Temporal Scaling of Hydrological Signals Reveals Forcing Persistence and Watershed Filtering Across the Contiguous U.S.

Abstract Streamflow temporal persistence reflects both hydrological inputs and watershed filtering, but their relative contributions remain difficult to distinguish. We analyzed spectral scaling in precipitation, evapotranspiration, effective precipitation, streamflow, and the effective precipitation‐to‐streamflow transfer function across 671 contiguous U.S. watersheds. Streamflow and transfer‐function scaling exponents vary more widely across watersheds than those of the hydrological inputs. Combining effective‐precipitation and transfer‐function exponents identifies three regimes: rapid‐response watersheds with weak input persistence and filtering, widespread across the central and eastern U.S.; forcing‐persistent watersheds with persistent inputs but limited filtering, mainly in the West; and hydrologically buffered watersheds with strong filtering, concentrated in the western mountains, northern regions and Southeast. Western mountain buffered watersheds have greater snow influence and baseflow contributions, whereas southeastern buffered watersheds have little snow but more permeable soils and subsurface materials. The two exponents distinguish input persistence from watershed filtering in shaping streamflow persistence.