Salinity Models for Florida Bay – Status and Recommendations
The assessment of benefits to Florida Bay from restoring the hydrology of the Everglades depends absolutely on salinity projections obtained from validated, predictive models.
Research and development efforts have pursued two radically different approaches to salinity prediction. These are the detailed representation of tide and wind-driven water movements in a hydrodynamic model, i.e. the RMA-10 model, and the relatively coarse, tidally averaged representation by a mass-balance model, i.e. FATHOM. Both models are operational and can be used to calculate salinity values within Florida Bay.
More work is needed to document the predictive capability of the models. Currently, neither model has the confidence of the technical community. This is the main conclusion of a technical review of the models and from our experience in applying these models to simulate the response of Florida Bay salinity based on hydrologic scenarios from the Restudy Project. Confidence of the technical community is a prerequisite for the application of model in the planning and assessment of restoration activities.
The PMC recommends that a calibration/validation exercise be conducted to demonstrate the ability of each model to reproduce the observed response of salinity in Florida Bay to historical variation in climate and freshwater runoff. This will require, first, the assembly of a standard data set, i.e. runoff, climate, sea level and salinity, that characterizes both the forcing and the response of the Bay during a period in which significant variation occurred in salinity. To build confidence in the models among research scientists and resource managers, both groups must contribute to defining the standard data set. In other words, this task cannot be delegated to the modelers to do working alone. Rather, the standard data set should be compiled based on broad consultation.
Design of the calibration/validation exercise requires active guidance and participation by eventual users of the models. Users include scientists and engineers who will apply the models in the planning and assessment of restoration activities. Model validation must take into account the aspects of salinity variation that are considered important or essential for the models to capture. For example, what salinity patterns should be reproduced and to what degree of detail? In addition to their predictive capability, the eventual users of the models may wish to include criteria related to utility, i.e. a model’s ability to be used within the planning and assessment process.
The task of assembling the standard data set for model calibration/validation must be accomplished whether agencies decide to proceed with the evaluation of the present models or to pursue development of other salinity models for Florida Bay. Therefore, work on this task should proceed immediately to avoid further delay in the development of these tools.