Characterizing spatial connectivity of hydrological and sediment-transport processes at the watershed scale for an extreme storm event

Abstract:

Directly measuring sediment loads at multiple spatial scales within a watershed for better management planning is often practically limited, particularly during extreme storm events. In this study, we used a physically based watershed model, Dynamic Watershed Simulation Model (DWSM), to obtain these values, based on which we examined spatial patterns of sediment dynamics in a watershed with the area of 311 km²for an extreme storm event. By dividing the watershed into 42 overland elements and 21 connected stream segments and fitting the hydrograph and sedigraph of at its outlet, we tested the simultaneously predicted hydrological and sediment variables in these overlands and stream segments using the measured event runoff volume (V) and sediment yield (SSY_e) at both a sub-watershed and the outlet of the study watershed. Then, using these predicted variables, we examined the relationship between SSY_e and area (A), precipitation (P), mean slope (S), soil K-factor (K), and percent cropland(%L) for all overland elements (i.e., the small spatial scale), and the changes of SSY_e with peak flow (Q_peak) and event runoff depth (h). We found that SSY_e at this scale cannot be explained well by these variables. Next, we investigated the relationships between SSY_e and A, sediment delivery ratio (SDR), and Q_peak of a series of first-order and nested sub-watersheds, respectively, which led to good correlation. The different spatial patterns between two spatial scales suggest that sediment transport is more controlled by surface conditions of the overlands, while by in-channel hydraulics at the sub-watershed scale for this event. Further analysis also revealed that both flow and sediment-transport regimes were changed during this event. Our study demonstrated a new means of characterizing spatially and temporal variable hydrological and sediment connectivity.