Exploring the limits for sky and sun glint correction of hyperspectral above-surface reflectance observations.

Above-surface radiance observations of water need to be corrected for reflections on the surface to derive reflectance. The three-component glint model (3C) [Opt. Express25, A742 (2017)OPEXFF1094-408710.1364/OE.25.0000A1] was developed to spectrally resolve contributions of sky and sun glint to the surface-reflected radiance signal $ {L_r}(\lambda ) $Lr(λ), and for observations recorded at high wind speed and with fixed-position measurement geometries that frequently lead to significant sun glint contributions. Performance and limitations of 3C are assessed for all relevant wind speeds, clear sky atmospheric conditions, illumination/viewing geometries, and sun glint contamination levels. For this purpose, a comprehensive set of $ {L_r}(\lambda ) $Lr(λ) spectra was simulated with a spectrally resolved sky radiance distribution model and Cox-Munk wave slope statistics. Reflectances were also derived from an extensive four-year data set of continuous above-surface hyperspectral observations from the Long Island Sound Coastal Observatory, allowing to corroborate 3C processing results from simulations and measurements with regard to sky and sun glint contributions. Simulation- and measurement-derived $ {L_r}(\lambda ) $Lr(λ) independently indicate that spectral dependencies of the sky light distribution and sun glint contributions may not be neglected for observations recorded at wind speeds exceeding $ 4\, m/s $4m/s, even for sun glint-minimizing measurement geometries (Sun-sensor azimuth angle $ \Delta \phi = 90 {-} {135° } $Δϕ=90-135°). These findings are in accordance with current measurement protocols for satellite calibration/validation activities. In addition, it is demonstrated that 3C is able to reliably derive water reflectance for wind speeds up to 8 m/s and $ \Delta \phi { \gt 20° } $Δϕ>20°.

Validation of a spectral correction procedure for sun and sky reflections in above-water reflectance measurements.

A three-component reflectance model (3C) is applied to above-water radiometric measurements to derive remote-sensing reflectance Rrs (λ). 3C provides a spectrally resolved offset Δ(λ) to correct for residual sun and sky radiance (Rayleigh- and aerosol-scattered) reflections on the water surface that were not represented by sky radiance measurements. 3C is validated with a data set of matching above- and below-water radiometric measurements collected in the Baltic Sea, and compared against a scalar offset correction Δ. Correction with Δ(λ) instead of Δ consistently reduced the (mean normalized root-mean-square) deviation between Rrs (λ) and reference reflectances to comparable levels for clear (Δ: 14.3 ± 2.5 %, Δ(λ): 8.2 ± 1.7 %), partly clouded (Δ: 15.4 ± 2.1 %, Δ(λ): 6.5 ± 1.4 %), and completely overcast (Δ: 10.8 ± 1.7 %, Δ(λ): 6.3 ± 1.8 %) sky conditions. The improvement was most pronounced under inhomogeneous sky conditions when measurements of sky radiance tend to be less representative of surface-reflected radiance. Accounting for both sun glint and sky reflections also relaxes constraints on measurement geometry, which was demonstrated based on a semi-continuous daytime data set recorded in a eutrophic freshwater lake in the Netherlands. Rrs (λ) that were derived throughout the day varied spectrally by less than 2 % relative standard deviation. Implications on measurement protocols are discussed. An open source software library for processing reflectance measurements was developed and is made publicly available.

Variability of adjacency effects in sky reflectance measurements.

Sky reflectance Rsky(λ) is used to correct in situ reflectance measurements in the remote detection of water color. We analyzed the directional and spectral variability in Rsky(λ) due to adjacency effects against an atmospheric radiance model. The analysis is based on one year of semi-continuous Rsky(λ) observations that were recorded in two azimuth directions. Adjacency effects contributed to Rsky(λ) dependence on season and viewing angle and predominantly in the near-infrared (NIR). For our test area, adjacency effects spectrally resembled a generic vegetation spectrum. The adjacency effect was weakly dependent on the magnitude of Rayleigh- and aerosol-scattered radiance. The reflectance differed between viewing directions 5.4±6.3% for adjacency effects and 21.0±19.8% for Rayleigh- and aerosol-scattered Rsky(λ) in the NIR. Under which conditions in situ water reflectance observations require dedicated correction for adjacency effects is discussed. We provide an open source implementation of our method to aid identification of such conditions.