Abstract
Data from Landsat-8 and Sentinel-2A/2B are often combined for terrestrial monitoring because of their similar spectral bands. The bidirectional reflectance distribution function (BRDF) effect has been observed in both Landsat-8 and Sentinel-2A/2B reflectance data. However, there is currently no definition of solar zenith angle (θsz) that is suitable for the normalization of the BRDF-adjusted reflectance from the three sensors’ combined data. This paper describes the use of four machine learning (ML) models to predict a global θsz that is suitable for the normalization of bidirectional reflectance from the combined data in 2018. The observed θsz collected globally, and the three locations in the Democratic Republic of Congo (26.622°E, 0.356°N), Texas in the USA (99.406°W 30.751°N), and Finland (25.194°E, 61.653°N), are chosen to compare the performance of the ML models. At a global scale, the ML models of Support Vector Regression (SVR), Multi-Layer Perception (MLP), and Gaussian Process Regression (GPR) exhibit comparably good performance to that of polynomial regression, considering center latitude as the input to predict the global θsz. GPR achieves the best overall performance considering the center latitude and acquisition time as inputs, with a root mean square error (RMSE) of 1.390°, a mean absolute error (MAE) of 0.689°, and a coefficient of determination (R2) of 0.994. SVR shows an RMSE of 1.396°, an MAE of 0.638°, and an R2 of 0.994, following GPR. For a specific location, the SVR and GPR models have higher accuracy than the polynomial regression, with GPR exhibiting the best performance, when center latitude and acquisition time are considered as inputs. GPR is recommended for predicting the global θsz using the three sensors’ combined data.
Highlights
The polar orbit satellite Landsat-8, launched by National Aeronautics and Space Administration (NASA) [1,2] and the Sentinel-2A and 2B satellites, launched by ESA [3], have similar spectral bands
The Support Vector Regression (SVR) and Gaussian Process Regression (GPR) models have higher accuracy than the polynomial regression, with GPR exhibiting the best performance, when center latitude and acquisition time are considered as inputs
Both Landsat-8 and Sentinel-2A/2B have sun-synchronous polar orbits. Their view angles are ±10.3◦ (Sentinel-2) and ±7.5◦ (Landsat-8) from the nadir view when acquiring observations, resulting in non-Lambertian surface directional reflectance effects. The magnitude of these effects varies as a function of geometry, and is usually described by the bidirectional reflectance distribution function (BRDF)
Summary
The polar orbit satellite Landsat-8, launched by National Aeronautics and Space Administration (NASA) [1,2] and the Sentinel-2A and 2B satellites, launched by ESA [3], have similar spectral bands Together, these three satellites provide 10–30 m moderate spatial resolution multi-spectral global coverage. Compared with a single satellite, the combination of the three satellites, taking advantage of their complementary revisit interval patterns, provides a 2.9-day global median average revisit interval [4,5] This would benefit numerous remote sensing applications, such as deforestation [6], fire monitoring [7], agriculture dynamics [8], and ice velocity detection [9].
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
