The U.S. should build a hurricane storm surge barrier

This is the first in a series of posts that will describe a portable hurricane storm surge barrier that I think could--and should--be built.  The barrier would be capable of being deployed to protect approximately 100 miles along the U.S. Gulf Coast or East Coast in a matter of days.  I will eventually outline one possible design for such a system.  My point in outlining one possible design is not to suggest that such a design is the best design, but simply to illustrate that such a system is in no way some science-fiction fantasy that could never be built and deployed.

Why build such a barrier?  In a word, "Katrina."  Or simply, "New Orleans."  But also, "Miami."  "Fort Lauderdale."  "Tampa/St. Petersburg."  "Jacksonville."  "Galveston."  "Virginia Beach."  "Washington DC."  "New York City."

The simple fact is that a major hurricane could strike anywhere along the Gulf Coast (1600 miles) or East Coast north to approximately Boston (another approximately 1600 miles).  Currently, measures are being designed or envisioned that can partially protect single cities (e.g. levees and flood gates for New Orleans, flood gates for New York City).  However, it would make more sense from an economic and engineering standpoint to design something that could protect approximately 100 miles of coastline days before a storm hits, rather than to protect individual cities by designing something that may not be challenged for many decades (during which time better systems could be built).

The damage from hurricane Katrina was virtually all caused by storm surge (rather than wind damage or inland flooding due to rain).  Wikipedia currently estimates that the total economic damage to Louisiana and Mississippi may exceed $150 billion.  That's just a single hurricane, and a single city.  Moving to New York City, it's been estimated that there is an approximately 1-in-4 chance that a Category 3 or hurricane will hit New York City/Long Island in the next 50 years.  Here are storm surge maps calculated by the SLOSH computer model, for various hurricanes hitting New York City and Long Island 

Hurricane Storm Surge maps for NYC, calculated by SLOSH

Roger Pielke Jr. and others have estimated that if the Great Miami hurricane were to hit Miami in 2020, damages would be approximately $500 billion.  (That's half a TRILLION dollars!)

Preview of AGU presentation: The $500 billion hurricane

Clearly, the risk of such economic consequences alone warrants significant thought about how to design and deploy such a hurricane storm surge protection system.

Comments

[GK]

Mark,

Coming back to the 21st Century economic growth topic, I see that your analysis was in 2004, off of data until 2000.

Now, can you do a new analysis that includes World GDP data until 2008 (where GDP projections for 2008 are fairly well defined at this point). The data exists on the IMF website.

The inclusion of more data through 2008 might help refine your forecast. Surely you agree that it is time for an update that includes the 2001-08 data.

[Mark Bahner]

GK,

According to my calculations of November 2, 2005, the number of human brain equivalents (HBEs) added in the year 2007 is only about 1000. And it's only 1 million in 2016. It's not until it it hits 1 billion in 2024 that growth should really shoot through the roof. Or at least somewhere in the 2016 to 2004 period.

http://markbahner.typepad.com/random_thoughts/images/annual_computer_power_additions.JPG

Mark

[Mark Bahner]

Interesting. That made for a very small jpeg. Here's the post:

http://markbahner.typepad.com/random_thoughts/2005/11/why_economic_gr.html

[GK]

Mark,

What I am talking about is taking the 2000-08 data for the linear regression. You previous linear regression was with data only up till 2000.

In other words, does the actual data from 2000-08 change anything?

[GK]

Mark,

What I am talking about is taking the 2000-08 data for the linear regression. You previous linear regression was with data only up till 2000.

In other words, does the actual data from 2000-08 change anything?

[GK]

Mark, BTW, you are taking a CAGR of your growth in HBEs at over 100%. That is too high.

What you should take is 7% (PC growth) X 58% (Moore's Law) = 69% a year.

This is still fast, but much slower than what you have.

Kurzweil glibly says that computing power 'doubles every 12 months'. Actually, it is every 18 months. Big difference. Thus, the 1 billion HBEs will not be in 2024, but rather 2040.

Kurzweil does tend to me right in the *nature* of his predictions, but does tend to overestimate the timing. In his 1999 book, his 2009 predictions are turning out to be too ambitious.