Donate Bitcoin

Donate Paypal


PeakOil is You

PeakOil is You

Would Someone Kindly Tell Jeffrey Brown...

What's on your mind?
General interest discussions, not necessarily related to depletion.

Would Someone Kindly Tell Jeffrey Brown...

Unread postby PenultimateManStanding » Mon 09 Jun 2008, 11:55:52

I don't post at theoildrum.com but some of you perhaps do. Jeffrey Brown, aka westexas, is well known for his Export Land Model which proposes that the exports of oil exporting counties will fall faster than production declines because of rising internal consumption. That makes sense. But he made a statement today that can only harm his credibility. He should edit it. If you have his email, send him a line for me please.

$this->bbcode_second_pass_quote('westexas', 'B')ased on the HL plot, which assumes approximately a bell-shaped curve (strictly speaking, parabolic I believe), the North Sea peaked when it was about 50% depleted.


I don't know if the Hubbert curve is a classic bell curve, but it is definitely not parabolic. The bell curve is the probability density function, also known as the Gaussian function.

Normal Distribution

Image

Image

The top curve is a parabola, the bottom a normal distribution. Note that even if you make a parabola, say, -x*2 so it goes down instead of up, it doesn't "flare out" asymptotically on the y-axis as the bell curve does.
Turn those Machines back On! - Don Ameche in Trading Places
User avatar
PenultimateManStanding
Expert
Expert
 
Posts: 11363
Joined: Sun 28 Nov 2004, 04:00:00
Location: Neither Here Nor There

Re: Would Someone Kindly Tell Jeffrey Brown...

Unread postby PenultimateManStanding » Mon 09 Jun 2008, 13:22:10

Has no one told Jeff? Leanan? Maybe nobody notices or cares. Still, it makes him look a bit silly with such a high school math gaffe.
Turn those Machines back On! - Don Ameche in Trading Places
User avatar
PenultimateManStanding
Expert
Expert
 
Posts: 11363
Joined: Sun 28 Nov 2004, 04:00:00
Location: Neither Here Nor There

Re: Would Someone Kindly Tell Jeffrey Brown...

Unread postby PenultimateManStanding » Mon 09 Jun 2008, 15:29:56

OK, so I'm talking to myself again. Nothing new. But the Gaussian function is beautiful. I love the purity of it. It says that if you take the inverse of a standard deviation (the inflexion point of the curve) times the radians of a circle and multiply all that times base euler's number exponent negative one half of the square of the ratio of distance from the mean divided by the standard deviation, you get a bell! Gloria Mathematica!
Turn those Machines back On! - Don Ameche in Trading Places
User avatar
PenultimateManStanding
Expert
Expert
 
Posts: 11363
Joined: Sun 28 Nov 2004, 04:00:00
Location: Neither Here Nor There

Re: Would Someone Kindly Tell Jeffrey Brown...

Unread postby dinopello » Mon 09 Jun 2008, 15:59:05

The Normal distribution is nice for modeling worlwide oil production because of the central limit theorum (Lyapunov's). Individual wells or fields may not match very well, but the aggregate comes out as modeled nicely by the Normal.

I think Brown is referring the the (so-called) Hubbert Parabola

Looks like a roll your own distribution to best fit the observed data. My guess would be that most regions are asymetrically produced with the decline having a long tail (production obviously starts abruptly at 0).
User avatar
dinopello
Light Sweet Crude
Light Sweet Crude
 
Posts: 6088
Joined: Fri 13 May 2005, 03:00:00
Location: The Urban Village

Re: Would Someone Kindly Tell Jeffrey Brown...

Unread postby PenultimateManStanding » Mon 09 Jun 2008, 16:26:41

$this->bbcode_second_pass_quote('Shannymara', 'W')ell, maybe people are reluctant to correct his math if they aren't certain about it themselves. Here's his email address for you:

westexasATaolDOTcom
Thanks, Shanny. I sent this message:

We've been discussing your remarks of this morning that Hubbert's bell curve is parabolic at peakoil.com. It's not a major issue but nonetheless I maintain that you must have been a little groggy at that moment because any sort of bell curve is clearly not parabolic. The closest parallel is a Gaussian function. There may be something I'm missing.

Scott, aka penultimatemanstanding.
Turn those Machines back On! - Don Ameche in Trading Places
User avatar
PenultimateManStanding
Expert
Expert
 
Posts: 11363
Joined: Sun 28 Nov 2004, 04:00:00
Location: Neither Here Nor There

Re: Would Someone Kindly Tell Jeffrey Brown...

Unread postby PenultimateManStanding » Mon 09 Jun 2008, 16:40:02

$this->bbcode_second_pass_quote('dinopello', '
')I think Brown is referring the the (so-called) Hubbert Parabola
It's been a while, but I did well in my studies of differential equations. I haven't had occasion to use them. Nonetheless, a statement that a Bell Curve is "strictly speaking" a parabola is clearly wrong.
Turn those Machines back On! - Don Ameche in Trading Places
User avatar
PenultimateManStanding
Expert
Expert
 
Posts: 11363
Joined: Sun 28 Nov 2004, 04:00:00
Location: Neither Here Nor There

Re: Would Someone Kindly Tell Jeffrey Brown...

Unread postby dinopello » Mon 09 Jun 2008, 16:50:07

$this->bbcode_second_pass_quote('PenultimateManStanding', '')$this->bbcode_second_pass_quote('dinopello', '
')I think Brown is referring the the (so-called) Hubbert Parabola
It's been a while, but I did well in my studies of differential equations. I haven't had occasion to use them. Nonetheless, a statement that a Bell Curve is "strictly speaking" a parabola is clearly wrong.


It's cute how you get all worked up about this stuff.

I 100% agree that saying "a Bell Curve is strictly speaking a parabola" is wrong.

I thought he was saying that the production curve can be approximated by a Bell curve, although, in fact the data is better fit by a parabola. The way he constructed that paragraph I can see why your interpretation is reasonable, though.
User avatar
dinopello
Light Sweet Crude
Light Sweet Crude
 
Posts: 6088
Joined: Fri 13 May 2005, 03:00:00
Location: The Urban Village
Top

Re: Would Someone Kindly Tell Jeffrey Brown...

Unread postby PenultimateManStanding » Mon 09 Jun 2008, 17:01:50

$this->bbcode_second_pass_quote('dinopello', '
')It's cute how you get all worked up about this stuff.
It's just something to do. heh heh. But hey! You know I'm right.
Turn those Machines back On! - Don Ameche in Trading Places
User avatar
PenultimateManStanding
Expert
Expert
 
Posts: 11363
Joined: Sun 28 Nov 2004, 04:00:00
Location: Neither Here Nor There
Top

Re: Would Someone Kindly Tell Jeffrey Brown...

Unread postby PenultimateManStanding » Mon 09 Jun 2008, 17:43:05

When I was a little child I loved numbers. I wanted to know at 4 years old what it meant to put a number above a line and have another number under it. If I had it all to do over again, I would have been a mathematician.
Turn those Machines back On! - Don Ameche in Trading Places
User avatar
PenultimateManStanding
Expert
Expert
 
Posts: 11363
Joined: Sun 28 Nov 2004, 04:00:00
Location: Neither Here Nor There

Re: Would Someone Kindly Tell Jeffrey Brown...

Unread postby PenultimateManStanding » Wed 11 Jun 2008, 11:42:50

Jeffrey Brown replied that the parabola was, as dinopello mentioned, a reference to a method to approximate cumulative production under the curve and not the bell curve itself. So his quote
"Based on the HL plot, which assumes approximately a bell-shaped curve (strictly speaking, parabolic I believe), the North Sea peaked when it was about 50% depleted," was a slip-up with English, not math, I suppose. It was nice of him to reply.
Turn those Machines back On! - Don Ameche in Trading Places
User avatar
PenultimateManStanding
Expert
Expert
 
Posts: 11363
Joined: Sun 28 Nov 2004, 04:00:00
Location: Neither Here Nor There


Return to Open Topic Discussion

Who is online

Users browsing this forum: No registered users and 1 guest

cron