Scientists and policymakers increasingly recognize that headwater regions contain numerous temporary streams that expand and contract in length, but accurately mapping and modeling dynamic stream networks remain a challenge. Flow intermittency sensors offer a relatively new approach to characterize wet stream length dynamics at high spatial and temporal resolutions. We installed 51 flow intermittency sensors at an average spacing of 40 m along the stream network of a high-relief, headwater catchment (33 ha) in the Valley and Ridge of southwest Virginia. The sensors recorded the presence or absence of water every 15 min for 10 months. Calculations of the wet network proportion from sensor data aligned with those from field measurements, confirming the efficacy of flow intermittency sensors. The fine temporal scale of the sensor data showed hysteresis in wet stream length: the wet network proportion was up to 50% greater on the rising limb of storm events than on the falling limb for dry antecedent conditions, at times with a delay of several hours between the maximum wet proportion and peak runoff at the catchment outlet. Less stream length hysteresis was evident for larger storms with higher event and antecedent precipitation that resulted in peak runoff > 15 mm/day. To assess spatial controls on stream wetting and drying, we performed a correlation analysis between flow duration at the sensor locations and common topographic metrics used in stream network modeling. Topography did not fully explain spatial variation in flow duration along the stream network. However, entrenched valleys had longer periods of flow on the rising limbs of events than unconfined reaches. In addition, large upslope contributing areas corresponded to higher flow duration on falling limbs. Future applications that explore the magnitude and drivers of stream length variability may provide further insights into solute and runoff generation processes in headwater regions.
Keywords: Flow intermittency; Hysteresis; Stream length; Temporary streams.