Katrina Changed Everything: New Software Predicts How Water Will Spread

Answers to at least some of the problem are now on the way, thanks to a team led by the Department of Homeland Security’s (DHS) Science and Technology Directorate (S&T), and they come in the form of some remarkable computer software.

When dam and levee owners and emergency planners want to know what flood water over a breached levee or dam may do as it spreads, they must resort to technical specialists who use numerical modeling software to solve very complex equations that describe how water will spread over a particular terrain. Through complex equations, specialists calculate how water will move around physical objects such as hills, buildings, vegetation, bridges, and railroads. With such factors in play, calculating and modeling flood inundation caused by a dam failure can take a lot of time and resources and keep emergency planners and dam owners up at night worrying.

When the levees were breached in New Orleans during Hurricane Katrina, the Ninth Ward was inundated. (Photo Credit: FEMA)

Powerful software tools have been combined into a seamless Web application, combining speed with sophisticated technology to visualize a flood, address consequences, and properly train emergency responders.

And this new tool is fast. Really fast. If a flood would take 24 hours to inundate downstream areas, this software tool could potentially model the inundation in less than 24 minutes.

S&T combined the talents of several agency experts and academics to better understand what the owners and operators would need from the software. S&T worked with dam experts at the Office of Infrastructure Protection (which serves as the Dams Sector-Specific Agency) within the DHS’ National Protection and Programs Directorate to develop the flood simulation tool, and with experts at the University of Mississippi — specifically the National Center for Computational Hydroscience and Engineering (UM-NCCHE)’s computational hydroscientist, Mustafa Altinakar, and his team.

This effort was funded by S&T’s Southeast Region Research Initiative (SERRI) and managed by S&T’s Infrastructure Protection and Disaster Management Division’s Mike Matthews. The key component of the project is DSS-WISE™ (Decision Support System for Water Infrastructural Security) and the underlying flood simulator, CCHE2D-FLOOD™, which provides unmatched ‘number-crunching’ speed. The flood simulator can replicate flooding caused by any cataclysm less fateful than The Great Deluge: a breached levee, a failed dam, a surging tide, a tsunami — even water waves caused by massive landslides.

In 2010, when one-fifth of Pakistan’s land was underwater, hydraulic engineers at the U.S. Army Corps of Engineers (USACE) used DSS-WISE™ to help the country reallocate resources. Time was essential, and to achieve its unprecedented speed, the “DSS-WISE™ guys,” as Altinakar affectionately calls them, use several methods to ensure success. First, DSS-WISE™ selectively prioritizes affected regions. It also processes only the model’s “skeleton,” or wireframe, while applying the “skin” afterward. Finally, it divides the flood path model into tens of millions of geometric cells, using parallel processing to parcel them out to separate processors.

The other critical piece of the puzzle is the Dams Sector Analysis Tool, or DSAT. This powerful Web-based application — developed by the Dams Sector-Specific Agency in collaboration with USACE Headquarters’ Office of Homeland Security, who co-sponsored the development of DSAT — is a one-stop shop where dam owners and operators have secure access to state-of-the-art analytical capabilities within a user-friendly graphical environment. Dam owners and operators use algorithms in DSAT to identify and prioritize the most critical dams within their portfolios. Considering that there are more than 84,000 dams across the country, this is no easy task. DSAT also incorporates a state-of-the art geospatial viewer that provides powerful query capabilities as well as access to real-time information (earthquakes, weather, etc.).

The DSAT interface is extremely intuitive and mastered with little training. With DSAT, a dam owner or operator can prepare the input data required for the flood simulation using DSS-WISE™. For example, to characterize a potential dam failure scenario, operators would define the reservoir, identify the main dam, note structures using satellite imagery, and specify the type failure to be considered: a “sudden and complete failure” or a “gradual and partial breaching.” DSAT does the rest — drawing data from the National Inventory of Dams (NID), maintained by the USACE. The data are then bundled into a data file and emailed to a dedicated server at Ole Miss, where the simulation is run. When the simulation ends, the server automatically notifies the user, who may then upload the results on DSAT, where they are rendered onto a map.

“It works similarly to Apple’s Siri,” says Altinakar, referring to the iPhone’s intelligent digital assistant. “There’s no way all that processing could occur in the user’s computer — or phone — so it’s handed off to an external server. It looks simple to the consumer, but I assure you, it’s not.”

The two software systems — DSS-WISE™ and DSAT — are both effective enough to stand on their own, but their integration into a powerful system elevates the capacity for flood simulation. The DSAT geospatial viewer includes a function called DSS-WISE™ Prep. Select your dam on a map, fill in a few facts, direct DSAT how high the reservoir will be when the flood starts, and click Begin. The request is bundled into a data file and automatically sent to the DSS-WISE™ flood simulator. As the simulation unfolds, the consumer will not see heavy activity but will immediately receive automatic progress reports by email.

The DSS-Wise™ Prep module was launched on the DSAT portal on February 20, and days later, it welcomed its first user, delivering results in just 15 minutes. By March, queries poured in from dam owners, state dam safety officials, and emergency managers in seven states — each looking to lower costs, work faster, and make sounder planning decisions.

Like all SERRI projects, flood modeling projects have combined science and technology with validated operational approaches to solve local and regional problems that have a national impact.

Looking back on the tragedy of Katrina, writer Chris Rose ended his 2006 book with a chapter titled, “A New Dawn.” In it he wrote: “Last year ended with everything so unsettled; just a million questions piled up on the curbside like so much debris, the answers just beginning to be formulated in our heads…. It’s just one small step at a time, small triumphs… Who says there’s no good news?”

The powerful software from DHS — now easily available to dam owners and emergency planners — is just some of that good news.

Article Link

Leave a Reply