31.5 Assessing the Model: Workshop 2
rent model agreed with historic or observed
changes in vegetation patterns. That is, the group
was unable to test the current set of rules, although
the model can qualitatively match current patterns.
Another gap was how to model longer-term vegetation dynamics, especially how communities respond to chronic disturbances such as flooding
stress and reestablishment following different levels of flooding. Another item discussed was the
need for a nutrient submodel that would link hydrology, sediments, and vegetation dynamics.
Habitats
The habitat group's discussions focused on weaknesses in indicators. Questions arose as to whether
it would be better to model responses of key species
or to develop habitat indicators. Related to the development of habitat indicators, participants were
uncertain whether enough empirical information
exists to model habitat changes based only on depth
and duration of inundation or if other factors need
to be considered. As with the vegetation subgroup,
model credibility was a key source of uncertainty,
and participants cited the need to test the submodel
against historic data.
Human Uses: Navigation and Recreation
Uncertainties related to human activities on the
river included gaps in information and the need to
determine appropriate indicators. For example, are
data available to determine how different management changes would affect key commercial and
recreational uses of the river? Other identified gaps
include the need to model negative interactions between recreational and commercial traffic and a
need to accurately identify changes in dredging requirements associated with different management
schemes.
31.5.4 Confronting Uncertainties
To address the information gaps and uncertainties
described, workshop participants suggested a variety of model refinement, model assessment, and
scenario development activities.
Model Refinement
A number of suggestions were made regarding potential model refinements. These involved modifications to existing submodels, creation of new submodels or routines, and additions to the user
interface.
One debate at the workshop was whether to continue development of two models (i.e., at the pool
469
and system scales) or to concentrate work on a
system-scale model with aggregated or broad-scale
indicators for sediment, vegetation, and habitats.
This was compounded by the difficulties of simulating the hydrology and sediment dynamics at a
1 DO-meter resolution in the pool. At this spatial resolution, the model is extremely slow for somewhat
reliable results (on the order of 10 minutes per simulation year). One option suggested was to use
higher-powered hydraulic models, the results of
which could then be plugged into the pool-scale
model. The debate remains largely open, with no
selection of preference made at the workshop.
A number of modifications were suggested for
each of the submodels, including the following:
1. Compiling data and building better representations of sediment transport in the tributaries
2. Developing a model of sediment-landform
building
3. Economic assessments for potential structural
modifications (e.g., construction and operating
costs of changes to dikes, levees, and channels)
4. Developing a water quality model
5. Developing stage-area relationships for habitat
and vegetation responses
6. Compiling data on recreational and navigational demands
7. Developing composite dredging cost curves
8. Connecting pool models to evaluate cumulative effects
9. Adding remaining pools to the river system
model (i.e., pools 11 to 26)
10. Refining vegetation models with other factors,
such as flooding and water stresses, nutrients,
and temperature
Workshop participants suggested modifying the
user interface to permit comparisons among various output maps. One suggestion was to have the
option for simultaneous display of three maps: two
of the maps would represent results under different scenarios and the third would highlight differences and similarities between the other two maps.
Other suggestions were to have an automatic pause
at the end of a year of simulation and to have supplemental files that could be accessed to explain
aspects of model code, functional relationships, or
parameters. Another suggestion was to have the
ability to export graphics or data files for use in
other applications.
Model Assessment
A recurrent theme in the January 1997 workshop
was the need to critically evaluate or assess the sub-
rent model agreed with historic or observed
changes in vegetation patterns. That is, the group
was unable to test the current set of rules, although
the model can qualitatively match current patterns.
Another gap was how to model longer-term vegetation dynamics, especially how communities respond to chronic disturbances such as flooding
stress and reestablishment following different levels of flooding. Another item discussed was the
need for a nutrient submodel that would link hydrology, sediments, and vegetation dynamics.
Habitats
The habitat group's discussions focused on weaknesses in indicators. Questions arose as to whether
it would be better to model responses of key species
or to develop habitat indicators. Related to the development of habitat indicators, participants were
uncertain whether enough empirical information
exists to model habitat changes based only on depth
and duration of inundation or if other factors need
to be considered. As with the vegetation subgroup,
model credibility was a key source of uncertainty,
and participants cited the need to test the submodel
against historic data.
Human Uses: Navigation and Recreation
Uncertainties related to human activities on the
river included gaps in information and the need to
determine appropriate indicators. For example, are
data available to determine how different management changes would affect key commercial and
recreational uses of the river? Other identified gaps
include the need to model negative interactions between recreational and commercial traffic and a
need to accurately identify changes in dredging requirements associated with different management
schemes.
31.5.4 Confronting Uncertainties
To address the information gaps and uncertainties
described, workshop participants suggested a variety of model refinement, model assessment, and
scenario development activities.
Model Refinement
A number of suggestions were made regarding potential model refinements. These involved modifications to existing submodels, creation of new submodels or routines, and additions to the user
interface.
One debate at the workshop was whether to continue development of two models (i.e., at the pool
469
and system scales) or to concentrate work on a
system-scale model with aggregated or broad-scale
indicators for sediment, vegetation, and habitats.
This was compounded by the difficulties of simulating the hydrology and sediment dynamics at a
1 DO-meter resolution in the pool. At this spatial resolution, the model is extremely slow for somewhat
reliable results (on the order of 10 minutes per simulation year). One option suggested was to use
higher-powered hydraulic models, the results of
which could then be plugged into the pool-scale
model. The debate remains largely open, with no
selection of preference made at the workshop.
A number of modifications were suggested for
each of the submodels, including the following:
1. Compiling data and building better representations of sediment transport in the tributaries
2. Developing a model of sediment-landform
building
3. Economic assessments for potential structural
modifications (e.g., construction and operating
costs of changes to dikes, levees, and channels)
4. Developing a water quality model
5. Developing stage-area relationships for habitat
and vegetation responses
6. Compiling data on recreational and navigational demands
7. Developing composite dredging cost curves
8. Connecting pool models to evaluate cumulative effects
9. Adding remaining pools to the river system
model (i.e., pools 11 to 26)
10. Refining vegetation models with other factors,
such as flooding and water stresses, nutrients,
and temperature
Workshop participants suggested modifying the
user interface to permit comparisons among various output maps. One suggestion was to have the
option for simultaneous display of three maps: two
of the maps would represent results under different scenarios and the third would highlight differences and similarities between the other two maps.
Other suggestions were to have an automatic pause
at the end of a year of simulation and to have supplemental files that could be accessed to explain
aspects of model code, functional relationships, or
parameters. Another suggestion was to have the
ability to export graphics or data files for use in
other applications.
Model Assessment
A recurrent theme in the January 1997 workshop
was the need to critically evaluate or assess the sub-