In situ validation of MERIS marine reflectance off the southwest Iberian Peninsula: assessment of vicarious adjustment and corrections for near-land adjacency

• Published in: International journal of remote sensing 2014-03, Vol.35 (6), p.2347-2377
• Main Authors: Cristina, Sónia Cláudia Vitorino, Moore, Gerald Francis, Goela, Priscila Raquel Fernandes Costa, Icely, John David, Newton, Alice
• Format: Article
• Language: English
• Subjects: Animal, plant and microbial ecology; Applied geophysics; Atmospheric correction; Biological and medical sciences; Buoys; Calibration; Coastal waters; Corrections; Deep water; Earth sciences; Earth, ocean, space; Exact sciences and technology; Fundamental and applied biological sciences. Psychology; General aspects. Techniques; Histograms; Imaging techniques; Internal geophysics; Microprocessors; Ocean colour; Oceans; Offshore; Optical properties; Optical thickness; Procedures; Reflectance; Regression analysis; Reprocessing; Resolution; Scattering; Stations; Statistical methods; Statistics; Teledetection and vegetation maps; Wavelength; Wavelengths
• ISSN: 0143-1161; 1366-5901
• DOI: 10.1080/01431161.2014.894657

Water-leaving reflectance (ρ w ) data from the European Space Agency ocean colour sensor Medium Resolution Imaging Spectrometer (MERIS) was validated with in situ ρ w between October 2008 and November 2011, off Sagres on the southwest coast of the Iberian Peninsula. The study area is exceptional, since Stations A, B, and C at 2, 10, and 18 km offshore are in optically deep waters at approximately 40, 100, and 160 m, respectively. These stations showed consistently similar bio-optical properties, characteristic of Case 1 waters, enabling the evaluation of adjacency effects independent of the usual co-varying inputs of coastal waters. Using the third reprocessing of MERIS with the standard MEGS 8.1 processor, four different combinations of procedures were tested to improve the calibration between MERIS products and in situ data. These combinations included no vicarious adjustment (NoVIC), vicarious adjustment (VIC), and, for mitigating the effects of land adjacency on MERIS ρ w , the improved contrast between ocean and land (ICOL) processor (version 2.7.4) and VIC + ICOL. Out of approximately 130 potential matchups for each station, 38-77%, 74-86%, and 88-90% were achieved at Stations A, B, and C, respectively, depending on which of the four combinations were used. Analyses of ρ w comparing these various procedures, including statistics, scatter plots, histograms, and MERIS full-resolution images, showed that the VIC procedure compared with NoVIC produced minimal changes to the calibration. For example, at the oceanic Station C, the regression slope was closer to unity at all wavelengths with NoVIC compared to VIC, whereas, with the exception of wavelengths 412 and 443 nm, the intercept, mean ratio (MR), absolute percentage difference (APD), and relative percentage difference (RPD) were better with NoVIC. The differences for MR and APD indicate that there was marginal improvement for these two bands with VIC, and an over-adjustment with RPD. ICOL also showed inconsistent results for improving the retrieval of the near-shore conditions, but under some conditions, such as ρ w at wavelength 560 nm, the improvement was striking. VIC + ICOL showed results intermediate between those of VIC and ICOL implemented separately. In relation to other validation sites, the offshore Station C at Sagres had much in common with the Mediterranean deep water, BOUSSOLE buoy, although the matchup statistics between MERIS ρ w and in situ ρ w were much better for Sagres than for BO