Spatial and temporal patterns of suspended sediment transport in nested urban watersheds

Abstract

Patterns of sediment yield have been widely investigated across a variety of scales, but few of these studies have been carried out in urban watersheds. The design of this study represented a unique opportunity to use high-frequency sensor data to compare sediment yields across spatial and temporal scales in an urbanized watershed. Near real-time turbidity and discharge data were collected continuously for four years at five stream gages over three nested spatial scales in the highly impervious 14.2-km2 Dead Run watershed, located in Baltimore County, Maryland, USA. Monthly suspended sediment yields varied by over two orders of magnitude across all stations. Two mid-watershed stations on tributaries to the mainstem of Dead Run are located only about 500 m apart and about 500 m upstream of their confluence, yet displayed observable differences in the timing of their highest sediment yields. Average annual suspended sediment yield was higher at the headwater station with older development (78.2 t/km2/yr) than at the headwater station with more recent development (47.0 t/km2/yr). Average annual yields increased with watershed size from the headwater stations to the mid-watershed stations (87.0 t/km2/yr and 60.4 t/km2/yr). Additionally, yields at the station farthest downstream (81.7 t/km2/yr) were slightly larger than the yield calculated from the sum of both mid-watershed stations (74.6 t/km2/yr), which together represented about 80% of the upstream drainage area.Several additional patterns highlighting spatial heterogeneity of the system were observed. The ratio of sediment yields between the headwater stations increased steadily over time while the runoff ratio remained almost constant. Yields at one mid-watershed station were consistently higher than yields at the other mid-watershed station, despite similarity in runoff totals. Upstream watersheds show year-to-year variation in sediment yield. In contrast, yields at the farthest downstream station show minimal variation. This observation is consistent with the possibility that internal storage and remobilization tend to modulate downstream yields even with spatial and temporal variation in upstream sources. Observable inconsistencies in sediment yields between subwatersheds at monthly time scales suggest that there is a substantial degree of heterogeneity in terms of sediment response to large storm events. Despite significant internal variability, the overall range of sediment yield values at each site was consistent with values reported for multiple urban mid-Atlantic Piedmont watersheds. Comparison with global values indicates that sediment yields from Dead Run are in the same general range as the 25% quartile of urban sediment yields.