Flood hazard planning requires the accurate estimation of total water elevation due to predicted tide, surge, and wave runup to design flood protection structures and improve coastal risk planning for severe storms. The beach geomorphology and nearshore hydrodynamic conditions impact the conclusive flood inundation mapping in complex environments. The conventional approaches of flood modeling are limited due to either i) simple static estimates, ii) the application of a coupled circulation and phase-averaged wave models in coarse resolution, iii) failing to calibrate and validate with in-situ data, or iv) not considering sea-level rise projections in mapping the flood extent. We used a fully nonlinear Boussinesq wave model (FUNWAVE-TVD) on the nearshore area with a high-resolution grid to determine total water elevation on the shores. The model adopted the boundary conditions from coupled circulation and wave model of the Long Island Sound (Liu et al., 2020).We applied the model to Branford, Norwalk, and New Haven, CT to represent the complex topo-bathymetric features and the structural interference in shallow water wave dynamics. We compared the model results with the FEMA base flood elevation, the North Atlantic Coast Comprehensive Study (NACCS), and (Liu et al., 2020)’s Long Island Sound FVCOM-SWAVE model. The FUNWAVE-TVD model is found to model wave processes more accurately in shallow water regions compared to the empirical equation application of FEMA and coupled circulation-phase averaged model application of NACCS and FVCOM-SWAVE. We also examined local sea-level rise predictions of storms with 1% and 10% annual exceedance probability by the year 2050 in Connecticut and found that the flood extent of these two storms showed little to no difference. We suggest the planning approaches should consider the increase in the frequency of the storms in the predicted inundation zones due to sea-level rise.