Earth Imaging Journal

Home

About Us

Articles

Press Releases

Image Gallery

Media Kit

Free Subscription

Industry Resources

Industry Directory

NASA Links

Missions/Launches

Calendar



	By Maarten Bomers and Robert Hoddenbach, Fugro-Inpark (www.fugro.nl), Leidschendam, The Netherlands.

	Water management is critical to The Netherlands. More than half of the country’s 16,000 square miles rely on artificial barriers to keep the country dry and therefore habitable for the 16.5 million people who live there. Without water embankments, locks, pumping stations and canals, more than the half of the country would be susceptible to flooding. The task of water management falls to two authorities, the decentralized water boards (waterschappen) and the directorate-general for public works and water management (rijkswaterstaat). The former oversees flood embankments at local and regional levels; the latter controls embankments of the major rivers and canals, Lake IJssel, the North Sea and the Wadden Sea. The Netherlands boasts some of the world’s largest storm surge barriers. These engineering feats, however, were constructed in the wake of a large-scale disaster. In the early morning of Feb. 1, 1953, the South West province of Zeeland was hit by massive flooding due to the combination of a spring tide and a severe storm that lasted more than 20 hours. Approximately 700 square miles were flooded, 75,000 people were forced from their homes and more than 1,800 people lost their lives. In response, the Dutch government immediately launched the Delta Act, which freed the way for the construction of a major storm surge infrastructure. Approximately 700 square miles were flooded, 75,000 people were forced from their homes and more than 1,800 people lost their lives. In response, the Dutch government immediately launched the Delta Act, which freed the way for the construction of a major storm surge infrastructure. More than 50 years later, Zeeland is still considered particularly vulnerable and continues to receive special attention and enormous financial budgets to help protect the land and people from the sea. In March and April of this year, Netherlands-based Fugro-Inpark used its helicopter-mounted FLI-MAP mapping system (www.flimap.nl or www.flimap.com) to conduct a 450-square-mile light detection and ranging (LiDAR) pilot study for the largest water board in the province (see "About the FLI-MAP System", below ). Beset with challenging project specifications, the goal of the pilot was to investigate the quality and usability of high-density, high-accuracy LiDAR elevation datasets for water management studies, including controlling the heights and widths of the 600 miles of levees and calculating the capacities of water basins, etc.


	Data Acquisition The water board in Zeeland has an ongoing need for accurate, up–to-date information on elevation. Until recently, the board used elevation data derived from the nationwide Actual Height Model of The Netherlands (AHN). For the sensitive area of Zeeland, however, it was determined that this elevation dataset didn't meet the density and accuracy requirements needed for the water management studies. Flood risk management assets protect thousands of homes and businesses from potential coastal and fluvial flooding. Therefore, accurate survey information of these assets is vital to the water boards' approach, both for the daily asset management and to the decision-making process for asset improvement and replacement. With conventional, lower density fixed-wing LiDAR data, there are difficulties in obtaining an adequate representation of important hydraulic features such as levee crests, flood defense walls and outfalls. Low-level LiDAR, with its much higher point density, would enable the water board to identify and evaluate these smaller but critical features. Fugro-Inpark used two FLI-MAP 400 systems mounted on Eurocopter AS 350 helicopters to collect high density and high accuracy LiDAR data, still imagery, video imagery and line scan imagery. Automatically adding natural colors to the laser points, the line scan imagery gives the whole point cloud an extremely realistic character. The survey was conducted at a height of 1,200 feet above ground level. To cover the complete area, approximately 2,700 miles of flight lines were needed. In total, 20 billion LiDAR points and more than 100,000 still images were collected. Project Deliverabless The difference in specifications between the existing AHN elevation data set and the new FLI-MAP-derived "AHN-2" datasets are dramatic (see "Specification Differences" chart, below).


	Data collection lasted 12 days. Ground crews established eight Global Positioning System (GPS) base stations throughout the area of interest to differentially correct the airborne elevation data. In addition to the GPS base stations, one of the two active GPS networks of The Netherlands was used to calculate Virtual Reference Station wherever needed. In general, GPS baselines were no longer than 15 miles. At the end of each acquisition flight, the LiDAR processor made an initial quality check of the data on a personal computer to ensure there were no gaps in the flight lines and whether the minimum point density of 10 points/square meter was fulfilled for the complete flight. The data then were shipped to Fugro-Inpark's Leidschendam production office for further processing and quality control. Note that the point density is increased by a factor of 160! The height grids needed to be delivered as digital elevation models—with vegetation and buildings left in the data set—and as bare-earth elevation models. There's a marked difference between the dataset consisting of the raw point cloud and the derived grids in the amount of processing performed and the detail that can be abstracted. The point cloud data have had minimal processing and as a result still include some outlier results from birds or system noise. The grid data have had greater processing with outliers and noise removed. Although it's much faster to used gridded data in analysis and modeling, such data lack the details of the raw point cloud data. Besides the grids, the most important products that were delivered were:: • Raw laser data • Filtered laser data (vegetation and other objects removed) • Line scan imagery • All videos (forward facing and nadir facing) • All still imagery (forward facing and nadir facing) • Quality control results • Topographic mapping test results



	Topographic Mappingg The water board, once confronted with the first results, reasoned it might be possible to use the data for extracting breaklines from the dataset. The most important breaklines for water management engineers are the levee crown lines and toe lines. In addition to the breaklines, the water board was interested in extracting cross profiles of the levees. The FLI-MAP team had performed breakline extraction and profiles generation for years, but this has always been done based on dense datasets that only covered the corridor of the levee and its immediate surroundings. For the first time, the data were extracted from an area-covering dataset. This new development clearly shows that the area-wide dataset can be used for hydraulic modeling and for accurately analyzing the geometry of the levees. The images of the laser data show a repeating rainbow pattern. The exact color of each laser point is determined by its elevation. Typically a rainbow pattern that repeats itself each 10 inches gives clear overviews of terrain morphology. As with contour lines, the closer the colors are together, the steeper the terrain. A breakline may be extracted at locations where there is a sudden change in distance between the different colors of the rainbow pattern. Semi-automatic routines are used to extract these breaklines. The breaklines and the cross profiles extracted from the FLI-MAP data were quality controlled by comparing them with terrestrially surveyed breaklines and profiles. Quality control was undertaken by the water board, and the results were impressive: the average height difference between the breaklines and the profiles based on terrestrial measurements and FLI-MAP LiDAR data was less than 0.5 inches, which means that the absolute error of the FLI-MAP data is, according to this test, negligible.



	Ongoing Development According to Willem Rijn, the project manager at the water board in Zeeland, the pilot project results were better than expected. "We have never had such detailed information of our complete area and our levees," he says. "This will give us more possibilities then I could ever hope for." Historically, Dutch water boards have separated their hydrologists from the levee experts. For the first time in history, both groups of specialists are using a single data set for their studies. Although the future can't be predicted, the pilot project is part of a bigger picture. After the project's success, it has become almost certain that soon all of The Netherlands (16,000 square miles) will be surveyed according to the specs of the Zeeland project. Part of this requirement will be fulfilled in the coming months, when 2,300 square miles in the northwest of the country will be surveyed to acquire areawide high-density, high-accuracy data. It shows that the Dutch have realized that dense and accurate areawide data on elevation are of utmost importance for the ongoing battle against the sea. Authors' Note: In the United States, the FLI-MAP systems are operated by John Chance Land Surveys (www.jclsurvey.com), also part of the Fugro group of companies. In the aftermath of Hurricane Katrina, FLI-MAP was contracted by the U.S. Army Corps of Engineers to conduct damage assessment surveys on more than 400 miles of levees around New Orleans.



See more Featured Articles See Featured Images Subscribe to Earth Imaging Journal Go to Home Page

[none]