|
Items in cart [0] |
Questions? Call 888-624-8373 |
PAPERBACK PDF BOOK Free Resources |
||||||||||||||||
|
||||||||||||||||
|
|
|||||||||||||||
| ||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||
| Copyright © 2009. National Academy of Sciences. All rights reserved. Terms of Use and Privacy Statement |
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 74
1
4
Resolving Problems:
Essential Study Elements
Science serves a critical role in seeking solutions to significant envi-
ronmental quality problems and mediating the conflicts that arise among
parties with different perceptions of a problem and its potential solutions.
The San Joaquin Valley Drainage Program (SJVDP) is a clear example
of science in this role. Whether in the San Joaquin Valley or elsewhere,
science performs key functions in a process that involves collecting and
analyzing data, proposing alternative solutions, and articulating trade-offs.
Science deals well with defining the objective properties of water, but
it is less able to address issues that involve value judgments. The degree of
excellence, or quality, of water is a concept that requires value judgments.
This interface between science and human values challenges even the best
problem-solving techniques.
In the Kesterson case, for example, science can determine the concen-
trations of selenium that are toxic to waterfowl or define the relationship
between increasing salt concentrations and crop production. But science
cannot judge which is more valuable, the crop or the waterfowl, nor can
science assign values to predicted outcomes. As was discussed in Chapter
3, it is in this sense that the parallel involvement of ethics, law, economics,
politics, and public policy has made the problems in the San Joaquin Valley
particularly difficult to define and solve.
Finding solutions to environmental problems like those caused by irri-
gation drainage requires difficult choices. Thus the equity and effectiveness
of the process used to seek, evaluate, and implement potential solutions
become critically important. For this reason, good study design is essential.
l
74
OCR for page 75
ESSENTIAL STUDY ELEMENTS
75
Concurring with an earlier National Research Council Study, Ecological
Knowledge and Environmental Problem-Solving Concepts and Case Studies
(National Research Council, 1986), the framework presented here is "in
essence, an admonition to think before acting and to use established sci-
entific principles." Although that National Research Council study focused
on the environmental impact assessment process, this committee's activities
have strongly reinforced many of the same messages. This report also
reinforces many pot ints about sound study design for long-term monitoring
as were highlighted in another National Research Council study, River and
Dam Management: A Review of the Bureau of Reclamation's Glen Canyon
Environmental Studies (National Research Council, 1987~.
The purpose of this chapter is to highlight key elements that the
committee believes are essential in addressing complex problems and that
are likely to prove important to future research efforts. The chapter
introduces five basic functions that characterize problem solving. The first
three elements (i.e., recognizing the problem, defining the problem, and
collecting and compiling data) are examined at length in this chapter.
Chapter 5 examines the final two interpretive elements (identifying and
evaluating alternative responses).
This chapter reflects the committee's deliberations and evaluation
of the problem-solving process, but it has also benefited from the work
of several authors who have explored complex problem solving in depth
(Robertshaw et al., 1978; Simon, 1981; Salthe, 1985; Warfield, 1973; Bald-
win, 1975; Optner, 1965~. Particular attention is paid to how complexity
and uncertainty affect the environmental problem-solving process.
ESSENTL`L STUDY ELEMENTS
Attempts to solve irrigation-induced water quality problems whether
the problem is related to selenium, boron, a pesticide, or something else-
cannot succeed unless the process used to identify, evaluate, and eventually
implement the responses is sound. In its guidance to the U.S. Department
of the Interior and the SJVDP, and by this report, this committee continually
has emphasized the need for formal decisionmaking and effective public
participation in this process, and it has stressed the importance of carefully
integrating technical, socioeconomic, and institutional considerations.
Certain key study elements are critical when decisionmakers attempt to
seek balanced solutions to significant environmental problems. In general,
problem-solving endeavors such as the SJVDP and the National Irriga-
tion Water Quality (NIWQP) should incorporate the following five basic
elements:
1. Recognize the problem.
2. Define the problem.
OCR for page 76
76
IRRIGATION-INDUCED WATER QUALITY PROBLEMS
3. Assess the data base and collect additional data.
4. Identify alternative responses.
5. Evaluate the alternatives.
Table 4.1 summarizes the general process that needs to be followed
to generate responses to major environmental problems when values, view-
points, and science may be in conflict. The process begins with the an-
tecedent conditions that set the stage for the occurrence of a problem.
These antecedents are the environmental variables that create the setting.
As was reviewed in Chapter 2, they can include the hydrologic or geologic
characteristics, ecological or biological factors, or other physical elements
that characterize the existing conditions. In addition, as was reviewed in
Chapter 3, the social and cultural context the economic, social, and politi-
cal setting also creates a backdrop against which a problem occurs. These
elements all contribute in various ways to the complexity of the problem
and the ultimate effectiveness of various proposed solutions.
In the San Joaquin Valley, for example, two of the most important
antecedents were the geology of the area (i.e., the fact that the soils were
rich in selenium) and the nature of the agricultural economy (i.e., the
history and importance of irrigation in the valley). If either of these two
variables had been different, the problem at issue would not have occurred
or at least would have been significantly different.
As was mentioned earlier, defining the problem is a critical and difficult
step in the problem-solving process. Implementation of any solution is
impossible if people do not agree as to the nature of the problem (Vlachos
et al., 1979), because how the problem is defined ultimately determines
the nature of the solutions that are possible. Of course, difficulties arise
because different people have different perspectives and thus will define
different problems. The compromise is to define the problem broadly
and then specify concrete, feasible goals that serve, as best possible, the
different perspectives. Rarely, if ever, is it possible for all parties to be fully
satisfied, and some value judgments will have to be made.
Ultimately, one clear problem must be defined a process that may
involve some difficult value judgments because without this basic under-
standing, obtainable goals cannot be set and alternative solutions cannot be
analyzed in context. If the problem-definition process is adequate, in the
end local, regional, and national interests should be appropriately balanced.
Table 4.1 also lists categories of available responses. These are the
generic tools available to address irrigation-induced water quality problems
anywhere, whether in the San Joaquin Valley or elsewhere in the United
States or the world. They are basic types of responses that might be applied
regardless of site. These can be technical, institutional, or a combination
of both approaches. Appropriate responses can be developed only after
OCR for page 77
ESSENTIAL STUDY ELEMENTS
TABLE 4.1 General Process for Developing Strategies to
Respond to Irrigation-Induced Water Quality Problems
Sequence of Steps
Essential Components
Recognizing the problem
Defining the problem,
assessing and collecting data
Identifying alternatives
Evaluating alternatives
Detection of anomalies
o Chemical parameters
o Physical parameters
0 Biota
o Social impacts
o Economic impacts
Antecedent conditions
0 Hydrological
o Biological
o Geological
o Ecological
Social and Cultural Context
o Historical setting
o Competing and conflicting
demands
o Inherent complexity
o Widespread support for irrigated
agriculture
o Subsidization of water and crops
0 Expectation of continued support
o Institutional constraints
Possible responses
0 Source control
o Drainage water treatment
o Transport and disposal
0 Price adjustments
0 Legal changes
0 Institutional changes
o Economic changes
o Social changes
Criteria for evaluating responses
0 Technical soundness
0 Economic viability
0 Legal appropriateness
0 Social acceptability
0 Political feasibility
o Ecological appropriateness
77
OCR for page 78
78
IRRIGATION-INDUCED WATER QUALITY PROBLEMS
careful interpretation of the data. These responses are discussed further in
Chapter 5.
Given the dynamic nature of any problem-solving process, one cross-
cutting issue critical to the search for appropriate responses is public par-
ticipation. Public participation is important throughout the various stages
of any problem-solving endeavor, but it is particularly necessary during
the definition of the problem and the assessment of alternative responses
(Ingram and Ullery, 1977~. In fact, the success of any proposed solution
will ultimately depend on the public's confidence that the decision process
was open and complete. Public participation is important because it is in-
evitable in any large, public debate that there will be differing views present
among the people affected. In other words, there is no one "public" but
rather many "publics" that must be given access to the decision-making
process: farmers (both irrigators and nonirrigators), business people, envi-
ronmentalists, local and regional residents, and a host of others with varied
rationales for involvement. All sides desire a chance to be heard and to
share in the decision-making responsibilities.
Public participation brings competing interests together, communicates
information, identifies research needs, and helps in the understanding of
scientific uncertainty. It is a forum for decision makers and the public to
listen and learn from each other. Public participation is not a frill; it is a
necessity that has been established by law and upheld by the judiciary. Thus
the question is no longer whether there should be public participation, but
how it can be done most effectively.
Recognizing the Problem
The first step toward solving a problem is recognizing that it exists.
Although problem recognition is difficult and often occurs by happenstance,
examples of strategies to facilitate problem recognition include the baseline
monitoring of chemical and physical parameters as conducted by the U.S.
Geological Survey, remote-sensing efforts by the National Aeronautics and
Space Administration and the National Oceanic and Atmospheric Admin-
istration, and isolated long-term ecosystem monitoring efforts supported
by the National Science Foundation. Many of these efforts are designed
primarily to explore the long-term behavior of natural systems. The number
of parameters monitored often is small, and the geographical coverage of
the studies is limited.
Monitoring detects change. Defining change in the natural world
as a problem, no matter how the change is discovered, depends on a
value judgment by some part of society because damage is a human value
concept. Science can serve society through sustained research vigilance, but
OCR for page 79
ESSENTIAL STUDY ELEMENTS
79
the decision determining which changes are problems (damage) requires
close coordination between monitoring and evaluation.
No common shortfalls interfere with many schemes for problem recog-
nition. One difficulty occurs because the scientific and technical programs
charged with monitoring generally are separated from the value judgment
methods that could identify a change as damage. The second difficulty
is that the technical institutions responsible for monitoring are often the
same institutions responsible for causing the changes, often in the name
of resource development. This creates a built-in bias to ignore unintended
problems as long as the original objectives of the projects are being met.
Few formal strategies exist for efficient problem recognition covering
a wide range of circumstances. Consequently, many problems are first
recognized through dramatic, attention-getting events such as the deaths
and deformities of birds at Kesterson National Wildlife Refuge (NWR), or
through serendipity during studies designed for other purposes. Thus the
threshold of severity that must be reached before a problem is recognized
can vary significantly. The first indications that a problem may exist tend
to be based on the following:
inferences drawn from prior experiences;
· detection of anomalies in chemical or physical parameters;
· detection of anomalies in the biota; or
· detection of socioeconomic impacts.
Many of the potential problems uncovered during the problem-recogni-
tion phase will turn out to be spurious, and therefore such associations need
to be assessed carefully to enable judging the strength of the association
and the likelihood of causal relationships. Again, the San Joaquin Valley
offers a vivid example of how problem recognition can occur: because
the selenium contamination at Kesterson NWR was unexpected, the mass
media played an unprecedented role in the problem-recognition process.
Given the experience gained at Kesterson NWR, monitoring for trace
elements at other sites may be better able to detect emerging problems
when the changes are more subtle and before drastic problems have arisen.
Defining the Problem
The next critical step in solving any problem is to define the nature of
the problem (Box 4.A). Problem definition for complex, multidisciplinary
environmental problems requires agreement among competing interests.
Developing a process to negotiate an acceptable statement of the problem
is a necessary part of problem definition and should precede the setting of
goals.
The importance of an interdisciplinary team undertaking a formal
OCR for page 80
80
IRRIGATION-INDUCED WATER QUALITY PROBLEMS
BOX 4.A Defining the Problem
It is important to negotiate an acceptable problem definition early in any
research effort because different observers will have different perspectives,
focus on different symptoms, and have different goals. The views expressed
can vary widely. A farmer may see the problem as one of diminishing
agricultural productivity, and the causes as increased salinity, rising water
tables, diminished irrigation supplies, or contaminated irrigation water. A
water resources management agency may see the problem as the excessive
accumulation of harmful pollutants in the hydrological system. An environ-
mental activist may focus on the loss of natural environmental attributes
caused by the expansion of agricultural systems. The impacts of diminished
in-stream flows on aquatic wildlife, recreation, and drinking water supplies,
or the simple degradation of natural landscapes, can also be issues.
From a national perspective, the major problem may be the significant
costs required to maintain the current agricultural system. Over the years,
the nation has developed an agricultural production system which, on the
one hand, appears to be very efficient in producing plentiful supplies of
inexpensive commodities, but which, on the other hand, requires billions of
dollars of subsidies for its maintenance. Other people will identify still other
types of problems. And even within any general problem area, different
individuals may define the problem differently.
All of these different definitions have some credibility. And indeed,
the full definition of "the problem" may include elements of all the different
definitions and more. But all too often, little attention is given to defining
exactly what the problem is, and this failure will often become a major cause
of subsequent confusion and conflict among those responsible for identifying
a solution.
problem decomposition was discussed above. The problem-definition phase
defines the purpose of the research and the future outcome that is desired.
The participants must resist the urge to find quick solutions during the
problem-statement phase. Sometimes a team of experts with no vested
interest in the outcome-in other words, a panel of outside experts is
best suited to evaluate the problem objectively. In any event, the inclusion
of formal value judgment is required.
It is absolutely essential to define a problem before seeking to solve
it. Although this may sound simple or obvious, it is not. When asked to
specify the problem they are trying to solve, farmers, scientists, engineers,
citizens, federal and state agency staff, and other interested parties may all
see the problem differently or may focus on different symptoms. Public
participation must be incorporated at this early stage of problem definition.
Acknowledging the existence of different views of a problem is important
OCR for page 81
ESSENTIAL STUDY ELEMENTS
81
because it means that the problem solvers will be less likely to proceed
down a short-sighted path.
How the problem is defined whether explicitly or implicitly will
determine which options are examined and implemented. Obtainable goals
can be set only if the problem to be solved is clear and agreed upon by
all parties. Different options will produce different effects on the various
interest groups. A response at the local level may aggravate the problem
at the state or national level. The simplest engineering response may be
an environmental mistake. A temporary remedy may preclude a future,
permanent cure. The benefits of each alternative for all affected parties
must be carefully assessed, and decisionmakers must remember that all
potential solutions have costs in money, resources, energy, and social
costs. Who will pay becomes an essential consideration, and, as for the
other questions raised, the answer depends very much on the perspective
from which the question is asked. It may be that no answers are possible
in which all the parties win, so that compromise is more often than not the
only realistic goal.
This committee cannot stress enough the importance of clearly defining
exactly what problem is being addressed and of making early problem
definition a crucial element in all attempts to study and solve irrigation-
induced water quality problems in the future. Each level of a problem
contains its own set of intertwined subproblems. If certain elements of the
problem are ignored, unforeseen repercussions will result. For example,
plugging the drains that discharged into the ponds at Kesterson NWR
was a response that did reduce the flow of contaminants into the refuge.
However, it did not solve several larger problems, such as what should be
done with the sediments that contain dangerous residues or who should
pay for cleanup. It did not address the obvious question of how to maintain
agricultural production without drainage, or the more subtle question of
how to compensate for the wetland habitats that were lost.
Data Assessment and Acquisition
Assessing the Data Base
Once a problem has been defined and goals set, problem solvers
should next assess the existing data base (National Research Council, 1986;
Larkin, 1984~. Too often, people confronted with a complex problem
tend to assume that the required data are absent and immediately begin
acquiring new data. Existing data often are ignored, underused, or treated
as suspect. Although existing data may have been developed from studies
with different objectives, they still can provide valuable insights about the
nature of the system interactions, a key component when dealing with
complex problems.
OCR for page 82
82
I~GATION-INDUCED WATER QUALITY PROBLEMS
From the existing data base, researchers should attempt to establish a
baseline, a trend, and some idea of the required endpoint. The question
of how (by what criteria) to judge the problem must also be addressed.
The research team should first screen the facts. This involves separating
objective information from expert opinions, educated guesses, speculations,
and other questionable data. But this step inevitably includes some value
judgments, and these opinions need not be discouraged: they simolv need
to be noted for what they are.
The assessment of quantitative and qualitative information derived
from ecological field studies is particularly difficult to deal with because such
studies often exhibit poorly quantified precision and accuracy. Nevertheless,
such studies may be the only ones that integrate the effects of combined
stresses on the environment. A well-designed exploratory analysis of the
existing data base can help decisionmakers assess the quality of this data
base and identify significant data gaps. If the facts are insufficient, further
information must be acquired.
Before acquiring new data, however, problem solvers need to develop
a formal statement of the specific measurement objectives, including a
complete list of variables to be measured. The reporting units, expected
ranges, required detection limits, relative prediction (upper limit), and
accuracy (maximum absolute bias) objectives must be specified for each
target parameter before measurement begins. Clearly stated data objectives
are necessary to the design of a quality assurance and quality control
procedure at the beginning of the measurement process (Box 4.B).
Often a quality control plan is developed too late to be of real use in
assessing the quality of information being acquired. A clear statement of
data objectives also helps field and laboratory personnel assemble candidate
measurement procedures and examine their cost-effectiveness. Questions
of sampling strategy, definitions of sample representativeness, and similar
issues all require specific objectives. Attention to method selection, devel-
opment, and optimization should precede the adoption of routine analytical
measurement procedures. Appropriate laboratories and investigators must
also be selected to perform the work.
C' ~1 ~
AcqumngAddihonal Data
Existing data can provide important input into the decision-making
process, but they will likely need to be supplemented with new data specif-
ically tailored to the situation. Thus data acquisition is a key, and often
time-consuming, stage in environmental problem solving. One area of par-
ticular importance is public health. The potential threat to public health
posed by the increasing exposure of people and wildlife to water contam-
ination mandates a closer examination of the importance of these data.
OCR for page 83
ESSENTLi4L STUDY ELEMENTS
BOX 4.B Quality Assurance and Quality Control
Uncertainty plays a constant and important role in problem solving.
To properly evaluate alternatives, a quantitative mechanism for monitoring
uncertainties should be included in all experimental designs. The design of
a quality assurance (QA) plan and implementation of quality control (QC)
procedures should occur early in any study. A member of the study team
should be assigned responsibility for QA/QC. When a project incorporates
the work of many people in several.places, a QA/QC manager should be
appointed as early as possible. In addition to the manager's coordinating
role, several broad QA objectives can be addressed only by a QA/QC
manager. Establishing QA program guidelines for data precision, accuracy,
completeness, representativeness, and comparability requires a whole-project
perspective that individual participating laboratories and data-gathering task
groups cannot provide.
Decisions about the utility of data for answering particular questions
depend on the objective of the study, the sampling design, and protocol.
There are data adequate to answer some questions that would be useless for
answering others. In addition, the degree to which one may assess sample
representativeness depends on the precision limits of the analytical methods,
i.e., whether field variability may be distinguished from laboratory impreci-
sion. Thus, once the goals for analytical precision have been established, one
may define representativeness and establish a protocol for assessing whether
the objectives have been achieved. If this is carefully done and meticulously
documented, then future analysts can be confident that data are, or are not,
appropriate for new analyses. These are important criteria for long-term data
sets that must serge time series analyses, analyses of change, and analyses
about the effects of experimental manipulation or management.
83
In particular, monitoring, dose response studies, and exposure assessments
(Box 4.C) play a key role in risk assessment and in the evaluation of
alternative responses to a problem.
Monitoring, or the routine collection of data, is used in ecological
studies in two basic ways. Anticipatory monitoring is designed to track the
effects of activities that might be cumulative or pose hazards to human
health. Monitoring during or after an action or project is designed to show
what ecological changes resulted (Baker, 1976~. Properly done, monitor-
ing provides continuous indexes of environmental quality that can signal
environmental degradation or improvement (National Research Council,
1986~.
Monitoring often is avoided because it is expensive and the return of
information for each dollar spent seems small. The additional expenditure
may be difficult to defend because the contribution that monitoring data
OCR for page 84
84
IRRIGATION-INDUCED WATER QUALITY PROBLEMS
BOX 4.C Exposure Assessment
Exposure assessment is a process used to estimate the rates at which
substances are absorbed by organisms through all mechanisms: ingestion,
inhalation, and by absorption through the skin. The absorbed dose often
differs significantly from the dose externally applied to the organism, which
is usually called the exposure.
A valid measurement of the organism's exposure to a chemical would
require measuring that chemical in the food, drinking water, air, and sur-
faces with which the organism would come into contact. Environmental
concentration measurements that do not consider the chemical and phys-
ical forms of the contaminants provide an imperfect basis for estimating
absorbed doses because these forms affect the gastrointestinal absorption
efficiency, the percutaneous transfer coefficient, and other important param-
eters. Ideally, the analytical data should also provide information on the
physical and chemical form of the substance being analyzed. In practice,
, most exposure assessments do not incorporate such sophistication, and the
resulting environmental assessments are weak.
When a possible environmental contamination problem is initially in-
vestigated, a large number of potential contaminants should be sought in
those areas where they would be expected to accumulate, by natural pro-
cesses, to unusually high concentrations. Next, the team should clearly
define the geographical extent of the problem and the major environmental
media for those contaminants uncovered in the preliminary phase. Then
the team should concentrate on making exposure measurements for humans
and selected organisms. Monitoring data collected during reconnaissance
should not be used to make definitive risk assessments without clarifying the
tentative nature of such assessments.
provide to assurance of safety and effectiveness generally is not evident
during the initial years.
Monitoring is, however, quite important during all phases of a water
resources investigation, and its importance will increase as water quality
problems become more frequent and the sources of contamination more
abundant and diverse. Survey monitoring, for example, tracks ambient
conditions, detects changes, and identifies problem areas on a routine
reconnaissance basis. As anomalies are detected, it may be necessary to
supplement existing monitoring networks with additional measurements to
obtain a better understanding of the study system.
Monitoring should not be restricted to the period of study of a partic-
ular problem situation but should be continued after packages of solutions
have been selected and implemented. This continued monitoring provides
a means to assess the effectiveness of the strategic response chosen and
permits identification of other potential anomalies. Of course, one serious
OCR for page 85
ESSENTIAL STUDY ELEMENrTS
85
problem in designing any monitoring system is the assumption that someone
knows what substances to monitor. Had a well-designed water monitoring
system existed in the San Joaquin Valley prior to the discovery of problems
at Kesterson NWR, it probably would not have provided advance indication
of the selenium contamination because there was no basis in experience
that warned researchers to monitor explicitly for selenium.
The National Irrigation Water Quality Program (NIWQP) the cre-
ation of which was inspired by the San Joaquin Valley experience provides
an example of the kind of basic data collection and monitoring that is nec-
essary to identify irrigation-related contamination problems. This program
Is an attempt to anticipate and identify contamination problems before they
take on Kesterson-like proportions, and it relies on a series of evaluative
steps ranging from desk reviews to reconnaissance-level field studies to
detailed field studies at sites showing potential problems. Given that the
nation now is aware of these types of problems and their potential con-
sequences, this committee believes some program of this type will remain
necessary in the long term.
Interpreting the Data
Converting the assembled data Into information is as important to
problem solving as experimental design is to data gathering. One ele-
ment of any information-gathering process should be an information base
management system. The data acquisition/interpretation plan supplies an
operational mechanism for information exchange and catalyzes the interdis-
ciplinary interaction. Interdisciplinary research requires a concerted effort
to force researchers to address the data/information base in its broadest
interpretive context.
A well-designed information management system should be more than
just computer software or a commercial data management system. It should
incorporate human creative elements using interpretive aids to display the
data conveniently, summarize its information, induce thinking about its
content, and facilitate its use as an instrument of reasoning. The data base
manager is a key individual in the interdisciplinary team. The individual
should have the technical expertise needed to critically evaluate the data
and function as an aid to retrieval and analysis. All projects benefit from
the broader perspective of a competent generalist. Thus one central role
for information management is to provide the day-to-day continuity that
keeps the systems design approach productive. Chapter 5 addresses data
interpretation in more depth, as it is fundamental to the tasks of identifying
and evaluating alternative responses and strategies.
OCR for page 86
86
IRRIGATION-INDUCED WATER QUALITY PROBLEMS
COMPLEXITY AND STUDY DESIGN
Many descriptions of environmental problems begin by stating that the
problems are complex. Although this may appear to be a statement of
the obvious, study designs seldom exhibit a truly thoughtful examination
of the claimed complexity. Irrigation-induced water quality problems are
indeed complex. But one central role for natural science is "to show
that complexity, correctly viewed, is only a mask for simplicity" and "to
find pattern in apparent chaos" (Simon, 1981~. The solutions to complex
problems are not always themselves complex, although they must take into
account the relevant complexity. Complexity can be addressed through
study design. Two types of complexity descriptive and interactive-need
to be considered.
Descriptive Complexity
Descriptive complexity results from observers with different perspec-
tives and institutions with different missions using different approaches
to dissect a system into subsystems (Box 4.D). This often results in poor
problem definition, and one consequence of this is that too much time is
spent trying to solve the wrong problems. Descriptive complexity occurs
because inherent differences of scale whether spatial or temporal are
addressed differently by individuals with differing objectives. For example,
a farmer whose objective is economic survival will describe a problem in a
more short-term light; a resource manager, given a mandated responsibility
for the resources being managed; is likely to describe the complexity more
broadly in terms of both time and space.
The difference in the perception of beneficial or adverse effects also
varies given the perspective-environmental, agricultural, or societal. To
accommodate descriptive complexity, dynamic and flexible approaches to
problem solving are necessary. The approaches must be interdisciplinary
and must involve the public.
The distinction between "interdisciplinary" and "multidisciplinary" is
more than semantic. Simply including studies from several disciplines does
not ensure that relevant system interactions will be uncovered or that
integrative, interpretive solutions will be obtained. Humans are adept at
recognizing complexity but often avoid implementing remedies that require
widespread changes in behavior. The complexity itself becomes an excuse
for lack of action. This makes problem definition and planning key steps
in problem solving.
Complex problems often have a hierarchical structure, and solutions
require the use of formal systems analysis. Systems are collections of things
that function together, and the study of these collections is called systems
OCR for page 87
ESSENTIAL STUDY ELEMENTS
BOX 4.D Descriptive Complexity in the San Jonquils Valley
Some of the complexity that characterizes the situation in California's
San Joaquin Valley came about because events inevitably mixed together
people who held incompatible values. Different people express their goals in
different terms: tons of cotton, waterfowl, human health, money, influence,
or esthetics. Thus the San Joaquin Valley case is also a "complexity of
values," generally expressed as special interests.
Nowhere was the potential for conflicting perspectives more apparent
than in the seemingly simple task of exactly defining the problem in the San
Joaquin Valley. Is the principal issue how to better manage water on the
farm so that the volume of drainage, and thus contaminated waste water, is
minimized? Or is it how to protect water quality and in-stream values? Is
the question one of broad economic benefits for the nation, or of continuing
the historic agricultural lifestyle in the San Joaquin Valley? Is the issue
the protection and enhancement of wildlife resources, particularly waterfowl
habitats? Are the events in the valley an isolated problem or are they
representative of a broad national issue?
These questions reflect the different perspectives, levels of authority,
and interests of the many people involved and potentially affected by the
answers. Local governments often perceive the issues much differently than
do federal agencies. In fact, although many institutions exist to examine
separate pieces of the water use puzzle, none has shown the breadth and
flexibility needed to integrate water policy across the disciplines.
87
analysis (Haith, 1982~. There are obvious advantages in treating environ-
mental problems as systems. Problems can be considered in their totality,
and the most effective points of control can be sought. A consequence of
a systems perspective on environmental quality is the broadening of pos-
sible control options and subsequent opportunities for efficient, integrated
management strategies (Haith, 1982~.
A formal, collaborative systems analysis can help identify the levels
of the problem hierarchy and provide a useful mechanism for breaking
down the problem into its essential elements. This decomposition allows
the various parts of a complex system to be considered in isolation, but
still in the context of the whole. Decomposition makes it easier to identify
any part of the problem that needs particular attention, and it is easy to
put things back together when the decompositions are formal. This is how
science seeks to simplify.
The levels of the hierarchy are characterized by shared properties,
such as spatial scale and temporal frequency, each subject to different
degrees of resolution during data collection phases of empirical science
(Box 4.E). Formal examination of the levels of the hierarchy permits the
OCR for page 88
88
IRRIGATION-INDUCED WATER QUALITY PROBLEMS
BOX 4.E The Problem Hierarchy in the San Joaquin Valley
When a problem is analyzed, spatial resolution determines how much
area is examined and temporal resolution determines the length of time.
Three geographic scales are obvious in the San Joaquin Valley example:
Kesterson National Wildlife Refuge (NWR), the San Joaquin Valley, and the
arid West. Three time scales are appropriate: short term (years), medium
term (decades), and long term (centuries). Cleanup at Kesterson NWR
represents a local, short-term goal. Reassessing water management in the
San Joaquin Valley addresses a watershed on the medium term. Achieving
a balance between sustainable agriculture and environmental values would
be a long-term goal.
A hierarchy of problem levels is clearly present. The discharge of
drainage water into the ponds at Kesterson NWR has left a serious toxic
cleanup problem; this is commonly referred to as the "Kesterson problem."
Second, the plugging of the contributing drains has aggravated drainage
problems for much of the irrigated land on the west side of the San Joaquin
Valley, and this is referred to as the "San Joaquin Valley drainage problem."
Third, the documentation of toxic concentrations of selenium in the drainage
water raises the spectre of similar problems elsewhere in the West, and this
broad issue is called the "irrigation-induced water quality problem."
It is easy to see how these differences in perspective add to complex-
ity. The drainage problems in the San Joaquin Valley have been examined
according to a variety of different organizational strategies. Economists,
politicians, ecologists, and legal analysts each simplified the issues by assum-
ing that the others' views were fixed. The extradisciplinary information then
was included as a constant or discarded as irrelevant.
problem-solving team to identify essential communication channels in the
interdisciplinary structure.
Many disagreements, particularly those that are hard to settle, are
characterized by the disagreeing parties having addressed the issue at
different levels, or in different contexts, with different values influencing
their perspectives.
Such disagreements are not resolved by factual information. When it
is realized that no answers are possible in which all parties win, the estab-
lishment of a common ground followed by compromise becomes the only
realistic goal. - In essence, there are no win-win situations, only compro-
mises in which all parties must give a little to attain a solution acceptable
to all. The systems approach to problem solving provides a rubric for such
problem definition and a structure for interdisciplinary collaboration.
OCR for page 89
ESSENTIAL STlJDY ELEMENTS
89
Interactive Complexity
While descriptive complexity is the result of problem perception, in-
teractive complexity is characteristic of natural systems. It is the result of
direct or indirect interactions among variables within a subsystem or of
interactions between subsystems. For example, the question of cadmium
toxicity to humans cannot be adequately addressed without understanding
the status of zinc. The toxic interaction between these two elements is
mediated by natural biochemical processes occurring in human cells. Most
environmental problems contain many significant interactions. Thus an
engineer, chemist, or biologist might view complexity in terms of the num-
ber and magnitude of system interactions. Unless these interactions are
identified and understood, the solutions proposed are likely to fail.
In a complex problem the relevant complexity must be accounted for,
but the key to useful solutions is to reject the irrelevant complexity and
uncertainty. It has become all too common to claim that "everything is
connected," but good study design reflects the fact that most things can
be looked at separately and that most connections are weak and can be
ignored. On the other hand, it is essential to recognize and deal with
significant interactions and to be aware that the sum of a number of weak
interactions may be significant. It would be an error to think that when a
dominant cause has been identified, the other factors are irrelevant. Thus
a good study design should create a data-gathering structure that is capable
of discovering unanticipated interactions and determining the magnitude
of expected interactions. Failure to adequately address interactional com-
plexity during problem definition leads to short-term solutions that can be
long-term disasters.
In addition to variable interactions within a domain of study, interac-
tions between domains also introduce complexity that must be addressed in
a study design. Although adequate theory may exist to predict interactions
within a study domain, theory that identifies and permits quantitative assess-
ment of interactions between apparently disparate domains is lacking. For
example, interactions between elements of the technical domain and the
social and economic domains that is, the linkages between science and
values are poorly understood. Furthermore, few theoretical constructs
exist to link phenomena occurring at different scales. This means that
predictability at large scales (regional, continental, or global) or far into
the future is not yet possible.
UNCERTAINTY
Given a good problem representation and a recognition of relevant
interactions, the problem solver must then map the consequences of the
OCR for page 90
go
IRRIGATION-INDUCED WATER QUALITY PROBLEMS
alternatives, recognizing that some predictions will be more certain than
others. Thus, to understand the nature of complexity and its consequences
in a problem-solving endeavor, the problem solvers must examine the role
of uncertainty.
Since the consequences of actions may be far reaching and long lasting,
a predictive capability is useful for assessing the effects of human activity.
Accurate prediction, however, requires a theoretical understanding of the
phenomena to be predicted as well as reliable data. Unfortunately, available
scientific theories often are incomplete, and the available data are uncertain.
Successful problem solving must be based on a strategy that addresses
complexity and recognizes that uncertainty is an inherent part of any
problem.
Uncertainties can be of two types: those that people know how to
remove without extraordinary effort, and those that people do not know
how to remove without extraordinary effort or that may not even be
recognized in the problem.
The first type of uncertainty includes the random errors associated
with measurement, and measurement limitations imposed by methods with
insufficient sensitivity, data gaps, and so on. Although sometimes problem-
atic, these can be minimized without too much effort. If these uncertainties
cannot be ignored, then problem-solving procedures must determine the
added cost of reducing the uncertainty to acceptable levels and must com-
pare that cost with the cost of not having the additional information. For
example, improving the measurement precision by a factor of two could
easily increase the cost of the measurement by a factor of four. It is
not always clear that the reduced uncertainty in a few measurements will
proportionally improve the final uncertainty in complex systems. In any
case, quantitative information regarding the measurement process and con-
tinuous performance surveillance are essential parts of problem solving.
The crucial role of a quality assurance and quality control program will be
described in more detail below.
The second type of uncertainty deals with uncertainties that arise out
of science's incomplete understanding of how things work. Uncertainties of
this type include the variabilities of human behavior, the weather, political
events, and similar factors where judgments are based on assumptions
rather than facts. Even with extraordinary efforts, these uncertainties
cannot be modeled adequately. Hidden variables perturb the system in
unanticipated directions, which makes prediction very uncertain even when
the uncertainties related to measurement type have been minimized.
Since these uncertainties arising from exogenous events cannot be
quantified, attempts are made to deal with them by estimating the prob-
abilities that they might occur (another assumption) using risk analysis
(sordid and Rodricks, 1987; Hogan and Hoel, 1989), as highlighted in
OCR for page 91
ESSENTL4L STUDY ELEMENTS
BOX 4.F Risk Assessment
Risk assessment is a process that seeks to estimate the likelihood of
occurrence of adverse effects due to specific exposures to chemical, physical,
and biological agents in humans as well as ecosystems. The assessments may
involve qualitative as well as quantitative estimates. Risk assessments often
must be made from fragmentary data and with data that were collected for
purposes not related to making risk assessments. Thus, by their very nature,
risk assessment processes emphasize extrapolations and are sometimes prone
to inaccuracies.
Risk assessment is but one phase in a much larger process that seeks
to prevent adverse effects on public health or ecosystems or the economy.
It is closely linked with exposure assessment and with risk management.
Risk management combines political, legal, and engineering approaches to
manage risks.
Potential risks are estimated by considering the probability of occur-
rence, the potential effects, and the exposure, all in order to make the
assessment of potential risks associated with the exposure to chemicals more
tractable. There are, however, generic limitations to risk assessment. For
instance, the number of substances for which an adequate amount of in-
formation exists for credible risk assessments is limited. Risk assessments
for complex mixtures and for intermittent and fluctuating exposures are
unreliable. Risk assessments for the protection of ecosystems are only in
their early developmental stages.
91
Box 4.F. Risk analysis, based as it is on assumptions, contains significant
uncertainties. An alternative or supplementary uncertainty management
strategr is to build feedback controls into the study design and solution
implementation plan so that plans can be altered as data improve or as
more is learned about the system.
The three elements outlined here recognizing a problem, defining the
problem, and assessing the data base and collecting additional data are
essential steps in any problem-solving endeavor. ~ identify appropriate
responses-ones that adequately and fairly respond to the stated goals of
the problem-solving endeavor requires careful analysis. Technical, eco-
nomic, legal, ecological, social, and political criteria must all be evaluated
in an attempt to weigh the relative advantages and disadvantages of each
proposed approach. The identification and evaluation of appropriate re-
sponses is of course the cornerstone of any problem-solving endeavor.
These critical steps receive detailed attention in Chapter 5.
CONCLUSIONS
Environmental quality problems tend to be complex, difficult to re-
solve, and controversial. However, a problem's complexity should not be an
OCR for page 92
92
IRRIGATION-INDUCED WATER QUALITY PROBLEMS
excuse for taking no action to solve it. A substantial degree of formality
will be necessary to address complex problems successfully. This will help
reduce wasted effort, increase the scientific integrity of the process and the
solutions ultimately proposed, and foster their public acceptance.
When designing studies to resolve environmental problems, it Is im-
portant to recognize the nature of complexity- both to acknowledge it
and to remove the excuse that, because of the complexity, the problem Is
intractable. The hierarchical nature of a large, complex problem ~nvolv-
ing many disciplines and interest groups must also be recognized because
communication Is possible only at the same hierarchical level.
A well-conducted study plan employs, in order, the elements of prob-
lem recognition, problem definition, data assessment and collection, gen-
eration of alternative solutions, and evaluation of these alternatives. In
structuring any study, explicit attention must be paid to quality assurance
and quality control, data and information management, monitoring, risk
assessment and uncertainty, public participation, and conflict management.
One aspect of complexity is that no environmental problem is solely tech-
nical or solely institutional in nature. All involve technical, legal, social,
and institutional components.
A formal systems analysis framework will aid in giving appropriate
weight to each of these disciplines and in enhancing communication. A
wide range of alternative potential solutions needs to be displayed and
analyzed formally. This not only avoids the obvious pitfall of overlooking
important possibilities, but it also provides a basis for establishing the costs
of the preferred alternatives compared to others. It also increases the
credibility of the study recommendations.
As indicated in Chapters 2 and 5, the solutions to most environmental
problems will involve important technical components. However, such
solutions cannot be solely technical but rather must also deal in legal, social,
economic, and institutional domains, as emphasized in Chapters 3 and 5.
These components should be Integrated throughout the problem-solv~ng
process. Viable long-term solutions must be chosen based on societal
judgment, and these can be assessed only when accurate information on
the economic, legal, and institutional environment Is available.
REFERENCES
Baker, J. M. 1976. Biological monitoring-principles, methods and difficulties. In Marine
Ecology and Oil Pollution. J. M. Baker, ed. John Wiley & Sons, New York.
Baldwin, M. M., ed. 1975. Portraits of Complexity: Applications of Systems Methodologies
to Societal Problems. Battelle Memorial Institute, Columbus, Ohio.
Haith, D. A. 1982. Environmental Systems Optimization. John Wiley & Sons, New York.
Hogan, M., and D. Hoel. 1989. Extrapolation to man. Pp. 879-891 in Principles and Methods
of Toxicology. 2nd Ed. A. Wallace Hayes, ed. Raven Press, New York.
OCR for page 93
ESSENTIAL STUDY ELEMENTS
93
Ingram, H. M., and S. J. Ullery. 1977. Public participation in environmental decision-making:
substance or illusion. Pp. 123-139 in Public Participation in Planning. ~ R. D. Sewell
and J. T. Coppock, eds. John Wiley & Sons, New York.
Larkin, P. ~ 1984. A commentary on environmental impact assessment for large projects
affecting lakes and streams. Can. J. Fish. Aquat. Sci. 41, 1121-1127.
National Research Council. 1986. Ecological Knowledge and Environmental Problem-Solving:
Concepts and Case Studies. National Academy Press, Washington, D.C., pp. 104-115.
National Research Council. 1987. River and Dam Management: A Review of the Bureau
of Reclamation's Glen Canyon Environmental Studies. National Academy Press,
Washington, D.C.
Optner, S. L. 1965. Systems Analysis for Business and Industrial Problem Solving. Prentice-
Hall, Englewood Cliffs, New Jersey.
Robertshaw, J. E., S. J. Mecca, and M. N. Rerick. 1978. Problem Solving: A Systems
Approach. Petrocelli Books, New York.
Salthe, S. N. 1985. Evolving Hierarchical Systems: Their Structure and Representation.
Columbia University Press, New York.
Simon, H. A. 1981. The Sciences of the Artificial. MIT Press, Cambridge, Massachusetts.
Ibrdiff, R. G., and J. V. Rodricks, eds. 1987. Toxic Substances and Human Risk. Plenum
Press, New York. 445 pp.
Vlachos, E., G. V. Skogerboe, G. E. Radosevich, P. C. Huszar, and W. Dock. 1979.
Socio-Economic and Institutional Factors in Irrigation Return Flow Quality Control.
Prepared for Robert S. Kerr Environmental Research Laboratory, Office of Research
and Development, U.S. Environmental Protection Agency (EPA Grant R-8034742~.
Warfield, J. N. 1973. An Assault on Complexity. Battelle Memorial Institute, Columbus,
Ohio.
Representative terms from entire chapter:
water quality
