This paper aims to describe the variability of particulate absorption properties using a unique hyperspectral dataset collected in the Red Sea as part of the TARA Oceans expedition. The absorption contributions by phytoplankton (aph) and non-algal particles (aNAP) to the total particulate absorption coefficients are determined using a numerical decomposition method (NDM). The NDM is validated by comparing the NDM derived values of aph and aNAP with simulated values of aph and aNAP are found to be in excellent agreement for the selected wavelengths (i.e., 443, 490, 555, and 676nm) with high correlation coefficient (R2), low root mean square error (RMSE), mean relative error (MRE), and with a slope close to unity. Further analyses showed that the total particulate absorption coefficients (i.e., ap(443)average = 0.01995m−1) were dominated by phytoplankton absorption (i.e., aph(443)average = 0.01743m−1) with a smaller contribution by non-algal particles absorption (i.e., aNAP(443)average = 0.002524m−1). The chlorophyll a is computed using the absorption based Line Height Method (LHM). The derived chlorophyll-specific absorption ((a⁎ph = aph(λ)/ChlLH)) showed more variability in the blue part of spectrum as compared to the red part of spectrum representative of the package effect and changes in pigment composition. A new parametrization proposed also enabled the reconstruction of a⁎ph(λ) for the Red Sea. Comparison of derived spectral constants with the spectral constants of existing models showed that our study A(λ) values are consistent with the existing values, despite there is a divergence with the B(λ) values. This study provides valuable information derived from the particulate absorption properties and its spectral variability and this would help us to determine the relationship between the phytoplankton absorption coefficients and chlorophyll a and its host of variables for the Red Sea.
|Original language||English (US)|
|Number of pages||12|
|Journal||Remote Sensing Applications: Society and Environment|
|State||Published - Mar 16 2018|