Evaluation of the suitability of MODIS, OLCI and OLI for mapping the distribution of total suspended matter in the Barra Bonita Reservoir (Tietê River, Brazil)

• Autor(a) principal: Bernardo, Nariane [UNESP]
• Data de Publicação: 2016
• Outros Autores: Watanabe, Fernanda [UNESP], Rodrigues, Thanan [UNESP], Alcântara, Enner [UNESP]
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1016/j.rsase.2016.06.002; http://hdl.handle.net/11449/228177
• topic: Bio-optical models; Case-2 water; Inland waters; Multispectral sensors

The objective of this work was to evaluate the suitability of three remote sensors, namely, the Moderate Resolution Imaging Spectroradiometer (MODIS), the Operational Land Imager (OLI), and the Ocean Land Color Instrument (OLCI), for estimating total suspended matter (TSM) concentrations in the Barra Bonita reservoir. Although remote sensors have been widely explored for ocean and inland water applications in Brazilian reservoirs, a thorough comparison of sensors as a TSM monitoring tool has yet to be conducted. OLI data have been used for inland waters, but few studies on Brazilian aquatic systems have been performed. MODIS data were investigated due to their daily coverage, and OLCI data (scheduled for launch in December 2015) were analyzed because of their spatial (better than MODIS) and temporal (lower than OLI) resolution. In situ hyperspectral measurements were used as input to simulate MODIS, OLI and OLCI spectral bands while considering the spectral response function for each sensor. Simulated data and TSM concentrations were tuned to generate regional models using linear and non-linear regressions. The models were assessed using the coefficient of determination (R2), which had a range of between 0<R2<0.83, and the Root Mean Squared Error (RMSE), which ranged between 20.00%<RMSE <68.31%. The best TSM model retrievals were obtained from the OLCI (using the spectral band in the near-infrared range) and OLI (using the green spectral band) simulated data.