4 ms·

A similar thing has happened on the Mississippi river, which has had '500 year floods' in at least 2011, 2009, 2008, 1993 and 1927. I may be missing a few. Ass

A similar thing has happened on the Mississippi river, which has had '500 year floods' in at least 2011, 2009, 2008, 1993 and 1927. I may be missing a few. Assigning probabilities to rare events is meaningless unless you a) have data over a much longer time period than the recurrence time and b) the mechanism causing flooding is not changing. Neither of these holds in the case of Mississippi flooding, for which we only have about 150 years of observations. Unfortunately, multi-billion-dollar decisions about flood risk are made using these manufactured numbers. For example, in Missouri, land behind a 100-year levee is not required to have flood insurance.

maxerickson 9y ago
Should any land be required to have flood insurance? The NFIP supposedly reduces taxpayer cost (with the shaky risk management you mention), but it currently has huge potential liabilities compared to the premiums it collects (and owes the government a bunch of money from past payouts).
- cat199 9y ago
  Hey we've got a legal mandate to fund insurers now with our 'public private partnership' healthcare system.. pick this up as your issue and you might have a lucrative career in lobbying lined up..
  - matt4077 9y ago
    "Now"... The National Flood Insurance Act of 1968 established the government-backed flood insurance scheme. It has Ronald Reagan's signature. https://en.wikipedia.org/wiki/National_Flood_Insurance_Act_of_1968 https://en.wikipedia.org/wiki/National_Flood_Insurance_Act_o... The idea was and is twofold: first, it was seen as the only possibility to actually provide such insurance because insurance companies (even together) lack the financial strength to be able to insure risks that are statistically extremely dependent. Something like this flood costs dozens of billions, and it would bankrupt the whole industry. Secondly, it's in the national interest to promote such schemes that allow for more areas to be used productively.
    - HillaryBriss 9y ago
      > insurance companies (even together) lack the financial strength high deductible policies and reinsurance are other tools that can be used to help with this problem
cc439 9y ago
"Assigning probabilities to rare events is meaningless unless you a) have data over a much longer time period than the recurrence time" This is probably the most important factor nationwide in the US. There simply aren't written records of the weather in the vast majority of land in North America which cover the past 500 years. I wouldn't even be surprised if more than half the land in NA didn't have meaningful data over 200 years old. "b) the mechanism causing flooding is not changing." This is probably the biggest factor in Houston's extraordinary reoccurence of "500 year" floods. While Harvey's level of rainfall is truly unprecedented in all of American history, the other floods can easily be blamed on the fact that the rapid growth of Houston's sprawl has turned a gargantuan amount of land area into a concrete swimming pool. Like other posters have mentions, undeveloped land is very effective at slowing and absorbing water but a gutter is nothing but an aqueous superhighway. Flooding isn't just caused by the total volume of rainfall, it's caused by the speed with which that volume reaches natural outflows (rivers, creeks, estuaries, etc).

maxerickson 9y ago

Should any land be required to have flood insurance? The NFIP supposedly reduces taxpayer cost (with the shaky risk management you mention), but it currently has huge potential liabilities compared to the premiums it collects (and owes the government a bunch of money from past payouts).

cat199 9y ago
Hey we've got a legal mandate to fund insurers now with our 'public private partnership' healthcare system.. pick this up as your issue and you might have a lucrative career in lobbying lined up..
- matt4077 9y ago
  "Now"... The National Flood Insurance Act of 1968 established the government-backed flood insurance scheme. It has Ronald Reagan's signature. https://en.wikipedia.org/wiki/National_Flood_Insurance_Act_of_1968 https://en.wikipedia.org/wiki/National_Flood_Insurance_Act_o... The idea was and is twofold: first, it was seen as the only possibility to actually provide such insurance because insurance companies (even together) lack the financial strength to be able to insure risks that are statistically extremely dependent. Something like this flood costs dozens of billions, and it would bankrupt the whole industry. Secondly, it's in the national interest to promote such schemes that allow for more areas to be used productively.
  - HillaryBriss 9y ago
    > insurance companies (even together) lack the financial strength high deductible policies and reinsurance are other tools that can be used to help with this problem

cc439 9y ago

"Assigning probabilities to rare events is meaningless unless you a) have data over a much longer time period than the recurrence time" This is probably the most important factor nationwide in the US. There simply aren't written records of the weather in the vast majority of land in North America which cover the past 500 years. I wouldn't even be surprised if more than half the land in NA didn't have meaningful data over 200 years old. "b) the mechanism causing flooding is not changing." This is probably the biggest factor in Houston's extraordinary reoccurence of "500 year" floods. While Harvey's level of rainfall is truly unprecedented in all of American history, the other floods can easily be blamed on the fact that the rapid growth of Houston's sprawl has turned a gargantuan amount of land area into a concrete swimming pool. Like other posters have mentions, undeveloped land is very effective at slowing and absorbing water but a gutter is nothing but an aqueous superhighway. Flooding isn't just caused by the total volume of rainfall, it's caused by the speed with which that volume reaches natural outflows (rivers, creeks, estuaries, etc).