Hurricane Harvey 17

[JedClampett]

Not an engineering disaster, but there's going to be engineering fallout.

 

[IRstuff]

Just to clarify, NOAA has been unfairly blamed for an erroneous combination of statements. What NOAA has said is that the 4-day rainfall record for the continental US was surpassed by Harvey at 49.32", exceeding the previous record of 48.00" in 1978. The US record 4-day rainfall is 52" in Hawaii.

Therefore, the "no one expected..." statement earlier is just random reporter BS.

NOAA's archived rainfall prediction from Friday was for 30-40 inches of rain:

http://www.nhc.noaa.gov/archive/2017/al09/al092017.public.023.shtml

Note also "...CATASTROPHIC FLOODING EXPECTED DUE TO HEAVY RAINFALL AND STORM SURGE..."

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[Spartan5]

zdas04: It most certainly was a 500-year storm, if not greater.

I don't think you know what a 500 year storm is. It does not mean there will only be one every 500 years. It means that in any given year, there is a 1 in 500 chance of it happening. These things are defined. It is not a term randomly grabbed out of the air and slapped on a storm.

Considering that Houston has had 3 500-year storms in a row the past three years (a 1 in 125 million chance), one might think that these things are happening... well, more frequently than they should.

Given that the predicted return periods are based on past climatological observations, it might be safe to assume that the climate is... changing.

 

[dik]

To me, a 500 year storm is one of intensity that should only have a return period of 500 years... it doesn't mean that you can't have two in a week... it's just unlikely.

Dik

 

[zdas04]

Spartan5,
Your definition of a 500 year storm coincides with my definition. But who sets the probabilities? If you say that a storm that raises the water level in the Bayous by more than 20 ft has a one in 500 year probability, and it happens 4 times in a 117 years, then you might wonder if it is more likely that the interpretation of the climate record is flawed.

One thing that the climate observations have not taken into consideration is the change in land use. A 20 ft change in the Bayous in the year 1500 would have been a LOT more water than a 20 ft change today because the volume of the structures that 6.5 million people (in the SMSA) have constructed is huge, and the amount of pavement and elevated roadways is very large, both factors making the same volume of water rise to a much higher level. But the definition hasn't changed. A weather event that shows up in the archaeological record from the time of Christ would have a much smaller impact on the landscape than the same weather event today.

Rather than showing this event as a 500 year storm, which is smaller than at least two and probably three events in the 20th century, as a one in 125 million chance, change the definitions to account for land use and call it a 30 year flood.

And of course the climate is changing, the only constant in the natural world is that all things will try to move towards lower entropy (less organized) and when one thing lowers its entropy it will raise the entropy of something else. Nature is always changing, and the only way to mitigate that change is the input of huge amounts of energy.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[SwinnyGG]

To me, a 500 year storm is one of intensity that should only have a return period of 500 years... it doesn't mean that you can't have two in a week... it's just unlikely.

'500 year storm', '100 year storm', '1,000 year storm', etc are scientifically defined terms- yours/my/Spartan/zdas opinions do not matter with regard to how meteorological data is compiled and reported.

The '500 year' storm nomenclature is also a local one- and doesn't refer to weather patterns on a national or global scale.

For reference, here's a nice Washington Post graphic showing '500 year' events recorded over the last 7 years

https://img.washingtonpost.com/blogs/wonkblog/files/2017/08/500-year-floods.png

 

[msquared48]

Future probabilities are based on past known events.

Should the data on which those probabilities change over time, so should future probabilities.

Every storm event changes the base for the picture.

Mike McCann, PE, SE (WA)

 

[dik]

jgKRI: The intensity for the return period is strictly local. The Atacama desert 500 year return period may be a 2" rainfall.

Mike: Recent rain events will have a direct bearing on future probabilities... it's a moving target... and future 500 year return periods, for some affected areas, will be adjusted upwards.

Dik

 

[Spartan5]

It is not "my" definition of a 500 year storm. It is THE definition.

As for who "sets the probabilities" if you want some light reading on the matter, here is 295 pages that outlines the methodologies and data that is used to define the storms:

http://www.nws.noaa.gov/oh/hdsc/PF_documents/Atlas14_Volume2.pdf

And stop conflating the return period for storms with the return period for floods. The land use has no bearing on the return period for a storm. None. This event, the storm, is not smaller than any storm event in the 20th century. Specifically for Houston, or for anywhere else in the lower 48 for that matter. As for the floods, it would be great if they treated it like a 30 year flood... and actually deigned for it and mitigated the effects of our unfettered consumption of natural resources. Instead we celebrate Houston's haphazard sprawl across hundreds of square miles of land.

Regarding your comments on change, that "the only constant in the natural world is that all things will try to move towards lower entropy"; "all things" do not try to move towards lower entropy. Only energy. Take evolution for instance. Or, if that is against your system of beliefs, shake a jar of sand silt and clay up in some water and let me know if it is more or less organized when it all settles. Yes. Entropy is a thing. But as it pertains to climate change (anthropomorphic global warming), aside from the underlying fundamental law, it has nothing really at all to do with it.

 

[SwinnyGG]

Future probabilities are based on past known events.

Should the data on which those probabilities change over time, so should future probabilities.

Every storm event changes the base for the picture.

This is true.

The other factor at work when we talk about '500 year' events of any kind is that those reclassifications of events, which will inevitably occur, are less valid until the data set is long enough that the true frequency of these events can be properly scaled. In 10 years or 50 years or 100 years, this '500 year' event might actually be determined to be a 100 year or 200 year (or 10 year) event.

In simple terms- we are calling this a '500 year' event, but we don't have 500 years of detailed flood data for the Gulf Coast. We don't actually know this to be a 500 year event- that statistical determination is based on some assumptions.

 

[TGS4]

The determination of whether an event is a 1-in-10, 1-in-100, or 1-in-500 year event needs to consider the long history as well as the short history.

Back in June of 2013, Calgary (Alberta) and area suffered a major flooding event, due to upstream rain combined with said rain melting substantial mountain snow. The prevailing scientific opinions at the time were that an event of this magnitude was at least a 1-in->500 year event. However, a post-facto evaluation of the event and the geological history that the flood event unearthed indicated that it was actually more like a 1-in-30/40 year event (source - see the last paragraph on page 10). It just turned out that the base-time period had been an unusually (but still statistically plausible) dry period.

Combine that with the rampant growth seen in the greater Houston area (Harris County population up 35% from 2000 to 2016(est) - source) and the concurrent land use (zdas04) and a lack of upgrades to the overall flood-management system, due in part to the fractured municipal structure in the region, and you have what we see here.

If we were to call such a rainfall/flood event a 1-in-30 year event, that would change how the engineers would (have to) manage it. Mislabeling it as a 1-in->500 year event is the real "engineering disaster" here.

 

[JStephen]

The discussion above reminds me of an incident I read of in Colorado. Seems like this was back in the 30's or so. Anyway, they built a big dam, and prepared for the lake to form behind for their irrigating pleasure. Said lake didn't form. It turned out, they had monitored rainfall over a 10-year period, but that was an uncommonly wet 10 year period, not to be repeated. Sort of the reverse of the situations above.

 

[dik]

Spartan... it's a similar methology for storms and flood levels... just like snow loading...

We're saying the same thing... "Hydrologists are more likely to say that a 1000-year event has a 1000-year recurrence interval. The United States Geological Survey website further provides details on the procedure:"

This means that there is a 0.1% chance of it occurring in any year...

Dik

 

[zdas04]

I just came across a report of the comments from hydrologists and others involved in this subject. HDSC Comments has some very interesting discussion from actual experts in the field. They are nearly unanimous in decrying the quality of 500 year and 1000 year events, and their impact on the environment. I found their comments quite interesting.

Spartan5, "THE DEFINITION"???? Really???? That science is totally settled, right? I'm glad that JohnRBaker keeps giving you stars for attacking me, but your agenda is showing. I will not participate in turning this discussion of a major storm event in Houston into a global warming discussion. Sorry. My only reason for stepping in is that the hyperbole of arbitrary labels is hindering the discussion instead of facilitating it.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[dik]

DS: Great article... and it points out a common flaw of projecting far beyond normal lifespan...

Just because you can do something mathematically, doesn't mean you should...

Dik

 

[miningman]

Yeah i think TGS4 is close to nailing the crux of the issue here. Like him , I also lived in Calgary in 2013, and the water volumes were certainly impressive. But as the weeks went by, as an engineer, it was disappointing to observe the reductions in the " 1 in x year " speculations. But I suppose we shouldn't be too surprised. most of these were generated by the media or politicians, who aren't exactly reknowned for accuracy or checking sources .If such a projection were requested from a competent engineer, do you really think that in the middle of a major flood event , one really good set of data could be obtained?? Under these conditions, the references to alligators while draining swamps comes to mind. It is only after the immediate crisis has been contained, a certain level of normalcy achieved, that all the data can be reviewed by competent personnel.

 

[Spartan5]

Yes, David. The definition. I take it you read the 295 pages already?

http://www.nws.noaa.gov/oh/hdsc/PF_documents/Atlas14_Volume2.pdf

I'll look through the comments that you provided, but I highly doubt any of them support the position that you are advocating; that the 500-year storm should be called the 30-year storm Because at that point, what does the old 30-year storm become? The 5-year storm? And so on down the line. Surely, in the 295 page document I provided, they show there is enough data to at least determine what the 30-year storm is, no? Like I said, with respect, I think you are out of your element.

 

[zdas04]

Spartan5,
Read the link that I posted above. People that USE these classifications to design watersheds, spillways, set insurance rates, etc, find them to be nonsense. I may be out of my element, but the responders to the request for comments are deeply within their element and find the definition to be worthless. I'll go with that and pass on the your 295 page document.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[Spartan5]

David:
I read the link. Have you even read it? It says at the outset that there was support for both discontinuing the publication of the 1,000-year data, and for continuing to publish it. I'd hardly come to the conclusion that the consensus is that it is "nonsense" after having read it. Most of the comments against continuing to publish them are that they simply don't have the occasion to use it. There are plenty of comments from the professionals that do use it supporting continued publication.

Regardless of that, your document is a non-sequitur because it in no ways supports your contention that the 30-year storm is wrong and should be made the equivalent of the current 500-year storm.

You might actually stand to learn from reading something developed by leading professionals in the field (instead of a series of anecdotes). But don't let me get in the way of your bliss.

EDIT: Again... looking at the document, more than half of it is people supporting the continuation of development and publication of the 1,000 year storm data. That's a far cry from your statement that "They are nearly unanimous in decrying the quality of 500 year and 1000 year events, and their impact on the environment. I found their comments quite interesting." Did you read it?

 

[HamburgerHelper]

I wonder if some definition hijicks is about FEMA not being held to their 100 and 500 year flood plain maps. I was worried Friday night due to not having flood insurance on the two houses I have here in Houston even though it wasn't required for my loans and both are outside of the 500 year flood plains. The closest the water got was about 3-4 inches from coming into the garage. That was during heavy rain. This slow and steady rain I don't think can flood either house. I did lose a car though two years ago that was parked in front of my house ,outside of a 500 year flood plain, but that was due to extremely heavy rain.

 

[3DDave]

"the only constant in the natural world is that all things will try to move towards lower entropy"

The flaw in that statement is that the tendency is towards higher entropy. Any local reversals are at the expense of increased entropy that occurs to produce it.

Is there any cooperation between the meteorologists and the hydrologists and the city planners to determine the rainfall levels that will result in problematic conditions? Giving a 'X years' event seems worthless to me. If it doesn't offer even short term predictions about where people need to leave on a street-by-street basis then it seems to be of only academic interest.