Coastal flooding is a growing concern. Compound coastal flooding considers the joint impacts of marine and hydrologic events characterized by multiple flooding pathways (i.e., high offshore water levels, streamflow, energetic waves, precipitation) acting concurrently. Flood risks are commonly assessed using numerical models or statistical methods. Quantifying event uncertainty is critical to accurate flood risk assessment. This work develops a hybrid statistical-hydrodynamic flood modeling methodology to characterize flood mapping uncertainty in highly urbanized, tidally and wave dominated regions. Uncertainties associated with copula selection, sampling method, data record length, utilized rainfall gauge, and event choice along an isoline were considered. Univariate statistics are analyzed for individual sites and events. Conditional and joint probabilities are developed using a range of copulas, sampling methods, and hazard scenarios. Multiple copulas (Nelsen, BB1, BB5, and Roch-Alegre, Fischer-Koch) consistently passed a Cram\'er-von Mises test and presented similar event pairs, with the exception of the BB5 copula which was often more conservative (i.e., more severe event pairs). Sampling impacts are considered using annual maximum, annual coinciding, wet season monthly maximum, and wet season monthly coinciding sampling. Generally, annual maximum sampling yielded the largest (most severe) event pairs. However, in some cases wet season monthly coinciding sampling suggested higher marine water levels. Uncertainties associated with record length were quantified by creating subsets with different sizes from long data records (~100 years). Significant event pair variability was observed when using short data record lengths, although results stabilized at 70-years. Flood risk estimates using local rainfall gauges significantly varied suggesting microclimatologies must be considered in flood risk analysis. Validated Delft3D-FM hydrodynamic models were developed for multiple urbanized coastal communities. Compound events were simulated to quantify flood mapping uncertainties associated with statistical characterization.