you’re not using GPS in earthquake early-warning yet?

KH: We’re not doing earthquake early warning yet. It’ll cost some $20 million to $40 mmillion to do it right. We’re not doing as much as we could for very rapid estimation of the fault source. We’re working right now to get GPS even better integrated with the existing earthquake monitoring systems. We’re trying to do more with GPS in real-time.

TC: You’re mostly using GPS to do the research you were describing before.

KH: We’re very effectively using GPS for research. When an earthquake occurs, we’re now able to use GPS within a day. At the time of the Northridge earthquake in 1994, it took a week to make full use of the GPS data. In the future we want to be able to use the data within 10 minutes, and even instantly. Though we’re testing new ideas for this now, there’s a long way to go.

A group of us just sent in a major coordinated proposal to NASA. We’re trying to pursue the use of GPS in precise point positioning mode to do early warning for earthquakes, tsunamis, and volcanoes, all using these robust techniques for precise real-time positioning. We’re hoping to make some important technological advances, just using GPS as it exists currently. As new GPS capabilities come along, and new GNSS capabilities come along, we hope to make full use of those signals as well.

For real-time positioning today, it does help to have GPS and GLONASS. It certainly helps with any real-time positioning application to have more satellites in view, so I look forward to the future of several GNSS constellations with interoperable signals on several frequencies. For early warning of natural hazards, these advances will help greatly.