Welcome to DU!
The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards.
Join the community:
Create a free account
Support DU (and get rid of ads!):
Become a Star Member
Latest Breaking News
Editorials & Other Articles
General Discussion
The DU Lounge
All Forums
Issue Forums
Culture Forums
Alliance Forums
Region Forums
Support Forums
Help & Search
Science
Related: About this forumThese eight coastal cities sit on America's flood front line, and AI shows why
https://phys.org/news/2026-04-coastal-cities-america-front-line.htmlPaul Arnold, Phys.org

Study area map. (A) 100-mile buffer zones from the US Gulf and Atlantic coastlines. (B) Elevation map of the USGAC. (C) Location of the USGAC in the context of the CONUS. Credit: Dey and Shao, Sci. Adv. 12, eaec2079
New York, New Orleans and Miami are among the eight cities along the US Gulf and Atlantic coasts facing the highest flood risk, according to a new study published in Science Advances. Scientists developed a new AI-driven framework and combined it with historical flood-damage data to not only identify high-risk coastal areas but also pinpoint the underlying factors driving that risk.
Severe flooding events have become increasingly common in the United States, partly due to rising sea levels and intensifying hurricane activity fueled by climate change. In addition to the risk to life, these disasters result in billions of dollars of property and infrastructure damage.
. . .
The results revealed that eight cities face severe threats: New York, NY; Norfolk, VA; Charleston, SC; Jacksonville, FL; Miami, FL; Mobile, AL; New Orleans, LA; and Houston, TX.
New York City and New Orleans were the most vulnerable, though for different reasons. New York has the highest number of people at risk, with 4.75 million exposed under the GFD scenario and 4.40 million in the EFD scenario. This is because massive urban development has replaced natural soil with impervious surfaces such as concrete. Land subsidence further increases the city's vulnerability.
However, nearly 99% of the population and infrastructure of New Orleans are at risk under both scenarios. This is primarily because much of the city sits below sea level, so water cannot drain by gravity.
. . .
Severe flooding events have become increasingly common in the United States, partly due to rising sea levels and intensifying hurricane activity fueled by climate change. In addition to the risk to life, these disasters result in billions of dollars of property and infrastructure damage.
. . .
The results revealed that eight cities face severe threats: New York, NY; Norfolk, VA; Charleston, SC; Jacksonville, FL; Miami, FL; Mobile, AL; New Orleans, LA; and Houston, TX.
New York City and New Orleans were the most vulnerable, though for different reasons. New York has the highest number of people at risk, with 4.75 million exposed under the GFD scenario and 4.40 million in the EFD scenario. This is because massive urban development has replaced natural soil with impervious surfaces such as concrete. Land subsidence further increases the city's vulnerability.
However, nearly 99% of the population and infrastructure of New Orleans are at risk under both scenarios. This is primarily because much of the city sits below sea level, so water cannot drain by gravity.
. . .
2 replies
= new reply since forum marked as read
Highlight:
NoneDon't highlight anything
5 newestHighlight 5 most recent replies
These eight coastal cities sit on America's flood front line, and AI shows why (Original Post)
erronis
4 hrs ago
OP
IMPORTANT NOTE - the type of AI used for this was NOT generative AI, and in no way is an argument
highplainsdem
1 hr ago
#1
Thanks for the clarification. Another way of saying this: "advanced statistical models".
erronis
47 min ago
#2
highplainsdem
(62,664 posts)1. IMPORTANT NOTE - the type of AI used for this was NOT generative AI, and in no way is an argument
for hallucinating generative AI trained on stolen intellectual property.
I need to bring this up because the AI bros behind the genAI industry like to pretend genAI must be encouraged if we want the advances from other types of AI in science and medicine. And that simply isn't true.
From the link in the OP:
https://phys.org/news/2026-04-coastal-cities-america-front-line.html
The pair drew historical records of past flood damage and examined 16 factors. They fed all this data into three different types of AI (Random Forest, Support Vector Machine and Multilayer Perceptron) to identify which cities are at the greatest risk and why.
erronis
(24,154 posts)2. Thanks for the clarification. Another way of saying this: "advanced statistical models".
Stuff that has been around decades.