MERIS-based ocean colour classification with the discrete Forel–Ule scale

Abstract

Abstract. Multispectral information from satellite borne ocean colour sensors is at present used to characterize natural waters via the retrieval of concentrations of the three dominant optical constituents; pigments of phytoplankton, non-algal particles and coloured dissolved organic matter. A limitation of this approach is that accurate retrieval of these constituents requires detailed local knowledge of the specific absorption and scattering properties. In addition, the retrieval algorithms generally use only a limited part of the collected spectral information. In this paper we present an additional new algorithm that has the merit of using the full spectral information in the visible domain to characterize natural waters in a simple and globally valid way. This Forel-Ule MERIS (FUME) algorithm converts the normalized multiband reflectance information into a discrete set of numbers using uniform colourimetric functions. The Forel-Ule (FU) scale is a sea colour comparator scale that has been developed to cover all possible natural sea colours, ranging from indigo blue (the open ocean) to brownish-green (coastal water) and even brown (humic-acid dominated) waters. Data using this scale have been collected since the late nineteenth century, and therefore, this algorithm creates the possibility to compare historic ocean colour data with present-day satellite ocean colour observations. The FUME algorithm was tested by transforming a number of MERIS satellite images into Forel-Ule colour index images and comparing in situ observed FU numbers with FU numbers modelled from in situ radiometer measurements. Similar patterns and FU numbers were observed when comparing MERIS ocean colour distribution maps with ground truth Forel-Ule observations.The FU numbers modelled from in situ radiometer measurements showed a good correlation with observed FU numbers (R 2 = 0.81 when full spectra are used and R 2 = 0.71 when MERIS bands are used).

True Colour Classification of Natural Waters with Medium-Spectral Resolution Satellites: SeaWiFS, MODIS, MERIS and OLCI

“…A linear relation works very well with only a small standard deviation of 1.6°. MERIS band setting is therefore very appropriate for retrieving the intrinsic colour of natural waters and it confirms the findings of an earlier publication on the FUME algorithm [12]. …”

“…In the development of the water hue angle concept, Wernand et al [12] used (α M ) for the hue angle derived with the FUME algorithm for MERIS, while Novoa et al [15] introduced (α W ) for the hue angle of water. In this manuscript we will refrain from indices and use (α) as the hue angle that represents the “true” or “intrinsic” colour of a natural water, which can be approximated by satellite remote sensing reflectance measurements.…”

“…It is true that the concentration of the optically active substances in the water can vary substantially, but the effect on the true colour is normally characterized by smooth transitions. In a first study, Wernand et al [12] demonstrated that the MERIS band setting allows for very accurate reconstruction of the water colour. After the development of the FUME algorithm [12], all MERIS FR data near the Spanish Mediterranean coast have been processed to hue angle and Forel-Ule scale for the Citclops project [8].…”

“…In a first study, Wernand et al [12] demonstrated that the MERIS band setting allows for very accurate reconstruction of the water colour. After the development of the FUME algorithm [12], all MERIS FR data near the Spanish Mediterranean coast have been processed to hue angle and Forel-Ule scale for the Citclops project [8]. From a first inspection of these maps, a number of applications have become apparent—easy descriptor of (rapidly changing) events like river plumes, eddies, convergence of water masses—quick check for (small) observation errors like atmospheric correction near clouds and close to land—connection to the water-observation experience of citizens.…”

WACODI: A generic algorithm to derive the intrinsic color of natural waters from digital images

“…For coastal and inland waters more optically active constituents determine the water leaving radiance and color (IOCCG 2000;van der Woerd and Pasterkamp 2008). Wernand et al (2013b) developed a specific MERIS algorithm (named FUME) to determine the color of natural water, based on color angles (a w ) converted to the Forel-Ule index FU (Ule 1892;Forel 1895;Wernand and van der Woerd 2010;Novoa et al 2013;Wernand et al 2013a,b). The application of this algorithm to satellite images allows classifying the seas and oceans based on their color.…”