Angular-Based Radiometric Slope Correction for Sentinel-1 on Google Earth Engine

Andreas Vollrath,Adugna Mullissa,Johannes Reiche

doi:10.3390/rs12111867

Andreas Vollrath, Adugna Mullissa + Show 1 more

Open Access

https://doi.org/10.3390/rs12111867

Copy DOI

Abstract

This article provides an angular-based radiometric slope correction routine for Sentinel-1 SAR imagery on the Google Earth Engine platform. Two established physical reference models are implemented. The first model is optimised for vegetation applications by assuming volume scattering on the ground. The second model is optimised for surface scattering, and therefore targeted at urban environments or analysis of soil characteristics. The framework of both models is extended to simultaneously generate masks of invalid data in active layover and shadow affected areas. A case study, using openly available and reproducible code, exemplarily demonstrates the improvement of the backscatter signal in a mountainous area of the Austrian Alps. Furthermore, suggestions for specific use cases are discussed and drawbacks of the method with respect to pixel-area based methods are highlighted. The radiometrically corrected radar backscatter products are overcoming current limitations and are compliant with recent CEOS specifications for SAR backscatter over land. This improves a wide range of potential usage scenarios of the Google Earth Engine platform in mapping various land surface parameters with Sentinel-1 on a large scale and in a rapid manner. The provision of an openly accessible Earth Engine module allows users a smooth integration of the routine into their own workflows.

Highlights

Google Earth Engine (GEE) was one of the first web-based platforms adopting the paradigm shift of providing Earth Observation Analysis-Ready-Data (ARD) on a Big Data infrastructure for rapid large-scale analytics of geo-spatial datasets [1]
This paper provides the implementation of the two angular-based reference models for the radiometric slope correction and the masking of active layover and shadow areas on GEE
This study demonstrates that angular-based radiometric slope corrections for Sentinel-1 imagery on the GEE platform are feasible by using the two established physical reference models described

Summary

Introduction

Google Earth Engine (GEE) was one of the first web-based platforms adopting the paradigm shift of providing Earth Observation Analysis-Ready-Data (ARD) on a Big Data infrastructure for rapid large-scale analytics of geo-spatial datasets [1]. The provision of ARD supersedes the cumbersome and computationally intensive burden of preprocessing low-level satellite data for the user. It allows for immediate access of the imagery and let the user focus on the actual information extraction. This is especially attractive for working with Sentinel-1 imagery, as the complexity of preprocessing Synthetic Aperture Radar (SAR) data is one of the main reasons for its slow uptake by a wider user community [2]. Before the ingestion of the Sentinel-1 data onto the GEE platform, basic processing steps (noise removals, calibration and geocoding) are applied [3], resulting in geometrically terrain-corrected ARD products. To comply with the ARD data standards for land, suggested by the Committee on Earth Observation Satellites (CEOS) [4], the radiometric slope correction and the provision of an invalid data mask over areas affected by layover and shadow are required as well

Methods

Results

Discussion

Conclusion