36
institution, such as a farm, or a political jurisdiction, such as a city (Chap. 18) with
consideration of physical scales and boundaries, e.g., an environmental boundary
like a watershed (Chap. 19), or the boundaries of important material or energy
inputs and outputs. These boundaries are not just defined by space but also time
(e.g., growing seasons or political cycles). This is a Coupled Natural-Human
System (CNH).
The application of system science is to predict system behavior in order to (1)
design systems and (2) guide decision-making to maximize benefits and minimize
adverse impacts.
Models of systems (Chap. 15) are often considered in two ways: “bottom-up”
and “top-down.”
Bottom-up models start by experimentally isolating and understanding the
individual components of a system and then adding them together (or linking them)
to construct the system. For FEW systems, bottom-up models tend to emphasize the
environmental and technological aspects of a system. The challenge with bottom-up
models is that the whole is greater than the sum of the parts; that is, the isolated parts
do not add up to explain the whole, due to the complex interactions between the
parts. For example, efforts to model how water moves through the system may
capture environmental factors like precipitation (snow, rain), movement of water
through the hydrologic cycle, and even the built environment like dams and networks of water distribution, but miss the legal and policy structures that also control
water flows.
The main problem with bottom-up models is, therefore, that they are never complete or detailed enough to understand the system’s behavior as a whole—although
they may be very accurate for one subsystem or component.
A secondary problem with bottom-up models is that their representation of the
whole system’s behavior may be poor despite a good representation of the behavior
of the subsystems. For example, a weather model of a hurricane could get the energy
of the ocean surface precisely correct, and also its rainfall totals, but still fail to
accurately predict the trajectory of the hurricane as a whole.
Top-down models “deconstruct” a whole system into a few essential components, and then proceed to disaggregate each of the components into a hierarchy of
finer subsystems. For FEW systems, top-down models tend to emphasize economic
and policy aspects of a system and global or national processes. The problem with
top-down models is their limited predictability because of complicated and complex
systems where the large-scale pattern emerges from the interactions of many atomic
(small) parts; this yields surprises. For example, a top-down model of regional water
stress might be based on demographics and prosperity, which motivate financing
and policy, which leads to infrastructure, and withdrawals. This approach might
accurately project long-term water shortages and economic problems of a waterscarce arid region by evaluating aggregated supply and demand for water. However,
this model could not tell you much about whether any individual city or family is
going to run out of water. One city might be in serious trouble, and another immune
to the water stress, based on details that are only available at a finer level of
P. Saundry and B. L. Ruddell
institution, such as a farm, or a political jurisdiction, such as a city (Chap. 18) with
consideration of physical scales and boundaries, e.g., an environmental boundary
like a watershed (Chap. 19), or the boundaries of important material or energy
inputs and outputs. These boundaries are not just defined by space but also time
(e.g., growing seasons or political cycles). This is a Coupled Natural-Human
System (CNH).
The application of system science is to predict system behavior in order to (1)
design systems and (2) guide decision-making to maximize benefits and minimize
adverse impacts.
Models of systems (Chap. 15) are often considered in two ways: “bottom-up”
and “top-down.”
Bottom-up models start by experimentally isolating and understanding the
individual components of a system and then adding them together (or linking them)
to construct the system. For FEW systems, bottom-up models tend to emphasize the
environmental and technological aspects of a system. The challenge with bottom-up
models is that the whole is greater than the sum of the parts; that is, the isolated parts
do not add up to explain the whole, due to the complex interactions between the
parts. For example, efforts to model how water moves through the system may
capture environmental factors like precipitation (snow, rain), movement of water
through the hydrologic cycle, and even the built environment like dams and networks of water distribution, but miss the legal and policy structures that also control
water flows.
The main problem with bottom-up models is, therefore, that they are never complete or detailed enough to understand the system’s behavior as a whole—although
they may be very accurate for one subsystem or component.
A secondary problem with bottom-up models is that their representation of the
whole system’s behavior may be poor despite a good representation of the behavior
of the subsystems. For example, a weather model of a hurricane could get the energy
of the ocean surface precisely correct, and also its rainfall totals, but still fail to
accurately predict the trajectory of the hurricane as a whole.
Top-down models “deconstruct” a whole system into a few essential components, and then proceed to disaggregate each of the components into a hierarchy of
finer subsystems. For FEW systems, top-down models tend to emphasize economic
and policy aspects of a system and global or national processes. The problem with
top-down models is their limited predictability because of complicated and complex
systems where the large-scale pattern emerges from the interactions of many atomic
(small) parts; this yields surprises. For example, a top-down model of regional water
stress might be based on demographics and prosperity, which motivate financing
and policy, which leads to infrastructure, and withdrawals. This approach might
accurately project long-term water shortages and economic problems of a waterscarce arid region by evaluating aggregated supply and demand for water. However,
this model could not tell you much about whether any individual city or family is
going to run out of water. One city might be in serious trouble, and another immune
to the water stress, based on details that are only available at a finer level of
P. Saundry and B. L. Ruddell