A Landsat 8 OLI-Based, Semianalytical Model for Estimating the Total Suspended Matter Concentration in the Slightly Turbid Xin’anjiang Reservoir (China)

Yibo Zhang,Yunlin Zhang,Kun Shi,Yong Zha,Yongqiang Zhou,Mingliang Liu
DOI: https://doi.org/10.1109/jstars.2015.2509469
IF: 4.715
2016-01-01
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Abstract:Total suspended matter (TSM) directly determines the underwater light field distribution and thus affects the primary productivity in a water body. Estimation of TSM plays a vital role in monitoring, evaluating, and protecting water quality. Many empirical and semianalytical models have been established for clear open ocean waters or extremely turbid coasts/lakes. However, few are generally applicable to inland, optically complex, deep waters. Using data sets of observed in situ data for the slightly turbid water of the Xin'anjiang Reservoir, we developed a semianalytical algorithm to estimate the TSM concentration for slightly turbid waters (TSM < 25 mg L-1; TSM/Chla < 2.2 x 10(3)) [chlorophyll a (Chla)] using [R-rs(542)(-1) -R-rs(600)(-1)] x R-rs (668), (R-rs: remote sensing reflectance), with a coefficient of determination (R-2) > 0.9 and a normalized root-mean-square error (NRMSE) < 12%. The semianalytical model was then applied to 14 Landsat 8 OLI images from December 2013 to April 2015, with R-2 = 0.85 and NRMSE = 23%, indicating the feasibility of the semianalytical model for estimating TSM. The TSM concentration estimated from Landsat 8 OLI data in the Xin'anjiang Reservoir exhibited a significant spatial and seasonal difference. The spatial heterogeneity, significantly higher in incoming rivers than the main body of the reservoir, was due to watershed inputs and anthropogenic dredging activity. The temporal heterogeneity of TSM, significantly higher in summer and autumn than in winter and spring, was mainly caused by the seasonal rainfall and seasonal growth of phytoplankton. Our study showed that the semianalytical model for Landsat 8 OLI images could be used to quantitatively monitor TSM in slightly turbid, inland waters.
What problem does this paper attempt to address?