QUERY Inputs for World3-2003 scenarios

Jack Harich <jack@thwink.org&
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

QUERY Inputs for World3-2003 scenarios

Post by Jack Harich <jack@thwink.org& »

Posted by Jack Harich <jack@thwink.org>

I’ll be extending the LTG World3-2003 model as part of an analysis. The
first step is to identify the nodes that vary with the scenarios, find
the values for each node for each scenario, and then setup the model so
the scenarios can be easily run.

Vensim ships with the model in VMF format, with 13 CIN files. One would
expect the CIN files to agree with the third edition of Limits to
Growth. But unless I’m missing something, they do not.

The model is set for scenario 1. For scenario 2, I’d expect the
W303S02.CIN file to have the two parameter changes described on page 172
in the caption to figure 4-12, in the 2004 edition of LTG. But only the
first parameter is in the file. The parameter to ""postulate that
advances in resource extraction technologies are capable of postponing
the onset of increasing extraction costs, industry can grow 20 years
longer"" is missing. Running the model with only the first parameter
changed fails to duplicate the graphs in figure 4-12 for scenario 2.
Looking in the W303S03.CIN, W303S04.CIN and W303S05.CIN files, the
missing parameter is not there either.

But it is in Harmut Bossel’s ""System Zoo 3 Simulation Models"", 2007.
Page 179 has a beautiful table of 20 parameters for all 10 scenarios.
The missing parameter is named ""investment in obtaining resources time.""
Its value changes from 4000 in scenario 1 to 2002 in scenario 2. This
makes sense.

But guess what? ""investment in obtaining resources time"" is not in the
model file! Looking on page 194 of Bossel, we see the node right at the
top of the page. But looking in the Vensim model file, on the
nonrenewable resources page, we see A DIFFERENT VARIABLE!!! The name has
been changed to ""fraction of capital allocated to obtaining resources 1
table"", and the type has been changed from auxiliary to lookup. The two
models are different! What’s going on here? Are there multiple versions
of World3-2003?

Once I realized this discrepancy, I thought it was time to stop modeling
and consult my fellow modelers. Does anyone know if a correct, complete
table of scenario parameters exists for World3-2003, the version that
ships with Vensim? Or alternatively, why do the CIN files that ship with
it seem to disagree with the published model? Do the CIN file numbers
correspond with the 10 published model scenarios?

If so, then why are there 13 CIN files? Perhaps these are the unnumbered
experiments discussed after scenario 10? If so, the CIN files 9 and 11
do not agree with the book’s description of what might be pseudo
scenario 11 on pages 250 to 251. ""industrial output per capita desired""
is the same in both files. It’s not doubled, as the book describes.

Thanks from a perplexed modeler,
Posted by Jack Harich <jack@thwink.org>
posting date Sun, 30 Mar 2008 19:21:35 -0400

Jack Harich <jack@thwink.org&
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

QUERY Inputs for World3-2003 scenarios

Post by Jack Harich <jack@thwink.org& »

Posted by Jack Harich <jack@thwink.org>

Thanks Bob. I had noticed this morning that the 2nd CIN file reproduced
scenario 2 of World3-93 perfectly, and that 2nd edition book had 13
scenarios. So, mystery solved.

Here's a puzzle: Why hasn't someone noticed this before? Are there
really that few efforts to examine World3-2003 and run it around the
block? Is this yet another symptom of falling interest in SD? Or perhaps
just falling interest in World3?

Looking forward to the correct CIN files. Thanks for making this happen.

Posted by Jack Harich <jack@thwink.org>
posting date Mon, 31 Mar 2008 21:52:07 -0400

Bob Eberlein <bob@vensim.com&
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

QUERY Inputs for World3-2003 scenarios

Post by Bob Eberlein <bob@vensim.com& »

Posted by Bob Eberlein <bob@vensim.com>

Hi Jack,

The change files that comes with Vensim for the 03 World3 model actually
correspond to the 93 version of the model (and book) - perhaps some are
even the original book. There are 13 of them corresponding to the 13
alternative scenarios in the earlier books.

I have to admit just from reading the most recent book I am not sure
exactly what additional adjustments were made to the scenarios. Tom
Fiddaman has a query out to Jorgen Randers to see if we can get the
changes to make things line up. When that is available I will post the
changes files on our web site.

As to the question of why nobody has noticed before, I am not completely sure.
It is posisble, of course, that many people have noticed but not brought it
to my attention. It is also possible that those reading the book and wishing
to follow up have contacted the authors and received the supporting material
in Stella format and so have not looked at the Vensim version. It may also
simply be that the readers of the book either find the results so clear they
don't feel the need to replicate them or so unbelievable they don't see the
need to. The first edition of the book both brought a lot of attention, and
made up a lot of minds and people are stubborn creatures.

Bob Eberlein
Posted by Bob Eberlein <bob@vensim.com>
posting date Mon, 31 Mar 2008 20:45:01 -0400