- SPECIAL REPORTS
- THE MAGAZINE
For more than a generation, residents of West Jefferson Parish in Louisiana had taken for granted the series of canals and levees that are supposed to protect them from rising floodwaters following hurricanes. But in 1998, a tropical storm that struck land 300 miles away jarred people from their complacency.
Officials in the parish, many of whom had warned of potential disaster for years, turned to GIS technology to fully assess the flood danger and then generate a graphical illustration for state and federal agencies of what might happen if a storm hit the parish head-on.
Ultimately, the GIS models made such a persuasive case that the parish received both public and private support to build corrective measures to protect the West Bank of the Mississippi from tidal surges related to hurricanes.
Protecting the ParishWest Jefferson Parish is located southwest of downtown New Orleans and straddles the Mississippi River. Running north-south through the western edge of the parish lies the Harvey Canal, an 11-mile manmade channel containing a section of the Gulf Intercoastal Waterway.
Levees are supposed to line both banks of the canal to protect homes and businesses from high water. The west side levee is a solid continuous structure built in 1991 with federal funds by the Army Corps of Engineers. Containment structures on the east side, however, are a hodgepodge of wooden bulkheads, cement sea walls, consolidated earthworks and even an abandoned barge.
Maintenance of these structures is officially the responsibility of property owners there. Most of these back levee owners are industrial businesses involved in construction, fishing and offshore petroleum operations.
“The east side is not really a levee system, and there’s a difference in elevation between the two sides,” says Giuseppe Miserendino, deputy director of the West Jefferson Levee District (WJLD). “Water would spill over the east side of the canal if there were a tidal surge.”
Tidal surges pose serious dangers in that section of Louisiana where many residential and commercial areas reside below sea level. In fact, the entire system of canals, levees and river locks built in the area has been constructed to protect New Orleans and outlying suburbs from those tidal surges on the Mississippi and the Gulf of Mexico following a major storm.
“GIS modeling showed that West Jefferson Parish would fill with water like a bowl all the way [east] to the Mississippi if the Harvey levees were significantly breached,” says Dan Heiken, general manager of Professional Engineering Consulting Corp. (PEC), the Baton Rouge firm contracted to build the GIS.
WJLD, which is responsible for maintaining more than 80 miles of levees on the Mississippi West Bank, had been attempting to convince people of the flood threat since the early 1980s, but its arguments fell on deaf ears. Local government agencies had become more concerned with protecting the area against rain-induced flooding, not hurricane surges.
“There has been a whole generation that has grown up not knowing what it’s like to go without running water or electricity after a storm,” WJLD President Chip Cahill says. “The last big storm to hit was Hurricane Betsy in 1965, and it came when Harvey Canal was at low water. There was no flooding and people seemed to think we were immune.”
Minds were changed when Tropical Storm Frances hit Houston in September 1998. Although the storm made landfall 300 miles away and was rated as less than a 25-year level event, the surge forced water over and through the Harvey levees in a half dozen places. The damage was minimal, but the point was made—the east side levees had to be enhanced.
Building a Levee System GISAt the suggestion of the New Orleans District of Louisiana DOT, the levee district contracted the Biloxi, Miss., office of PEC to survey the east bank of the canal and develop a GIS for asset management and emergency response purposes. PEC utilized the GeoMedia GIS suite of products from Intergraph Mapping and GIS Solutions of Huntsville, Ala., to complete nearly every phase of the GIS project from GPS surveying through final presentation before local politicians.
“Although WJLD didn’t have an existing GIS, they were savvy to what the technology could do for them,” Heiken says. “One of their primary objectives was to have us create a thematic representation of elevation data to show where the levee was inadequate and where the water would go in the event of a 25-, 50- and 100-year storm.”
PEC subcontracted the GPS survey to Automated Measuring Systems (AMS) of Monroe, La. Using the Sokkia GSR-2300 GPS receiver with a Sokkia SDR-33 controller (Sokkia Corporation, Olathe, Kan.) and base station for differential post-processing, AMS collected 375 survey points with an accuracy of 3 mm on the six-mile long east side levee.
The east side levee is not continuous. Rather, it is divided into 137 pieces crossing 78 parcels of private property. GPS points were collected at high and low points on each section. Special attention was taken to survey levee defects, which were defined as pipes or gaps in the structure that would readily allow water to flow through or across to the other side.
“Defects usually occurred at property lines where two different types of levee structures did not join up properly,” Heiken says. “We found 16 of these defects.”
Accurate elevation values were extremely important during GPS collection. The Army Corps of Engineers had included that the levee had to extend at least six feet above sea level to adequately protect the back levee area from storm surge. The survey revealed that 106 of the 137 east side levee sections did not meet the Corps requirement.
“During the survey, we gathered other information to populate the GIS,” Heiken says. “Digital photos were taken of each section and of any defects.”
Additional attributes collected in the survey included the type and condition of each levee section, as well as the location and size of defects. WJLD supplied parcel data containing the name, address and contact details for property owners along the canal’s east side. All of these attributes were compiled in the GIS.
Assessing the Levees with GISFor the data collection and modeling portions of the GIS development, PEC used FieldWorks and InRoads software packages from Bentley Systems of Exton, Pa. Survey points were loaded into FieldWorks, which processed the x, y and z coordinates and drew lines between related points to generate a plan view vector map of the levee system.
PEC obtained a 1985 LaDOT survey of the area and used that data for comparison with the GPS elevation points. They also acquired aerial photos from Jefferson Parish, which were compared to the GPS plan map to ensure that levee boundaries had been precisely mapped. In a few cases, the photo revealed survey errors near a towering bridge, which likely had interfered with precise GPS collection. These errors were corrected with traditional survey techniques.
Once all the GPS points had been processed and corrected, they were imported into InRoads, a civil engineering package typically used for elevation modeling and cut-and-fill calculations for construction of highways, canals and landfills. InRoads generated a DTM showing a vertical elevation profile of the east side levee system.
“Then we made DTMs of the flood water heights estimated by the Corps for 25-, 50- and 100-year storm surges and compared them alongside the levee DTM,” Heiken says. “The results were pretty bad.”
Pretty bad in this case meant the comparison of water levels with levee profile showed that numerous sections of the structure would be breached by high water even during a 25-year storm. Many more would be flooded over if a 50- or 100-year hurricane hit.
Compiling Data in the GISAs the DTMs were being generated, PEC concurrently built the GIS with the GeoMedia Professional development package. This Windows-compatible software was ideal for the Harvey Canal application because it was designed to integrate, process and display multiple types of geographic data in varied formats.
PEC had several different data layers to assemble in the GIS: aerial photos in GeoTiFF, parcel data in ArcView Shape files, a 1985 levee survey in a NAD 27 projection and the new GPS survey in NAD 83. The older survey was especially important to the project because it contained coordinate points for ground elevations, roads and building footprints in the back levee area, which were not surveyed in the 1999 project.
“We wanted to overlay the new and old surveys in the GIS to include buildings and terrain features in our flood assessment,” Heiken says. “In another GIS, this would require rubber sheeting to rectify the two projections, but GeoMedia aligns files on-the-fly and displays them in any coordinate system and projection.”
The aerial photo served as the base map for the GIS, and the ground survey points can be overlaid for visual reference. The GIS enables WJLD users to view where each breach point would be located at various surge levels. In addition, PEC hyperlinked the ground photos to the GIS so that WJLD can view ground level conditions by clicking on any section of the GIS levee map.
InRoads was also used to compute the volume of material that would be required to fill each gap in the levee. The engineers converted these volumetric calculations into numbers of sandbags, and this information is stored as an attribute.
“If we get a call of a levee breach, we can punch a street address into the GIS and instantly see what the area looks like and find out how many sandbags are needed to plug it,” Miserendino says. “More likely though, we would fill those defects in advance as the storm approached.”
Making a Case with GISWith the breach points located on the digital map, PEC used the InRoads hydraulic analysis tools to model the flow of water over and through the structures. The system showed which roads and buildings in the back levee would be inundated as floodwaters rose in various parts of the parish. The model revealed that virtually any serious water breach could flood the entire bowl-shaped topographic depression between the Harvey Canal and the Mississippi River.
Because GeoMedia products are Windows-compatible, the flood models and DTM profiles were exported directly to Microsoft PowerPoint for use in public presentations. WJLD’s Cahill and PEC’s Heiken made the presentation before citizens groups, state legislators and Corps engineers to demonstrate the potential disasters that awaited the parish under certain high water conditions.
WJLD also used the GIS to cross reference levee defect and height inadequacy locations with parcel ownership records. This was crucial information because private property owners are responsible for levee maintenance in the parish. Sixty of the 78 east side owners were found to have insufficient levees.
The district sent hardcopy GIS maps along with a letter to each owner outlining the problem and suggesting how each particular levee section could be repaired to bring it up to Corps standards. WJLD also offered some building material at little or no charge.
“We met with the owners to present the findings and discuss options,” Cahill says. “The GIS was very convincing; most [property owners] rolled up their sleeves and got to work on repairs.”
More importantly, state legislators pledged funding to build a flood control gate at the canal’s north end and a series of cement-capped walls in front of the levees. With matching funds from the federal level, WJLD received more than $40 million to make the improvements it has sought for the last 15 years. An estimated 250,000 people will be protected in the Harvey back levee area once the work is completed.
“The GeoMedia GIS helped us get the funding, but its use doesn’t end there,” Cahill says. “We will continue to update it for use in everyday operations and emergency situations.”
WJLD has contracted PEC to perform another GPS survey on the west side of the canal for inclusion in the GIS. Plans are also underway to link real-time tidal surge sensors in the canal to the GIS and publish this information via the Web so WJLD and Corps technicians can monitor rapidly changing storm situations from remote locations.