Subject: Loop of the Week

Posted by John Sterman on 11/5/2003
In Reply To:Loop of the Week Posted by Jay Forrester on 11/4/2003

Message:

From John Sterman:

Thanks to Jay Forrester for commenting on the example of my
daughter's 7th grade geography text in which the author asks students
to determine where the water cycle begins. I argued that the
question itself reinforced an incorrect and harmful open loop,
sequential perspective. Jay suggests that the book should not be
faulted for describing the water cycle as having a beginning and an
end, and that a system dynamics model of the process could treat the
ocean as the starting point and as having an infinite capacity. I
don't agree.

Jay is right that the term cycle as used in "the water cycle" refers
not to an information feedback loop but to a physical circulation of
matter. The term cycle is commonly used to describe this and other
circular material flows such as the carbon cycle. The multiple uses
of the term cycle (a circular flow of material, an information
feedback loop, a type of dynamic behavior) creates confusion (this is
something I think we have all experienced in teaching system dynamics
and has been documented by Linda Booth Sweeney in her research). The
lesson is that we have to be careful to make clear which sense of the
term "cycle" we intend when we teach.

Jay also suggests it would be acceptable in a system dynamics model
to show the flow of water as starting from an infinite source (the
ocean) and then ending when water returns to the ocean (flowing into
an infinite sink). I believe this would be both incorrect and
misleading. A good system dynamics model of the water cycle should
include the circular (conserved) flow of water as it moves from one
state to others (e.g., from the ocean to atmosphere to snow/rain to
glacier/river/lake, etc.). It would also include the feedbacks that
regulate these flows.

It is true that the quantity of water in the ocean dwarfs the
quantity in all other states, but it would be both physically
incorrect and misleading to treat the stock of water in the oceans as
an infinite source/sink by representing it as a cloud. To do so
violates conservation of matter, and reinforces a pervasive and
damaging mental model in which there are infinite sources and sinks.
The earth is a thermodynamically open but materially closed system --
except for meteorites and space dust (coming in) and a few spacecraft
(going out), the quantity of matter on earth is constant. Models of
biogeochemical processes should not suggest otherwise by using
infinite sources or sinks. We should not inadvertently reinforce the
impression that the earth has infinite capacity to supply resources
or absorb wastes.

It is not true, as Jay wrote, that "the ocean in this example is
inactive in the dynamic movement of water." Every outflow from a
physical stock must have a first-order negative loop to capture the
physical fact that there can be no outflow if the stock is zero. For
the ocean this constraint will never become active, but by creating a
model in which it could you lose the opportunity for people to learn
this critical lesson. And the constraint is relevant to other bodies
of water that are drying up such as the Salton sea and the Aral sea.

Further, while the ocean is vast, it is not well mixed. The
so-called 'mixed layer' near the surface (top few tens of meters) is
reasonably well mixed but equilibrates with deeper waters over much
longer time horizons (on the order of decades to centuries). The
state of the mixed layer (temperature, salinity, concentrations of
dissolved O2, CO2, and other chemical species) has significant
feedback effects on the atmosphere and deep ocean, including the
water cycle and carbon cycle. Feedbacks from the state of the mixed
layer to the rate of evaporation are critical and must be included in
dynamic models of evaporation and transport, both on the time horizon
of weather and on the longer time horizon of climate.

It is also not true that "evaporation does not dynamically change the
size of the ocean or cause the ocean to have an effect on the
evaporation." It is true that the reduction in the quantity of water
in the ocean caused by evaporation is trivial compared to the total
mass of water in the oceans, but the water that evaporates from it
does reduce the quantity remaining in the seas. Further, evaporation
does change the state of the ocean; it significantly changes the
state of the mixed, surface waters; as these waters give up heat and
water to evaporation, they grow colder and more saline, then start to
sink to deeper layers, to be replaced the upwelling of deeper, colder
waters. These feedbacks are the source of currents such as the North
Atlantic thermohaline circulation, which includes the Gulf stream.

Of course many of these feedbacks are well beyond the level of a 7th
grade geography text. Yet that is no reason to compromise good
modeling practice to build a model that does not conform to
conservation of matter or that shows a one-way flow of water through
the environment. It is never too early to begin developing a
closed-loop, feedback view of the world, including physical processes
and human behavior. If textbooks and teachers reinforce the view
that such processes have beginnings and ends we should not then be
surprised when these students become adults who fail to see the
circular flows of matter on this finite planet or the feedback loops
that govern them.

John

Follow Ups:

Loop of the Week - Bill Barroway 11/6/2003 
Loop of the Week - Peter Welles 11/6/2003 
Loop of the Week - Richard Turnock 11/7/2003
Loop of the Week - Dexter Chapin 11/7/2003
Loop of the Week - Kevin Williamson 11/5/2003 
Loop of the Week - John Heinbokel 11/6/2003