Testing and Calibration Scenarios of a Hydrodynamic Model of Mosquito Lagoon, Florida

The objective of this paper is to present preliminary results of two calibration runs and an additional test for a hydrodynamic model of Mosquito Lagoon, Florida. Mosquito Lagoon (ML) is a sub-basin of the Indian River Lagoon (IRL) located along the East Central Florida coast. The configuration of this shallow coastal lagoon and the restricted connection to the ocean makes it a wind dominated system. Since ML is directly linked to the IRL through a narrow canal (Haulover Canal or HOC) the model grid extends into the northern IRL region. This work is focused on examining five different scenarios in order to understand the response of the system to observed and synthetic input data applied at the boundaries. All simulations are for a period of 180 days during 1998 (May-11 to November-5) and all other input parameters are kept equal. The two calibration simulations use the observed time series as input and differ only in their representation of the west section of the IRL region as either an open (case-CO) or closed boundary condition (case-CC). Results for these two scenarios indicate that water elevation record is best reproduced at HOC and at a second calibration station in the IRL (station c-IRL) in case-CO. Station c-IRL is located in the IRL approximately 5km southwest from the HOC. Overall, the mean trend of the modeled surface salinity and water temperature for HOC and c-IRL stations is also captured in the CO set up. It is therefore noted that assimilation of the south boundary (SB) forcing should be extended to cells along the western boundary of the IRL. The other model test includes three scenarios that have synthetic input time series at each open boundary along with differing wind conditions: observed-wind (SOW), no-wind (SNW), and minimal-wind (SMO). The synthetic water elevation data consist of a simplified sinusoidal signal combined with a low frequency component. Observed water elevation records of both calibration stations (HOC and c-IRL) mimic that of the south boundary (SB) signal. Therefore under synthetic boundary conditions, modeled data from HOC and c-IRL are compared to that of the synthetic SB data. Among these three synthetic scenarios only minor differences occurred in the amplitude range and along with an offset of the mean trend with respect to the water elevation. Overall in these three cases the water level oscillations are under-predicted but in phase.