Geoscience Reference
In-Depth Information
dispersive model) may be appropriate (Yeh et al., 1996; Liu et al., 2008), and in the latter case (i.e.,
nondispersion), it is appropriate to use the fully nonlinear shallow water wave theory to model
tsunami propagation. Because the tsunami wavelength becomes shorter on the shelf than in
the abyssal plain, and the bathymetry becomes more complex, a iner grid size is required for
the simulation model in the continental shelf (Scientiic Committee on Oceanic Research, 2001).
Once the tsunami commences the run-up onto the beach, hydrodynamic models need to
consider and parameterize natural and human conigurations (e.g., buildings, trees, mounds,
roads) by assigning proper values to the friction factors. More sophisticated numerical models
(e.g., based on 3-D Navier-Stokes model) may need to be implemented to assess the detailed
effect of tsunami forces on structures (Yeh et al., 1996). Although modelers acknowledge the
importance of natural and human conigurations in inluencing tsunami run-up, there is still
signiicant uncertainty of their effect on tsunami propagation and inundation models.
More than 10 years ago, NOAA/PMEL attempted to initiate a tsunami community modeling
activity to develop a functionality to coordinate and improve hydrodynamic models speciically
used for tsunami prediction practice (Titov et al., 1999). The activity was intended to provide
sharing of community models and databases, which would promote accelerating improvement
of the models and applications. Because of lack of support this activity did not materialize at
PMEL. However, similar concepts have been carried on elsewhere. The Arctic Regional Super-
computing Center (ARSC) and Northwest Alliance for Computational Science & Engineering
(NACSE) have developed the “Tsunami Computational Portal” (https://tsunamiportal.arsc.edu/).
The portal provides a common interface for running and comparing the results of differ-
ent hydrodynamic models for tsunamis and provides uniform access to the topographic and
bathymetry data. Another instance is a model repository within one of NSF's EVOs (Engineering
Virtual Organization, Inundation Science & Engineering Cooperative (ISEC): http://isec.nacse.
org/). The source codes of hydrodynamic models and full documentation are available in the
repository, and the site includes a discussion forum. There appears to be a trend toward making
hydrodynamic models, once proprietary codes are developed by individuals or organizations,
more transparent and more easily shared, facilitating further improvement and applications.
Conclusion:
Rapid improvements in the global and regional bathymetry datasets are
substantial and provide adequate data for inundation modeling efforts in most places
throughout the nation. However, some regions, such as coastal Alaska, still lack bathymetric
data of suficient quality and quantity, which impedes their hazard and evacuation
mapping efforts. Regarding hydrodynamic models, the accuracy and precision of run-up
models are hampered by uncertainties associated with the effects of coastal bathymetry,
topography, vegetation, and structures.
Recommendations:
To improve tsunami inundation modeling, the NTHMP, through
workshops convened with the USGS and academic institutions, should periodically
review progress in hydrodynamic models, including peer-reviews of modeling codes and
discussions on how to better incorporate variations in coastal bathymetry, topography,
vegetation, and buildings into inundation models.
