Flash floods occurring over complex terrain have claimed the lives of hundreds of people and accounted for hundreds of millions of dollars in time-inflated damages over the years. While early detection and warning of such events has improved over the past few decades, predictive skill of flash floods across a range of time-scales remains critically low. A significant contributor to the lack of forecast skill has resulted from poor (i.e. parameterized) representation of convective processes occurring over complex terrain. However, competing factors such as landscape form, terrestrial hydrologic conditions, and land use also interact to modulate a watershed's response to intense convective events. The response of the watershed to precipitation forcing is highly non-linear in both space and time and therefore demands a physically robust and spatially-distributed framework for adequate diagnosis and prediction of potential flooding hazards as well as spatially-distributed turbulent fluxes to the atmosphere. The system developed under this project aims to improve characterization of terrestrial hydrology during inter-storm periods via hydrologic data assimilation and seeks to advance our understanding of how the quality of QPE and quantitative precipitation forecasts (QPF) from radar-nowcasting and NWP systems, respectively, will impact flood predictions. We have coupled the Noah-distributed hydrological model, a hydrologically-enhanced variant of the vertical Noah land surface model, to the NCAR High Resolution Land Data Assimilation System (HRLDAS) and the WRF model and deployed the coupled system over the Colorado Front Range. The hydrologically-enhanced HRLDAS is used to spin-up land surface and channel flow conditions prior to forecast initiation. We present results from both HRLDAS and coupled model simulations of selected flash flood events from the study domain including the 1997 Fort Collins Flood. Precipitation from the NCAR-autonowcaster (ANC) and Advanced Weather Research and Forecasting (ARW) model and Noah-distributed modeled streamflow are evaluated in the context of preparing flash flood guidance with 0-3 hour and 1-day lead times.