Spectral variability of sea surface skylight reflectance and its effect on ocean color.

In this study, sea surface skylight spectral reflectance ρ(λ) was retrieved by means of the non-linear spectral optimization method and a bio-optical model. The spectral variability of ρ(λ) was found to be mainly influenced by the uniformity of the incident skylight, and a model is proposed to predict the ρ(λ) spectral dependency based on skylight reflectance at 750 nm. It is demonstrated that using the spectrally variable ρ(λ), rather than a constant, yields an improved agreement between the above-water remote sensing reflectance R(rs)(λ) estimates and concurrent profiling ones. The findings of this study highlight the necessity to re-process the relevant historical above-water data and update ocean color retrieval algorithms accordingly.

[Retrieve of red tide distributions from MODIS data based on the characteristics of water spectrum].

After comparing the spectral differences between red tide water and normal water, we developed a method to retrieve red tide distributions from MODIS data based on the characteristics of red tide water spectrum. The authors used the 119 series of in situ observations to validate the method and found that only one observation has not been detected correctly. The authors then applied this method to MODIS data on April 4, 2005. In the research areas three locations of red tide water were apparently detected with the total areas about 2 000 km2. The retrieved red tide distributions are in good agreement with the distributions of high chlorophyll a concentrations. The research suggests that the method is available to eliminating the influence of suspended sediments and can be used to retrieve the locations and areas of red tide water.

[Spectrum simulation based on data derived from red tide].

The present paper utilizes the absorption data of red tide water measured during the growing and dying course to retrieve imaginary part of the index of refraction based on Mie theory, carries out the simulation and analysis of average absorption efficiency factors, average backscattering efficiency factors and scattering phase function. The analysis of the simulation shows that Mie theory can be used to reproduce the absorption property of Chaetoceros socialis with an average error of 11%; the average backscattering efficiency factors depend on the value of absorption whose maximum value corresponds to the wavelength range from 400 to 700 nanometer; the average backscattering efficiency factors showed a maximum value on 17th with a low value during the outbreak of red tide and the minimum on 21th; the total scattering, weakly depending on the absorption, is proportional to the size parameters which represent the relative size of cell diameter with respect to the wavelength, while the angle scattering intensity is inversely proportional to wavelength.

[Spectral modeling of marine suspended particulate matter: comparison between ADA and Mie computations].

The present paper aimed at exploring the feasibility of modeling optical properties of marine suspended particulate matter (SPM) with anomalous diffraction approximation (ADA) method, which had advantage of computation speed over the classic Mie theory. Based on the marine SPM characteristics of refractive index and particle size, attenuation efficiency Q(c) absorption efficiency Q(a) and scattering efficiency Q(b) of SPM were computed by ADA and Mie to analyze their relative error delta(c), delta(a), delta(b) and the variations. The variations of delta(c), delta(a), delta(b) with the particle size, real and imaginary part of refractive index were different and complex. The minimum particle size R(min), with which ADA can be used to give computation results with enough accuracy (relative difference less than 5%), was determined and shown to be influenced by the refractive index and wavelength Classic Mie method, rather than ADA method, was recommended for the spectral modeling of ocean suspended particulate matter.

[Enteromorpha prolifera underwater spectral research based on simulation of radiation transmission].

The accumulation of Enteromorpha prolifera in huge amount in the Yellow sea in June, 2008 draws the attention from all over the world. It is an urgent requirement to monitor the wide range of Enteromorpha prolifera distribution by remote sensing. As to the Enteromorpha prolifera floating on the sea surface, effective monitoring by optical remote sensing has been basically achieved. As far as the underwater suspended Enteromorpha prolifera is concerned, the present paper carried out the radiative transfer simulation research on the above water spectral response, its variation with the suspending depth, the water turbidity and environmental conditions. It was found that with the increase in Enteromorpha prolifera suspending depth and water turbidity as well as the decrease in the thickness of Enteromorpha prolifera, the Enteromorpha prolifera information contained in the surface spectra would decrease. The influence of environmental factors such as water-gas interface roughness, cloud cover extent and sun zenith angle on the underwater suspended Enteromorpha prolifera spectra can be ignored. The maximum Enteromorpha prolifera depth that can cause surface spectrum changes is about 30 m in clean water and about 1 m in turbidity water.

[Study of red tide spectral characteristics and its mechanism].

In situ spectral data of different red tide, whose dominant species are leptocylindrus danicus, chattonella marina, skeletonema costatum, and mesodinium rubrum, were acquired by above water method utilizing spectrometer manufactured by FieldSpec Dual VNIR (USA). It is emphasized that the characteristic reflectance peak lying between 687 and 728 nm can be used to distinguish between red tide and normal sea water. Also the spectral discrepancy between different dominant species of red tide is pointed out, which could be utilized to identify certain red tide species by remote sensing technique. Mechanisms of phytoplankton red tide spectra peaks and vales are given. Spectral characteristics of mesodinium rubrum, a kind of protozoan, may be related to its symbiotic alga in its body and phytoplankton pigment crumb. So, research on ingestion preference, symbiotic property with algae, and fluorescence emission character of such symbiotic algae under normal temperature may be helpful for the deep understanding of mechanism of mesodinium rubrum spectra.

[Research on airborne hyperspectral identification of red tide organism dominant species based on SVM].

Airborne hyperspectral identification of red tide organism dominant species can provide technique for distinguishing red tide and its toxin, and provide support for scaling the disaster. Based on support vector machine(SVM), the present paper provides an identification model of red tide dominant species. Utilizing this model, the authors accomplished three identification experiments with the hyperspectral data obtained on 16th July, and 19th and 25th August, 2001. It is shown from the identification results that the model has a high precision and is not restricted by high dimension of the hyperspectral data.