Impact of Various Atmospheric Corrections on Sentinel-2 Land Cover Classification Accuracy Using Machine Learning Classifiers

Luka Rumora,Damir Medak,Mario Miler

doi:10.3390/ijgi9040277

Luka Rumora, Damir Medak + Show 1 more

Open Access

https://doi.org/10.3390/ijgi9040277

Copy DOI

Abstract

Atmospheric correction is one of the key parts of remote sensing preprocessing because it can influence and change the final classification result. This research examines the impact of five different atmospheric correction processing on land cover classification accuracy using Sentinel-2 satellite imagery. Those are surface reflectance (SREF), standardized surface reflectance (STDSREF), Sentinel-2 atmospheric correction (S2AC), image correction for atmospheric effects (iCOR), dark object subtraction (DOS) and top of the atmosphere (TOA) reflectance without any atmospheric correction. Sentinel-2 images corrected with stated atmospheric corrections were classified using four different machine learning classification techniques namely extreme gradient boosting (XGB), random forests (RF), support vector machine (SVM) and catboost (CB). For classification, five different classes were used: bare land, low vegetation, high vegetation, water and built-up area. SVM classification provided the best overall result for twelve dates, for all atmospheric corrections. It was the best method for both cases: when using Sentinel-2 bands and radiometric indices and when using just spectral bands. The best atmospheric correction for classification with SVM using radiometric indices is S2AC with the median value of 96.54% and the best correction without radiometric indices is STDSREF with the median value of 96.83%.

Highlights

Preprocessing of satellite data plays one of the key roles in result analysis
The National Aeronautics and Space Administration (NASA) provide LaSRC and LEDAPS corrections for the Landsat satellite mission, the European Space Agency (ESA) provides Sentinel-2 atmospheric correction for the Sentinel-2 satellite mission, Planet, Inc. provides their own atmospheric correction based on second simulation of a satellite signal in the solar spectrum (6S) for PlanetScope satellite imagery
t2insepl-2ecstpreacltrsailgsnigantuatruersesoofffifivve attmmoospsphehreicriccorcroerctrieocntsioands atonpdotfotpheoaftmthoespahtemreosphere (TOA) has the highest values for water class, while standardized surface reflectance (STDSREF) has the highest value for the other four classes for all dates in 2017

Summary

Introduction

Preprocessing of satellite data plays one of the key roles in result analysis. Atmospheric correction is one of the most important preprocessing steps because it can affect the final result. The National Aeronautics and Space Administration (NASA) provide LaSRC and LEDAPS corrections for the Landsat satellite mission, the European Space Agency (ESA) provides Sentinel-2 atmospheric correction for the Sentinel-2 satellite mission, Planet, Inc. provides their own atmospheric correction based on second simulation of a satellite signal in the solar spectrum (6S) for PlanetScope satellite imagery. This can be confusing for end users when comparing multiple sensors. Model-based atmospheric corrections relies on radiative modeling and requires data about atmospheric optical characteristics of image acquisition time [3]

Objectives

Methods

Results

Conclusion