Models of inshore marine systems are well-established with a long history of development and application to local coastal management issues. We can distinguish five broad classes of applications:
• Coastal inundation and changes in sea level
• Coastal circulation
• Sediment dynamics and transport
• Water quality and the fate and impact of pollutants;
• Coastal Ecosystem impacts and responses.
Coastal models of inundation and changes in sea level, including surface waves, tides, tsunami, storm surge and river flooding, are typically 2-D depth-averaged. The Advanced Circulation Model (ADCIRC) is a hydrodynamic circulation numerical model that simulates water level and
current over an unstructured gridded domain.141 ADCIRC can be run as a two-dimensional or
three-dimensional (2-D or 3-D) model for modeling tidally driven and wind and wave driven circulation in coastal waters; forecasting hurricane storm surge and flooding; and for modeling inlet sediment transport/morphology change studies, and dredging/material disposal studies. Because applications of these models address high priority threats to human life and infrastructure, they are widely used operationally and are arguably more mature than other
classes of coastal models.142
Coastal hydrodynamic models predict currents, mixing and transport, and underpin models of water quality, biogeochemical cycles and ecosystems. The spatial resolution, accuracy and skill of 3-D coastal circulation models have all increased with improvements in computational power and numerical methods. Greater attention is now being given to wave-current interactions and
coupled wave-hydrodynamic models.143
A distinguishing feature of coastal systems is the importance of interactions between the water column and the benthos. Models of sediment dynamics represent the dynamic processes controlling exchanges of tracers between the water column and the seabed, and may be used to
predict changes in benthic geomorphology144, changes in the concentrations of suspended fine
sediments, and modeling of other tracers adsorbed to sediment particles or dissolved in
interstitial pore waters.145 Models of sediment dynamics rely on semi-empirical
parameterizations of small-scale processes. While these semi-empirical process formulations are relatively mature and stable, model applications require parameter calibration against observations. The prediction of changes in benthic geomorphology has proven more challenging
than prediction of turbidity, and specialized geomorphological models may be used in engineering applications, especially in zones of wave breaking and swash on open coastlines. Representing the effects of structure and relief due to living benthos on bottom stress and sediment exchanges continues to be a challenge.
Receiving water quality models (RWQMs) have been developed to address the fate and impact of pollutants, including sediments, nutrients and toxins, in coastal systems. Biological process representations in RWQMs have changed from simple semi-empirical representations of biological oxygen demand (BOD), to full biogeochemical process models, representing the cycling of multiple elements (eg nitrogen, phosphorus, carbon) through pelagic and benthic
primary producers and consumers.146 Most RWQMs produce products and indicators related to
water quality, specifically indicators such as turbidity, nutrient levels, phytoplankton biomass and dissolved oxygen. Benthic primary producers such as seagrasses and macroalgae have been included as needed.
Most recently, end-to-end ecosystem models (primary producers to top predators), involving full
food webs, and benthic habitats, have been implemented for a number of coastal systems.147
These are discussed further in Section 5.2.3.
The operational status of these coastal models is uneven. With the exception of ecosystem models, the models are now relatively mature, and have been very widely applied to support tactical and strategic management decisions. There are well-known commercial packages such as
the DELFT-3D and Danish Hydraulic Institute (DHI)-MIKE series of models,148 which are
widely applied, often by commercial consultants and engineers. There are other open research
and community modeling platforms which have also been used in many applications.149
While this subsection and the previous one have focused on more sophisticated modeling approaches, especially data-assimilating dynamical models, there is an ongoing role and need for simpler models, especially for assessment and diagnosis. The LOICZ biogeochemical budget
methodology is an important example.150 This allows users to infer information about flushing,
production and nutrient fluxes from limited data sets using simple inverse techniques. It has been widely applied to estuaries and coastal systems globally. There is again a need to develop better
statistical techniques to put confidence limits around outputs from coastal inverse models.151
In summary then, some coastal modeling for inundation is already operational. Coastal circulation models and RWQMs might be said to be quasi-operational in the sense that modeling platforms and tools exist that are routinely applied to support coastal management. However, there is an opportunity and need to radically improve these tools by adopting some of the techniques and methods already demonstrated for ocean circulation models by GODAE. One can expect this development to be encouraged and supported by GODAE OceanView, but it will require a genuine partnership and integration of the existing ocean modeling communities (GODAE and IMBER) with coastal modeling communities. Coastal physical and
biogeochemical models are not simply inshore, high resolution extensions of ocean models. They include processes and components absent from ocean models, particularly those related to benthic components and benthic-pelagic coupling. This collaboration has already commenced in a number of nations and groups, and it would be reasonable to expect prototype data-assimilating inshore coastal models to be running in a number of locations after 3 years, with demonstration operational systems in 5 years.
Coastal RWQMs are used to predict and assess the impact of diffuse loads from catchments. The state of observation, modeling, assessment and prediction for catchment flows and loads offers strong parallels with the state of RWQMs. There is a strong history of catchment modeling, with a mix of physical process-based and empirical approaches. Observations have typically been sparse and under-sampled in space and time. With the exception of hydrological models for flood prediction, catchment models have been primarily used for assessment and management scenario evaluation. There are currently moves to implement more sophisticated automated observing systems, and to develop data-assimilating catchment models for prediction and assessment. While this report is not directed at observing systems for coastal catchments, it is well recognized that the development of coastal terrestrial and marine observing systems must proceed in an integrated manner.