And now you see how ridiculous this "Ike Dike" idea was. It would have protected nothing. The problem was and is is still the massive rain falls, not the incoming storm surge.
Energy industry could easily build such a "Ike Dike" but why should the city of Houston do such a thing? Bush lobbied strongly for it on behalf of the energy guys, but didn't get it. Now you see why.
I have not seen any evidence in journals that climate change will inherently reduce predictability of short or medium range weather forecasts. I believe you're implying that without climate change we could somehow be predicting hurricanes years ahead of time in the current year, which is really not the case.
> ... climate change will [not] inherently reduce predictability of short or medium range weather forecasts
In the broad sense that's totally correct: weather will still be weather, and the sensor data & computing resources we have to throw at prediction are only getting better.
One aspect of this specific conversation relating to climate change, though, would be the relative abruptness of multiple weather systems combining into much larger storms. Sandy, for example, became a Frankenstorm in part due to multiple slightly abnormal factors coming together. That trend should erode predictability across multiple axes.
On the order of centuries, compared with the otherwise ability to predict weather: I would expect to see a marginal decline in weather predictability as more events, and more extreme events, are added to the atmospheric system.
If you read the article you will see that it entirely failed to predict what just happened in Houston. The article goes on and on about Storm Surges, and building Dikes - neither of which had anything to do with what just happened.
ghshepard, true. The article focused on seawalls and surges, but still managed to get the damage correct: 600,000 residents effected, $100 Billion in damage, 145 mph winds. Regardless, knowing the effects of the storm -shown through the simulation - there could have been a plan for the ensuing catastrophe, e.g. evacuating the city, relief efforts; which seem ad hoc.
Thanks for posting; it was very informative. I'd been looking for a succinct explainer of the what and how of the flood disaster but kept running into situation updates instead.
That was a painful way to read that informative content. Here it is in a more readable form, links at the end, abbreviations like 'HTX' replaced with the complete words, sentences and paragraphs consolidated. Inline links might be nice, but HN-formatting works a whole lot better than this ridiculous "twitstream". Links are in my comment below: https://news.ycombinator.com/item?id=15124561 because the whole thing was too large to fit in one HN comment. Much less a 140-character tweet.
A Houston floodsplainer, by Matt Corbett [caveat]
There will inevitably be extreme hottakes regarding flood planning and monday-morning quarterbacking of officials. This is for context. Some good links: [1] [2] [3] [4].
Houston is on a flat, mostly featureless plain, which is naturally drained by a number of Bayous. ("The Bayou City" refers to Houston, not New Orleans.) The bayous all run (and drain) from west to east, converging on either the ship channel or San Jacinto Bay.
It also has varying development density. Here's a satellite photo which will roughly show that density: [Sat pic] (Note: I've highlighted 2 areas- Addicks & Barker reservoirs and the medical center, because I'll mention them later).
Houston has sandy soil and a high water table, and so has some limited ability to rely on absorption. (Related: No houses have basements and it would be nearly impossible to construct a subway.) Most of Houston is ~35-45'/12m above sea level.
Flooding risk is almost entirely from rain, not storm surges. Being on the Gulf Coast, Houston gets ~50" of rain a year. Gulf thunderstorms can get intense. Four-to-six-inches-in-8-hours storms happen about once a year. Flooding is essentially a rate problem: can you drain the water as fast as it comes? When the answer is 'no', water backs up along the drainage routes. [Sat pic with drainage routes] As a result, any person's flooding risk is mainly about proximity and elevation vs the nearest bayou.
The primary backups for the bayous handling too much water are the roads. In the 90s, Houston was getting large enough that relying on groundwater was starting to cause subsidence problems. The powers that be decided (wisely, mostly) to slowly convert all the roads into a giant rain collection network, so every time an asphalt road needed to be repaved, it got replaced with curb & gutter concrete with a big storm sewer underneath. This has been highly obnoxious to anyone living nearby when such a project was underway, but ultimately quite effective. It usually means that in flooding situations, roads briefly become rivers and then drain, saving houses from flood damage, but it's also a work in progress that has proceeded at the rate roads needed replacing, and it varies greatly by location.
The next backup for water are sections of freeways. Here, for example, is a section of I-69/US-59 (as [indicated on map].) Given the flood risk indication of the neighborhood immediately south, that sunken section serves a flood-relief purpose: [Flooded freeway]
Thus, flood control in Houston is and has been in a continual state of upgrade for 20 years. However, Houston has also been growing rapidly in that time, adding about 100-125 thousand people per year for 15 years. The result of this growth is that at any given time the flood control has been adequate, but only for the city T-5 years ago, not now. The currently least-adequate parts are usually around the geographic periphery and immediately downstream.
The key incidents forming city officials' decision making have been the experiences of Allison (2001), Rita (2005), Ike (2008) and the flooding events of the past 2 years (Memorial Day 2015 and Tax Day 2016).
Conceptually, Harvey is closest to Allison, which was a tropical storm that parked itself over Houston for 3 days and dumped 20" of rain [Tropical Storm Allison wiki]...
The debate of whether Houston should have issued a mandatory evacuation is more
complicated than many probably realize. In 2005, the evacuation of Hurricane Rita was a bigger calamity than the hurricane itself: https://apps.texastribune.org/road-from-rita/taking-on-traff.... For days, major highways looked like parking lots. Dozens died before Rita even reached Texas.
> Of the 139 deaths that the state linked to Hurricane Rita, 73 occurred before Rita reached Texas. Twenty-three people died in a bus fire. Ten others died from hyperthermia due to heat exposure.
It was clear in hindsight that many evacuees would have been better off riding out the storm at home. Texas took measures to improve future evacuations but state officials admitted to me in 2015 that they weren't enough. 3 years later, the Texas evacuation ahead of Hurricane Ike went much better. But that's partly because of this:
> The Ike evacuation was also aided by tens of thousands of residents refusing orders to leave. For many, that decision was driven by the still-fresh memory of Rita.
Texas has grown like gangbusters since Rita. Growth in highway capacity hasn't come close to matching that.
> Between 2005 and 2014, the population of a 40-county area covering southeast Texas and the Gulf Coast grew by 20 percent, or more than 1.5 million people. Over the same period, highway lane miles in that region grew by just 5 percent, or 1,707 miles, according to state data.
Houston is the 4th largest city in the country with 2.3 million people. So as Houston Mayor debated evacuation, he had to weigh whether he was directing millions to sit in traffic as Harvey reached landfall, and also whether he might be creating thousands of new people who refuse to evacuate ever again. This is not a defense of Sylvester Turner's decision not to evacuate Houston. Just offering context of how complicated the decision was.
Prepared or not, I think there is some element of it being really hard to get funding to go fix things proactively vs. tapping into a Federal organization which will fund the aftermath.
I wonder if there have been studies comparing the costs? The hurricane is coming either way and there will be a need to cleanup.
Its the same in some software organizations, fixing things so they are right takes a lot of convincing vs. band-aids when the problem comes up.
Hey HN, thank you to Federal taxpayers, who funded matching grants to bayou and watershed changes that, even in an incomplete state, have saved many houses and people. The article focuses on an incomplete big set piece that if fully funded would have likely starved the myriad smaller improvements that have made life better for a large number of people and done nothing for this week's disaster.
Everywhere I go I see the things people have done to mitigate weather events:
* commercial buildings with high water dams
* freeway underpass pumping stations
* new suburban neighborhoods and commercial centers with large water retention zones that become parks and playing fields in the 99% of the time they aren't underwater
* infill housing in existing neighborhoods that is raised on piers to protect people without displacing water needlessly
* regional sized water retention ponds for established neighborhoods that have become wetlands for migratory waterfowl and local wildlife.
Is there a lot left to be done? Sure. Could what is being done be done better? Sure, we learn all the time. I think learning is the reason why the current semi-organic approach is sound. The glass may be half empty or half full, but the fill rate is moving in a direction of increased storm resilience (which this week is toward empty for sure ).
23 comments
[ 2.9 ms ] story [ 63.5 ms ] threadIt is also linked from the article, but embarrassingly took me a minute to find it.
Energy industry could easily build such a "Ike Dike" but why should the city of Houston do such a thing? Bush lobbied strongly for it on behalf of the energy guys, but didn't get it. Now you see why.
In the broad sense that's totally correct: weather will still be weather, and the sensor data & computing resources we have to throw at prediction are only getting better.
One aspect of this specific conversation relating to climate change, though, would be the relative abruptness of multiple weather systems combining into much larger storms. Sandy, for example, became a Frankenstorm in part due to multiple slightly abnormal factors coming together. That trend should erode predictability across multiple axes.
On the order of centuries, compared with the otherwise ability to predict weather: I would expect to see a marginal decline in weather predictability as more events, and more extreme events, are added to the atmospheric system.
https://twitter.com/CorbettMatt/status/901959336850804737
A Houston floodsplainer, by Matt Corbett [caveat]
There will inevitably be extreme hottakes regarding flood planning and monday-morning quarterbacking of officials. This is for context. Some good links: [1] [2] [3] [4].
Houston is on a flat, mostly featureless plain, which is naturally drained by a number of Bayous. ("The Bayou City" refers to Houston, not New Orleans.) The bayous all run (and drain) from west to east, converging on either the ship channel or San Jacinto Bay.
It also has varying development density. Here's a satellite photo which will roughly show that density: [Sat pic] (Note: I've highlighted 2 areas- Addicks & Barker reservoirs and the medical center, because I'll mention them later).
Houston has sandy soil and a high water table, and so has some limited ability to rely on absorption. (Related: No houses have basements and it would be nearly impossible to construct a subway.) Most of Houston is ~35-45'/12m above sea level.
Flooding risk is almost entirely from rain, not storm surges. Being on the Gulf Coast, Houston gets ~50" of rain a year. Gulf thunderstorms can get intense. Four-to-six-inches-in-8-hours storms happen about once a year. Flooding is essentially a rate problem: can you drain the water as fast as it comes? When the answer is 'no', water backs up along the drainage routes. [Sat pic with drainage routes] As a result, any person's flooding risk is mainly about proximity and elevation vs the nearest bayou.
The primary backups for the bayous handling too much water are the roads. In the 90s, Houston was getting large enough that relying on groundwater was starting to cause subsidence problems. The powers that be decided (wisely, mostly) to slowly convert all the roads into a giant rain collection network, so every time an asphalt road needed to be repaved, it got replaced with curb & gutter concrete with a big storm sewer underneath. This has been highly obnoxious to anyone living nearby when such a project was underway, but ultimately quite effective. It usually means that in flooding situations, roads briefly become rivers and then drain, saving houses from flood damage, but it's also a work in progress that has proceeded at the rate roads needed replacing, and it varies greatly by location.
The next backup for water are sections of freeways. Here, for example, is a section of I-69/US-59 (as [indicated on map].) Given the flood risk indication of the neighborhood immediately south, that sunken section serves a flood-relief purpose: [Flooded freeway]
Thus, flood control in Houston is and has been in a continual state of upgrade for 20 years. However, Houston has also been growing rapidly in that time, adding about 100-125 thousand people per year for 15 years. The result of this growth is that at any given time the flood control has been adequate, but only for the city T-5 years ago, not now. The currently least-adequate parts are usually around the geographic periphery and immediately downstream.
The key incidents forming city officials' decision making have been the experiences of Allison (2001), Rita (2005), Ike (2008) and the flooding events of the past 2 years (Memorial Day 2015 and Tax Day 2016).
Conceptually, Harvey is closest to Allison, which was a tropical storm that parked itself over Houston for 3 days and dumped 20" of rain [Tropical Storm Allison wiki]...
[Good link 1] https://www.harriscountyfws.org/
[Good link 2] https://spacecityweather.com/
[Good link 3] https://www.texastribune.org/boomtown-floodtown/
[Good link 4] http://traffic.houstontranstar.org/cctv/transtar/
[Sat pic] https://pbs.twimg.com/media/DIRoKVNVYAAuzPc.jpg
[Sat pic with drainage routes] https://pbs.twimg.com/media/DIRokT7UwAEr53H.jpg
[Freeway indicated on map] https://pbs.twimg.com/media/DIRpK4-V4AAephw.jpg
[Flooded freeway] https://pbs.twimg.com/media/DIRpUBsVYAA2Gld.jpg
[Tropical Storm Allison wiki] https://en.wikipedia.org/wiki/Tropical_Storm_Allison
[Aman Batheja tweetstorm] https://twitter.com/amanbatheja/status/901867083608285184:
The debate of whether Houston should have issued a mandatory evacuation is more complicated than many probably realize. In 2005, the evacuation of Hurricane Rita was a bigger calamity than the hurricane itself: https://apps.texastribune.org/road-from-rita/taking-on-traff.... For days, major highways looked like parking lots. Dozens died before Rita even reached Texas.
> Of the 139 deaths that the state linked to Hurricane Rita, 73 occurred before Rita reached Texas. Twenty-three people died in a bus fire. Ten others died from hyperthermia due to heat exposure.
It was clear in hindsight that many evacuees would have been better off riding out the storm at home. Texas took measures to improve future evacuations but state officials admitted to me in 2015 that they weren't enough. 3 years later, the Texas evacuation ahead of Hurricane Ike went much better. But that's partly because of this:
> The Ike evacuation was also aided by tens of thousands of residents refusing orders to leave. For many, that decision was driven by the still-fresh memory of Rita.
Texas has grown like gangbusters since Rita. Growth in highway capacity hasn't come close to matching that.
> Between 2005 and 2014, the population of a 40-county area covering southeast Texas and the Gulf Coast grew by 20 percent, or more than 1.5 million people. Over the same period, highway lane miles in that region grew by just 5 percent, or 1,707 miles, according to state data.
Houston is the 4th largest city in the country with 2.3 million people. So as Houston Mayor debated evacuation, he had to weigh whether he was directing millions to sit in traffic as Harvey reached landfall, and also whether he might be creating thousands of new people who refuse to evacuate ever again. This is not a defense of Sylvester Turner's decision not to evacuate Houston. Just offering context of how complicated the decision was.
[Addicks & Barker reser...
California ill-prepared for the Big One, experts say https://phys.org/news/2016-07-california-ill-prepared-big-ex...
What Would Happen If Yellowstone's Supervolcano Erupted? https://www.livescience.com/20714-yellowstone-supervolcano-e...
America’s Crumbling Dams Are A Disaster Waiting To Happen http://www.huffingtonpost.com.au/entry/america-crumbling-dam...
Scientists say Seattle’s overdue earthquake could happen within 50 years http://mynorthwest.com/12944/scientists-say-seattles-overdue...
Ancient aliens are hibernating on Earth – and they are waiting to ATTACK - shock claims http://www.dailystar.co.uk/news/latest-news/618941/ancient-a...
There's nothing particularly interesting about old articles that appear to have predicted current events.
When it happens, someone will dig up the old article saying "are we ready for the floobledasch EVENT?!!!"
There is, if:
- they had valid grounds for making the prediction
- something could have been done to better prepare for the event
and
- the warnings were recklessly ignored.
I wonder if there have been studies comparing the costs? The hurricane is coming either way and there will be a need to cleanup.
Its the same in some software organizations, fixing things so they are right takes a lot of convincing vs. band-aids when the problem comes up.
Maybe if we keep hammering on about how people predicted this stuff the next prediction will get listened to a bit mroe.
Everywhere I go I see the things people have done to mitigate weather events: * commercial buildings with high water dams * freeway underpass pumping stations * new suburban neighborhoods and commercial centers with large water retention zones that become parks and playing fields in the 99% of the time they aren't underwater * infill housing in existing neighborhoods that is raised on piers to protect people without displacing water needlessly * regional sized water retention ponds for established neighborhoods that have become wetlands for migratory waterfowl and local wildlife.
Is there a lot left to be done? Sure. Could what is being done be done better? Sure, we learn all the time. I think learning is the reason why the current semi-organic approach is sound. The glass may be half empty or half full, but the fill rate is moving in a direction of increased storm resilience (which this week is toward empty for sure ).