Research on distribution of suspended sediment concentration in long time series based on mid-resolution imaging spectrometer data and quantile trend analysis

Abstract

In order to explore the long-term variation trend of suspended sediment concentration (SSC) in the Pearl River Estuary, a multi-voyage in-situ measurements and Moderate-Resolution Imagine Spectrometry (MODIS) Level-1 products were used to establish a quantitative inversion model of the SSC in the Pearl River Estuary. In situ data were gathered with multiple campaigns, while the MODIS sensor was passing over the study area. Then, the Mann–Kendall​ algorithm and the Sen’s slope estimation method were used to analyze the trend of various quantiles SSC. The accuracy of the SSC inversion model was verified by using multiple sets of measured data in which the correlation coefficient was 0.90 and the root mean square error was 4.76. The results showed that: (1) The SSC in the Pearl River Estuary shown a zonal distribution in the northeast–southwest direction, which was gradually reduced from the nearshore to the sea, with an average concentration of more than 90 mg L−1. (2) The SSC in the PRE displayed a decreasing trend, with an averaged SSC reduction rate of about 1.02 mg L−1 yr−1 during 2004–2016. Decreasing rate in high SSC was generally larger than those was in low SSC, with a marked spatial difference. (3) The seasonal variation of dam-induced reduction of the sediment load into the PRE contributed to most of the low-percentile SSC decreases. The water-depth change in the PRE induced by human activities dominated the high-percentile SSC trends. The decrease of the wind speed over the PRE also affected the high-percentile SSC trend.