A photo of Hurricane Katrina over southeast Louisiana. Photo courtesy of NOAA.
by Craig Dylan
Posted: October 1, 2009
Louisiana’s GULFNet provides reliable elevations when it really matters.
When record storm surges hit New Orleans and breached the 17th Street Canal levee in the wake of Hurricane Katrina at the end of August 2005, experts knew the situation was bad. “I was watching it on TV from where I’d evacuated, and I couldn’t believe it at first,” says Mark W. Huber with the U.S. Army Corps of Engineers (USACE). “‘Life’s not going to be the same now,’ I thought.”
Huber and others in the USACE immediately began organizing emergency survey operations that would continue for months. Jimmy Chustz, PLS, president of Chustz Surveying Inc. in New Roads, La., a private survey firm under contract to the USACE, may have been the first surveyor to get to New Orleans following the disaster. “It was devastating to watch it all happen,” he says. “The first thing I thought was that we needed to get in there immediately.”
Prior to Katrina, Chustz had stored boats and other equipment in a flood-proof location away from trees, which allowed his crew to deploy rapidly. “The levee broke on Monday, and we were there Tuesday morning,” he says.
Roy Dokka, PhD, travels to various locations throughout Louisiana equipped with a Trimble R8 GNSS receiver (on the vehicle's roof) and a Trimble TSC2 Controller to measure the tops of levees. Photo courtesy of Saundra Dokka.
However, once the surveyors arrived, they faced another challenge. The entire city was under water, including key bench marks, and most of the region’s monumentation was known to be off vertically by nearly a foot due to subsidence. What’s more, Louisiana State University’s (LSU) GULFNet, a network of continuously operating reference stations (CORS), had been hit hard by the storm conditions. “Some stations were blown down; others had lost power. And just getting to them was difficult because roads had been washed out,” says Roy Dokka, PhD, the LSU civil and environmental engineering professor who originally conceived GULFNet. A team from LSU headed by Tony Cavell, PLS, and other volunteers scrambled behind the scenes to re-establish the network. Receivers were repaired and replaced; solar panels were installed to provide emergency power, and satellite uplinks were established. Within one to two days, individual reference stations were once again operational, and the entire network was functioning within two weeks.
One important post-Katrina adjustment was relatively easy. Critical early survey tasks included damage assessment and the location of navigational hazards, and airborne surveying was the quickest way to get this work done. To improve the accuracy of airborne LiDAR, photogrammetric surveys and aerial photography, the USACE asked Dokka if he could increase the sampling rate of GULFNet’s reference stations. In response, Dokka increased the rate from once every 15 seconds to once per second, which made post processing more precise by lessening the time lag of the aircrafts’ GNSS receivers.
Looking west from the Trimble Zephyr Geodetic 2 antenna for station 1LSU on the LSU campus.
Meanwhile, Chustz and others were able to rely on GULFNet for positioning in areas where existing monumentation had been wiped out. “Communications were terrible, and the corps had evacuated to Vicksburg, Mississippi,” Chustz says. “We finally got through to them, and they sent us to the 17th Street Canal to run sections with single-beam hydrographic equipment. The whole city was under water, so we got around by boat. There weren’t many bench marks or known points available, but GULFNet CORS were accessible and really sped things up.”
As the water slowly receded, Chustz and his crew set up Trimble R8 GNSS or 5700 GPS receivers on whatever dry ground was available and performed static GNSS work to establish control points. These points were then used as references for the receivers during the hydrographic surveys. The Trimble TSC2 Controller was switched as needed between GNSS receivers and total stations.
Much of the early work was overseen by Huber, who was one of the first government surveyors on the scene. Although GULFNet was repaired quickly and seemed to be working well, Huber, like any surveyor, was looking for some verification.
Finally, on Oct. 12, 2005, NGS released new figures for area bench marks, giving Huber something to hang his hat on. As it turned out, GULFNet’s elevations and the NGS’s new figures agreed within 0.15 feet or less in critical areas, which, Huber says, made him extremely happy. “Under the circumstances, it was quite a relief to have our work verified,” he says.
In the following months, surveyors working under emergency conditions throughout southern Louisiana constantly turned to GULFNet for reliable positioning.
GULFNet Gets an Upgrade
The flooding from the levee failures devastated large areas; water often rose to roof level.
While GULFNet proved itself an invaluable resource following Hurricane Katrina, Dokka believed that using Trimble VRS technology to convert the system to a real-time network (RTN) could make it even more useful. “When we first built GULFNet, we didn’t know about VRS,” Dokka says. “So when we learned about it from NEI [Navigation Electronics Inc.] soon after Katrina in 2005, we integrated VRS into everything.”
The decision provided instant benefits for Huber and the other surveyors using the network. “We had been holding the GULFNet stations as fixed when we did our own network adjustments, which was incredibly useful,” Huber says. “But when VRS was added to the system, we began to do RTK without base stations for engineering studies and most of our other work.”
In 2009, the USACE did a post-Katrina assessment of every federal hurricane structure in southeast Louisiana--levees, canals and floodwalls--to analyze the region for “hot spots” in advance of the upcoming hurricane season. “Our in-house survey crew (just two guys) was able to get to every structure and shoot hundreds of miles of profile, as well, in just a couple of weeks,” Huber says. “Before VRS, it would have taken considerably longer, and before GULFNet, it would have taken months.”
Workers use boats to get around and GPS to navigate even as flood waters recede (note the water line on the building).
To illustrate another way the GULFNet RTN has helped him with his daily work, Huber likes to tell a story about the time he received a call about an apparent 7-foot horizontal bust in sections collected by a contractor at the 17th Street Canal.
Concerned, but with an idea about what was causing the problem, Huber hung up the phone and ran out to take check shots on the floodwall with a Trimble 5800 GPS Receiver and the GULFNet VRS. “Everything checked with the rest of our work,” he says. “So when I got back to the office, I called the surveying firm that performed the initial surveys and asked if maybe they’d been using international feet instead of U.S. survey feet--and that turned out to be the case.” The very minor difference, when applied to the large numbers that characterize state plane coordinates, created the 7-foot displacement. Working by himself, Huber was able to confirm his suspicions and resolve the error.
Dokka has similar stories. Soon after Katrina, the State of Louisiana asked him what it would take to conduct an independent analysis of levees and other flood structures in the southern half of the state. State officials thought it would cost millions and take years. “But,” Dokka says, “Tony Cavell and I were able to do it for them in three months at a fraction of the cost they had in mind.” In fact, flood-structure assessment has become so efficient with the VRS-enhanced GULFNet that the corps plans to take a “snapshot” look at the entire Louisiana system every year.
“After all,” Huber says, “on the levees where we can get a four-wheeler on top, one person can profile 40 miles a day, calibrating at bench marks along the way. Why not take advantage of that?”
Huber, who is now at the U.S. Army Geospatial Center, also sees a lot of “intangible” benefits to RTNs. “Networks like GULFNet level the playing field,” he says. “If we do a dredging project, for example, and I work on GULFNet to estimate how many cubic yards of material have to come out and the dredge comes in and measures from the same continuously updated reference system, that’s very comforting because their work is more closely related to mine. Even better, navigation can be on the same network. Simply put, if everyone is using the same reference system, we’re all comparing apples to apples.”
That Sinking Feeling
Cavell measures the height of a floodwall near a damaged section in New Orleans.
Though the emergency and practical uses of GULFNet are the most visible, its scientific value is also important. In effect, the statewide system functions as a giant laboratory device measuring subsidence and crustal deformation--critically important tasks in this coastal state.
Bedrock in southern Louisiana can be many hundreds of feet down and is essentially inaccessible, meaning that accurate bench marks unaffected by shallow subsidence processes are hard to come by. But accurate bench marks are even more important in this coastal area than they are in states with accessible bedrock because flood control and hurricane protection depend on accurate elevations applied consistently over large regional projects. Subsidence is a pressing concern. Dokka, who has served as the program director at the National Science Foundation, says that due to fault lines underneath the state and the massive amount of mud that moves “like a glacier,” subsidence can occasionally be quite rapid. From 1966 to 1970, for example, part of the New Orleans area sank 1.75 inches annually.
And it’s not just subsidence and monumentation issues that make life difficult for GNSS surveying in the Bayou State. Climate is also a factor. “Much of what we know about GNSS was learned in semiarid climates,” Dokka explains, “which means that humid summer afternoons in Louisiana are not a good time for GNSS surveyors.” That’s because the hot, heavy atmosphere causes more distortion in the GPS signal there than in drier states. The Trimble VRS network solution accounts for regional atmospheric delays encountered by surveyors in these conditions, allowing them to continue working despite the challenging environment.
GULFNet is also contributing to weather prediction by providing data to atmospheric scientists at NOAA’s National Weather Service, who use the data as part of their precipitable water vapor research. “They’re interested in the delay time of satellite signals passing through the atmosphere because moisture increases delay,” Dokka explains. “So if we’ve got a station collecting signals from 10 satellites, that’s like measuring moisture in 10 different directions at once, and they can use that data to improve their atmospheric models.” Then he laughs. “It’s interesting,” he explains. “The signal they use is the noise in my data, but to the weather guys, my signal is their noise.” Ultimately, it’s a win-win because the improved atmospheric models that the National Weather Service derives from the data improve the results GULFNet provides to surveyors.
A System That Comes Through
USACE’s Tommy Bacon checks the elevation at a West Bank pump station using the GULFNet VRS network.
GULFNet is owned and operated by LSU and is made up of 65 GNSS reference stations and two redundant server banks. Access to additional stations provided by the Texas Department of Transportation and the University of Southern Mississippi was arranged in order to provide full coverage in Louisiana. Stations are being added in the northern half of the state to improve coverage, and Dokka says there’s a lot of interest in densifying coastal areas. “But it’s a tough environment as we have to provide power and cellular communications links,” Dokka says, “and sometimes they aren’t available in the middle of nowhere.”
Redundant server banks mean that one bank can fail and users shouldn’t even notice. As a result, system downtime is minimal. Because rural Internet providers and power grids can fail, there are certainly times when individual stations go out, but Dokka says the system as a whole hasn’t had any significant downtime except during hurricanes.
Dokka says he’s proud of what GULFNet has done for Louisiana. “In addition to the flood structure assessments, we’ve also been able to survey evacuation roads in a few months and integrate that work into the National Weather Service’s official storm-surge models,” he says. “That work would have taken years, cost huge amounts of money and been less accurate before GULFNet.”
It isn’t often that survey work saves lives, but the work and planning that goes into GULFNet will surely do just that as planners and emergency personnel prepare for future hurricane seasons with accurate, timely information that simply wouldn’t have been available without the behind-the-scenes work of Dokka and LSU. “The evidence is clear,” Dokka says. “New Orleans and south Louisiana are sinking, and the survival of … these coastal communities depends on our ability to accurately measure the change so that mitigation strategies can be developed. GULFNet helps us develop sound solutions based on accurate measurements we can rely on.”
Sidebar: The Origins of GULFNet
Improved hurricane protection and ecosystems restoration planning require reliable predictions and assessments of future subsidence, and rebuilding after any hurricane requires readily accessible and accurate vertical control. Not only do levees and floodwalls need to be repaired, but planners have to prepare for the next big hurricane by surveying proposed evacuation routes and incorporating dynamic data into storm-surge models.
To take advantage of improved GNSS location technology and to make use of reference points--satellites--that weren’t continually sinking, Roy Dokka, PhD, and the Louisiana State University Center for GeoInformatics began to develop GULFNet in the early 2000s. While initially a simple CORS network, the new network was revolutionary for Louisiana. “We always knew everything was moving all the time,” Dokka says, “but now, at least, we could track it as it was happening.”
An important early decision was to use the best locations possible for reference stations. In Louisiana, that meant large public buildings. “It’s all about the monuments and how you install them,” Dokka says, adding that Navigation Electronics Inc. helped them overcome the challenges of sighting reference stations. “Because we have weak soils and no bedrock to speak of, we can’t just set them anywhere. But some of our buildings are set on piles that go down hundreds of feet--that’s about the best we can do in terms of minimizing movement.” Using mostly public buildings, like schools, ensures access over time.
Dokka has done well: Twenty-one of GULFNet’s reference stations are part of the national CORS network. According to a June 2009 NGS-commissioned report,* this part of the system is valued at $12 million annually.
A flythrough showing preliminary results of the 2009 levee survey, South Lafourche Levee District, courtesy of the LSU Center for GeoInformatics. The horizontal elevations were referenced NAD 83, and the vertical elevations to NAVD 88. Lidar DEM was corrected to NAVD 88.
Craig Dylan Craig Dylan is a freelancer with a land surveying background who specializes in writing for the AEC industry. For more information about Trimble VRS, visit www.trimble.com. More information about GULFNet can be found at c4g.lsu.edu.