Event‐Scale Controls on Suspended Sediment–Discharge Hysteresis in an Agricultural Headwater Watershed of the Upper Blue Nile Basin, Ethiopia

ABSTRACT

In rivers, suspended sediment concentration (SSC) may not respond instantly or proportionally to discharge (Q), peaking before or after peak Q during runoff events. In the Upper Blue Nile Basin, sediment load is highly episodic and storm‐driven, yet these dynamics remain poorly understood. This study characterizes SSC–Q hysteresis using 10 min–resolution hydrometric and sediment data from 48 events. Event‐based hysteresis was characterized using integrated quantitative indices and qualitative classification, revealing five patterns: anticlockwise (ACW), clockwise (CW), two figure‐of‐eight patterns (F8‐ACW/CW and F8‐CW/ACW), and a complex single‐peak pattern. In the Angodigud watershed, the ACW pattern was the most frequent, occurring under low‐discharge conditions (< 50 m ³  s ⁻¹ ) from early July to early September, accounting for 56% of all events and contributing 38% of the total sediment load, indicating delayed upland sediment supply, whereas CW pattern prevailed in September–October as vegetation limited hillslope inputs and channel sources dominated. The F8‐ACW/CW pattern was the dominant sediment contributor, accounting for 55% of the total sediment load and occurring in 23% of all events; its probability of occurrence increased when peak discharge exceeded 50 m ³  s ⁻¹ . In this study, peak SSC occurred before, after, and concurrently with peak Q in 19%, 44% and 38% of events, respectively. Despite this phase variability, SSC and Q were near‐synchronized (HI = −0.17), whereas peak Q slightly preceded peak SSC by 2% of the runoff duration, indicating a slightly muddier recession limb (HI _{_Lloyd}  = −0.17). Principal component analysis identified event, rising‐ and falling‐limb, maximum, and area‐normalized suspended sediment yields as the main drivers of hysteresis variability. High‐resolution SSC–Q characterization improves understanding of sediment dynamics and supports event‐based management, as well as the evaluation and improvement of sediment‐routing models in small, data‐limited agricultural watersheds.