River plume monitoring in a deep valley reservoir using HJ-1 A/B images

Abstract

River discharge carries suspended matter into lakes, reservoirs, estuaries, and coasts, resulting in obvious plumes in river mouths. River plumes are more likely to form in valley reservoirs due to the steep slopes in the basin, which may accelerate fluvial transport from upstream to the river mouth, causing a considerable particles load and increasing the appearance times and scale of river plumes during rainfall events. Information on the intensity and evolution of river plumes in valley reservoirs has important implications for understanding nearshore geomorphological processes and dynamics in the water quality surrounding river mouths. Taking the Xin’anjing Reservoir, a typical deep valley reservoir, as a case study, a plume index based on the relative reflectance height in the red band of HJ-1 A/B images was first proposed. Subsequently, we used seventeen HJ-1 A/B images during 10 heavy rainfall events from 2009 to 2014 to characterize the spatial–temporal variability in river plumes and illustrate the relation of the river plumes and total suspended matter (TSM) concentration to the amount of rainfall. The results indicated that 10 heavy rainfall events caused different plume areas varying from 41.26 to 148.12 km2, with an average area of 76.45 ± 35.03 km2. Significant positive correlations were observed between plume areas, TSM concentrations and rainfall amount (P < 0.005). The plumes would simultaneously emerge when the rapid surface flow reached the river mouth but increase to the maximum size after 2–3 days of these heavy rainfalls. These results confirmed that the HJ-1 A/B images combined with our proposed plume index could detect and monitor the formation and diffusion of river plumes in valley reservoirs during heavy rainfall events.