Tropical Cyclone (TC)-induced flooding is among the most destructive natural disasters worldwide, particularly affecting coastal watersheds through pluvial/fluvial and coastal flooding, as well as their combination (i.e., compound flooding). The increasing frequency of these events underscores the need for improved flood-mapping tools to support risk mitigation. However, many Caribbean islands, located along major TC pathways, lack flood modeling capabilities due to the high data and computational demands of existing methods, hindering flood frequency analysis and forecasting. This presentation highlights an integrated flood modeling framework designed to address these limitations in data-scarce regions. The framework is tested in several coastal watersheds in Puerto Rico's southern and eastern regions and focuses on three key objectives. The study evaluates the performance of five parametric precipitation models (PPMs) to identify the most suitable model for flood modeling applications, comparing them at both rainfall and ground-response levels. Then, the proposed approach couples two physics- based hydrologic and hydraulic models, the Gridded Surface Subsurface Hydrologic Analysis (GSSHA) model and the Storm Water Management Model (SWMM), to generate freshwater inundation maps, which are subsequently integrated with storm surge maps using a deep learning-based approach that accounts for multiple flood-generating mechanisms. Results are also compared with a comprehensive ocean circulation model (ADCIRC) that includes rainfall routing to assess the coupling technique and its approach to assess compound flood. Overall, this study advances flood modeling capabilities in data-scarce coastal regions and improves flood-risk assessment and disaster preparedness in TC-prone areas. The proposed framework provides a scalable methodology that can be adapted to other Caribbean islands and similar coastal environments vulnerable to TC impacts.