Detecting Natural Disasters Before They Happen

Written by Monica Savaglia
Posted July 16, 2018

A few months ago, Ellicott City, Maryland, experienced yet another massive flood. It’d only been two years since water last gushed through the streets of Ellicott City, which ruined homes and businesses in its warpath.

Following the floods, businesses and homes were finally rebuilt. And businesses were opening up their doors again. At least, that was until tragedy struck yet again on May 27th...

After the first storm, the National Weather Service put the timeline of another storm happening again at 1,000 years. After this most recent storm, Jeff Allenby, the director of conservation technology for Chesapeake Conservancy, an environmental group, said, “It’s the second time it’s happened in the last three years.”

Floods in Ellicott City are nothing new because it’s located where two tributaries join the Patapsco River. But the floods have been getting worse because of the development that's happened within the city. What used to be a “natural sponge of a forest” is now paved roads, rooftops, and lawns.

That’s part of the problem, but it’s also the world that we live in now.  

The U.S. Department of Homeland Security (DHS) had selected Ellicott City for a pilot program to deliver better flood warnings to residents via automated sensors because of its 2016 flood. This happened a few days before this year’s May 27th flood.

Natural disasters strike at the most unexpected times and with full force. So, imagine the lives that could be saved by if we were alerted before these disasters strike...

AI Here to Save the Day

Artificial Intelligence (AI) could be the tool and technology to help in making natural disasters… less of disasters.

It might not be possible to completely detect natural disasters before they occur. But AI technology could help in understanding and developing information that’ll send out alerts a lot sooner when a natural disaster is on its way. And this will allow for more people to be alerted and warned so they can take the necessary precautions to avoid the disaster.

And that will allow them to reach safety before destruction strikes.

Allenby has developed a tool that’ll help predict, plan, and prepare for future floods.

This new tool is the first of its kind. It’s a high-resolution map that shows what’s on the ground. This includes a map of buildings, pavement, trees, and lawns that stretch across 100,000 square miles from upstate New York to southern Virginia and into the Chesapeake Bay.

This tool and AI helps to generate an aerial map that shows objects as small as 3 feet square. This type of map is more precise than any other map that flood planners have used in the past.

Microsoft helped Allenby’s team teach its AI Earth algorithms that allow for the technology to identify objects on its own. At first, the technology wasn’t perfect. But it eventually started to improve. And its ability to recognize waterways, trees, fields, roads, and buildings became easier.

Chesapeake Conservancy plans to use AI to refresh these maps more frequently. In the past, it wasn’t easy for flood planners to make these maps. It required a lot of labor and money. But with AI, making these maps will become more accurate, up to date, and easier to create.

This is a groundbreaking achievement. It might not seem like much at the moment. But it’ll help in preventing and understanding land areas in a way that hasn’t been possible before. And that's because it had been hard to have up-to-date maps to study and allow disaster experts to plan ahead.

With Allenby’s work at the conservancy, he said detailed, up-to-date information is paramount when it comes to the design of stormwater-management systems. He said, “Looking at these systems with the power of AI can start to show when a watershed is more likely to flood.”

We often take for granted the work that goes into helping to protect us against and prevent natural disasters. So, when a disaster does strike and human lives and people’s livelihoods are sacrificed, we realize that we could have prevented such an unbelievable tragedy if we'd been better prepared and had more accurate data.

Until next time,

Monica Savaglia Signature Park Avenue Digest

Monica Savaglia
Editorial Director, Park Avenue Digest

Report: 5 Simple Rules
for Investing in a Bear Market