Calibration of the Normalized Radar Cross Section for Sentinel-1 Wave Mode

Type Article
Acceptance Date 2018 IN PRESS
Language English
Author(s) Li Huimin1, Mouche Alexis2, Stopa Justin E.3, Chapron Bertrand4
Affiliation(s) 1 : Univ. Brest, CNRS, IRD, Laboratoire d'Océanographie Physique et Spatiale, L'Institut Français de Recherche pour l'Exploitation de la Mer, IUEM, 29280 Brest, France, and also with IMT Atlantique, Lab-STICC, UBL, 29238 Brest, France
2 : Univ. Brest, CNRS, IRD, Laboratoire d'Océanographie Physique et Spatiale, L'Institut Français de Recherche pour l'Exploitation de la Mer, IUEM, 29280 Brest, France.
3 : Department of Ocean Resources and Engineering, University of Hawaii at Manoa, Honolulu, HI 96822 USA.
4 : Univ. Brest, CNRS, IRD, Laboratoire d'Océanographie Physique et Spatiale, L'Institut Français de Recherche pour l'Exploitation de la Mer, IUEM, 29280 Brest, France.
Source IEEE Transactions on Geoscience and Remote Sensing (0196-2892) (Institute of Electrical and Electronics Engineers (IEEE)) In Press
DOI 10.1109/TGRS.2018.2867035
Abstract

Sentinel-1 (S-1) is a two-satellite constellation for continuity of operational synthetic aperture radar (SAR) observations. Wave mode (WV) is the default mode over open ocean for S-1 to monitor global ocean waves and wind field. Therefore, proper radiometric calibration is essential to accurately infer these geophysical quantities. Based on the global data set acquired by S-1A WV, assessment of normalized radar cross section (NRCS) is carried out through comparison with CMOD5.N predictions over open ocean. The calibration accuracy quantified by NRCS residuals between SAR measurements and CMOD5.N demonstrates distinct features for two incidence angles (23.8° and 36.8°). Particularly, NRCS at 23.8° is overall consistent with CMOD5.N, while NRCS at 36.8° displays great deviation. Two recalibration methods are then implemented by examining the backscattering profile over Amazon rain forest and ocean calibration. Both methods show the necessity for recalibration and obtain comparable correction factors for WV1 and WV2, respectively. The NRCS residuals by applying both methods are significantly reduced toward zero. By comparison, ocean calibration is more efficient and practical to implement.

Full Text
File Pages Size Access
9 4 MB Access on demand
Author's final draft IN PRESS 11 5 MB Open access
Top of the page