technique, to describe changes at multiple scales in a regulated river to illustrate a multiscale analysis. They quantified how the distribution of energy
at multiple scales in a river cross-section was changed by impoundment.
The unregulated river channel possessed an evenly graded distribution of
subchannels, each characterized by relatively low energy. Some years after
regulation, the river had gradually changed into a high-energy main channel
flanked by small-scale subchannels. This multiscale change in channel-bed
form could not have been described with more conventional single-scale
approaches.
There are substantial advantages to multiscale analysis. First, it is a more
accurate representation of reality, so that the causes of and solutions to
environmental degradation can be more accurately determined. Second,
because scale is incorporated as a metric, different-sized organisms
(responding to features at different scales) can be evaluated in a single
analysis. By performing a spatial analysis as a first step, an investigator can
optimally size sampling or simulation to reflect the dominant scales within
a river system rather than impose an arbitrarily selected scale of analysis.
8.3.1.4 Declarative Modules
A well-recognized method for reducing conceptual and programming
complexity involves structuring a model as a set of distinct modules with
well-defined interfaces. Modular design facilitates collaborative model construction, allowing teams of specialists to work independently on different
modules. Modules can be archived in distributed libraries and serve as a set
of templates to speed future development.
The most common approach to model integration, which involves linking
procedural models through the use of distributed object formalisms, is
greatly limited by the fact that the various submodels are, by their nature,
overspecified as modules. That is, in the process of implementing a submodel in a procedural programming language, the modeler generally
“hard codes” many choices, such as programming language, spatiotemporal
representation, model control and input/output (I/O) interfaces, and
computing paradigm (e.g., serial or parallel message passing). These fixed
aspects are extremely limiting and irrelevant to the essential dynamics of
the model. To improve flexibility, it is useful to develop a formalism
for coding archivable modules that allows maximum generality and applicability of the modules. This formalism can be accomplished through
declarative module specifications containing only enough information to
specify the essential dynamics of the module and allowing a wide range
of customized procedural implementations (Maxwell 1999; Maxwell
and Costanza 1997a,b). This approach provides the high level of abstraction necessary for maximum generality, yet provides enough detail to
allow a dynamic simulation to be produced automatically. The approach
separates general specifications from site-specific specifications. Because
only the universal blueprints are included in the module specification,
8. Evolving Approaches and Technologies
145
at multiple scales in a river cross-section was changed by impoundment.
The unregulated river channel possessed an evenly graded distribution of
subchannels, each characterized by relatively low energy. Some years after
regulation, the river had gradually changed into a high-energy main channel
flanked by small-scale subchannels. This multiscale change in channel-bed
form could not have been described with more conventional single-scale
approaches.
There are substantial advantages to multiscale analysis. First, it is a more
accurate representation of reality, so that the causes of and solutions to
environmental degradation can be more accurately determined. Second,
because scale is incorporated as a metric, different-sized organisms
(responding to features at different scales) can be evaluated in a single
analysis. By performing a spatial analysis as a first step, an investigator can
optimally size sampling or simulation to reflect the dominant scales within
a river system rather than impose an arbitrarily selected scale of analysis.
8.3.1.4 Declarative Modules
A well-recognized method for reducing conceptual and programming
complexity involves structuring a model as a set of distinct modules with
well-defined interfaces. Modular design facilitates collaborative model construction, allowing teams of specialists to work independently on different
modules. Modules can be archived in distributed libraries and serve as a set
of templates to speed future development.
The most common approach to model integration, which involves linking
procedural models through the use of distributed object formalisms, is
greatly limited by the fact that the various submodels are, by their nature,
overspecified as modules. That is, in the process of implementing a submodel in a procedural programming language, the modeler generally
“hard codes” many choices, such as programming language, spatiotemporal
representation, model control and input/output (I/O) interfaces, and
computing paradigm (e.g., serial or parallel message passing). These fixed
aspects are extremely limiting and irrelevant to the essential dynamics of
the model. To improve flexibility, it is useful to develop a formalism
for coding archivable modules that allows maximum generality and applicability of the modules. This formalism can be accomplished through
declarative module specifications containing only enough information to
specify the essential dynamics of the module and allowing a wide range
of customized procedural implementations (Maxwell 1999; Maxwell
and Costanza 1997a,b). This approach provides the high level of abstraction necessary for maximum generality, yet provides enough detail to
allow a dynamic simulation to be produced automatically. The approach
separates general specifications from site-specific specifications. Because
only the universal blueprints are included in the module specification,
8. Evolving Approaches and Technologies
145