All:
I would like to see a specific case example of ". . . a model that is
capable of accurate predictions over some limited time scale" so I can begin
to understand the real power of models to demonstrate that they generate
less error than a disciplined mind. I am not implying that I do not "believe
in" models; I mean exactly what I said. An explanation of the input,
processing, and output also would help. In other words, can someone
"demystify" ecological models for me in the spirit of advancing this
particular discussion?
John, what "context" are you talking about? What, exactly, makes you cringe?
I like the idea of being explicit, with or without models. Why shouldn't
verbal hypotheses be just as explicit as those inserted into a computer
program?
WT
----- Original Message -----
From: "Christopher Brooks" <[email protected]>
To: <[email protected]>
Sent: Thursday, January 22, 2009 4:43 AM
Subject: [ECOLOG-L] ecological modeling
John -
Whenever I see the phrase "ecological modeling" in this kind of context I
cringe a little. What is typically meant is to generate some sort of
dynamical model (as opposed to a statistical model) with a goal of
projecting something (often population sizes) into the future. It is this
approach that is often attacked by empiricists as lacking the complexity
of
biological systems, etc. While one can often construct a model that is
capable of accurate predictions over some limited time scale, the
inability
to incorporate complete biological "reality" will necessarily decrease the
predictive power of such a model through time.
The most useful application of mathematical or simulation models, in my
opinion, are those that seek to represent one or more hypotheses (which
are,
of course, models themselves) about the mechanisms behind some
observation.
In this approach, the focus is on the construction and analysis of the
model. In the construction phase, one is forced to be explicit about the
assumptions they make (something that is not necessary with "verbal"
hypotheses). In the analysis phase, it is possible to generate predictions
that are dependent on the assumptions of the model, and/or to determine
the
sensitivity of the system to particular components of the system.
I would suggest that you seek a course that teaches you the tools of model
building (programming, differential equations, matrix models, etc.), but
emphasizes model analysis and the use of models as quantitative
hypotheses.
The Santa Fe Institute has a wonderful summer program that might be one to
look at for starters, but it is quite selective.
--
Dr. Christopher P. Brooks
Assistant Professor
Department of Biological Sciences
Mississippi State University
P.O. Box GY
Mississippi State, MS 39762
E-mail: [email protected]
Phone: 662.325.8591
Fax: 662.325.7939
http://www.msstate.edu/dept/biosciences/brooks/
--------------------------------------------------------------------------------
No virus found in this incoming message.
Checked by AVG - http://www.avg.com
Version: 8.0.176 / Virus Database: 270.10.10/1905 - Release Date: 1/20/2009
2:34 PM
