LEARNING CALIBRATION AND TESTING MODEL TO PREDICT FUTURE IMPACTS IN COASTAL ENVIRONMENTS

The complexity of coastal environments and the physical processes involved in their evolution demand integral approaches and the implementation of complex numerical models. To properly evaluate the accuracy of these models, field data are required for their calibration and testing. During recent decades, advanced 2D and 3D numerical models have been developed and used in coastal environments to improve the prediction capabilities at various spatial and temporal scales. Delft3D is one of such models, being a world leading 3D modelling suite to investigate hydrodynamic, sediment transport, morphology and water quality for fluvial, estuarine and coastal environments. Although this model is usually used for teaching purposes, it is usual to omit the processes of calibration and testing to students. This generates the incorrect belief that default parameters of the model reproduce accurately the natural processes.

In this work, we describe our experience using DELFT3D to simulate the dynamics of the Bay of Cádiz (south-western Spain) and to predict the possible effects of future anthropogenic interventions. In particular, we have focused on:
(1) the hydrodynamic evolution, the instantaneous and residual circulations, and the water exchange between the outer and inner portions of the bay;
(2) the sediment transport tendencies based on the calculations of the divergence of the residual current.

During the conference, the sequential methodology used to complete the calibration and testing processes of the DELFT3D will be presented, as well as the main results obtained after one month of simulation.

Moreover, the potential of using this type of studies and methodologies to introduce graduate students into research will be also discussed. In this work, students were responsible for the steps conducting the calibration and testing of the model:
(1) tidal forcing characterization,
(2) definition of the uniform wind and atmospheric forcing,
(3) calibration the model changing the roughness and the wind drag coefficient and
(4) model testing.

The supervisors checked their work, introducing them into the whole research processes: from the obtaining of field data and its treatment, to the dissemination of the results.

With our experience, we conclude that DELFT3D has a sound potential as a new research tool, being the implementation of the model a new research methodology of special interest for many Earth Sciences. Given it simplicity and low cost, it constitutes also a good method to introduce graduate student into research, since they can participate in all the research processes.