Hydrologic exchange flux (HEF) is an important hydrologic component in river corridors that includes both bidirectional (hyporheic) and unidirectional (gaining/losing) surface water‐groundwater exchanges. Quantifying HEF rates in a large regulated river is difficult due to the large spatial domains, complexity of geomorphologic features and subsurface properties, and the great stage variations created by dam operations at multiple time scales. In this study, we developed a method that combined numerical modeling and field measurements for estimating HEF rates across the riverbed in a 7 km long reach of the highly regulated Columbia River. A high‐resolution computational fluid dynamics (CFD) modeling framework was developed and validated by field measurements and other modeling results to characterize the HEF dynamics across the riverbed. We found that about 85% of the time from 2008 to 2014 the river was losing water with an annual average net HEF rates across the riverbed (Qz) of −2.3 m3 s−1 (negative indicating downwelling). June was the only month that the river gained water, with monthly averaged Qz of 0.8 m3 s−1. We also found that the daily dam operations increased the hourly gross gaining and losing rate over an average year of 8% and 2%, respectively. By investigating the HEF feedbacks at various time scales, we suggest that the dam operations could reduce the HEF at seasonal time scale by decreasing the seasonal flow variations, while also enhance the HEF at subdaily time scale by generating high‐frequency discharge variations. These changes could generate significant impacts on biogeochemical processes in the hyporheic zone.