Evaluation of Atmospheric Correction Methods of Sentinel-2 for Monitoring Paddy Rice Growth in Cianjur and Klaten Districts
Fadila Muchsin(1a), Supriyatna(2), Adhi Harmoko(3), Indah Prasasti(4)

1,2,3) Faculty of Mathematics and Natural Sciences, Universitas Indonesia, Indonesia
a fadila.muchsin[at]lapan.go.id

4) Remote Sensing Application Center - LAPAN, Indonesia

Abstract

Atmospheric correction is a critical pre-processing step required in quantitative remote sensing applications such as extraction of biophysical variables and multi-temporal analysis. The objective of this study was to evaluate surface reflectance of atmospheric correction methods base on scene-based empirical namely Dark Object Subtraction (DOS) and the radiative transfer methods namely the Second Simulation of Satellite Signal in the Solar Spectrum (6S). We applied these methods to Sentinel-2 images above rice fields in Cianjur and Klaten Regency, and compared surface reflectance from atmospheric correction methods with ground measured reflectance base on spectral response and vegetation indices (NDVI, EVI, SAVI and MSAVI). For that purpose, generative phase of rice growth was analysed by using statistical analysis. The result showed that the 6S method has better accuracy than DOS.

Keywords: Atmospheric correction, Sentinel-2, Rice growth phases, Statistical analysis, Vegetation indices

Topic: Earth and Planetary Sciences