A semi-analytical approach for detecting suspended particulate composition in complex turbid inland waters (China)

• A semi-analytical approach to detect particle composition in turbid inland waters.
• Successful application of the developed model to new Chinese satellite HJ1A/HSI data.
• Providing a new view of understanding particle composition for turbid inland waters.

Remote sensing detection of suspended particulate composition is of importance for understanding variability of water optical properties, especially in complex-material loading turbid inland waters. To achieve this goal, this study carried out multiple in situ bio-optical investigations during the period of 2008 ~ 2010, covering Lake Taihu, Chaohu, Dianchi, and Three Gorges reservoir in China. Using the collected in situ observed datasets, a semi-analytical approach was developed based on theoretical derivation of the water radiative transfer theory. This algorithm showed relatively high calibration accuracy, and was further evaluated by means of an independent validation dataset, producing low predictive errors (mean absolute percentage error, MAPE = 29.5%, and root mean square error, RMSE = 0.23, 10− 3). Then, the developed algorithm was applied into a new hyperspectral satellite data, namely the HJ1A/HSI image, mapping spatio-temporal distribution of suspended particulate composition for Lake Taihu waters. The findings in the present study indicate that the proposed approach has a great potential in detecting particulate composition of water bodies. This study provides us improved understanding of particulate composition characteristics and corresponding spatio-temporal distribution for complex turbid inland waters.