Water surface temperature estimation from Landsat 7 ETM+ thermal infrared data using the generalized single-channel method: Case study of Embalse del Río Tercero (Córdoba, Argentina) Original Research Article

Monitoring of warm distribution in water is fundamental to understand the performance and functioning of reservoirs and lakes. Surface water temperature is a key parameter in the physics of aquatic systems processes since it is closely related to the energy fluxes through the water–atmosphere interface. Remote sensing applied to water quality studies in inland waterbodies is a powerful tool that can provide additional information difficult to achieve by other means. The combination of good real-time coverage, spatial resolution and free availability of data makes Landsat system a proper alternative. Many papers have developed algorithms to retrieve surface temperature (principally, land surface temperature) from at-sensor and surface emissivity data. The aim of this study is to apply the single-channel generalized method (SCGM) developed by for the estimation of water surface temperature from Landsat 7 ETM+ thermal bands. We consider a constant water emissivity value (0.9885) and we compare the results with radiative transfer classic method (RTM). We choose Embalse del Río Tercero (Córdoba, Argentina) as case study because it is a reservoir affected by the outlet of the cooling system of a nuclear power plant, whose thermal plume could influence the biota’s distribution and biodiversity. These characteristics and the existence of long term studies make it an adequate place to test the methodology. Values of estimated and observed water surface temperatures obtained by the two compared methods were correlated applying a simple regression model. Correlation coefficients were significant (R2: 0.9498 for SCGM method and R2: 0.9584 for RTM method) while their standard errors were acceptable in both cases (SCGM method: RMS = 1.2250 and RTM method: RMS = 1.0426). Nevertheless, SCGM could estimate rather small differences in temperature between sites consistently with the results obtained in field measurements. Besides, it has the advantage that it only uses values of atmospheric water vapor and it can be applied to different thermal sensors using the same equation and coefficients.