An accurate assessment of flood inundation and damages on residential and industrial properties requires a comprehensive modeling of simultaneous flood duration, peak, and volumeand identifying their regional hydrologic drivers. The purpose of this study is to advance our understanding from regional hydrologic (percent storage of lakes, ponds, and forested area) and geomorphologic properties (drainage area, mean basin elevation, average basin slope, and perimeter of basin) to quantify the significant drivers that govern the variabilities in theflood duration, peak, and volume at the same time. We used long-term records of streamflow from 479 Hydro-Climatic Data Network (HCDN) sites across the United States that have flood stage information for computing flood duration, peak, and volume (i.e., flood attributes). These stations feature minimum effects in the watershed by the artificial adjustments of flow and human-induced disruptions in the natural stream channels. A joint copula-based modeling of flood attributes embedded in an inference model is proposed here to account for the joint behavior of flood duration, peak, and volume and multivariate dependencies of regional factors. Analysis of spatiotemporal flood attributes and quantifying the flood-generating processes at regional-to-global scales can promote the accuracy of flood damage models and a better projection of flooding risk in the context of socioeconomic factors and ocean-atmospheric complex interactions.