HICO-Based NIR–Red Models for Estimating Chlorophyll- $a$ Concentration in Productive Coastal Waters

Abstract

We present here results that demonstrate the potential of near-infrared (NIR)-red models to estimate chlorophyll-a (chl-a) concentration in coastal waters using data from the spaceborne Hyperspectral Imager for the Coastal Ocean (HICO). Since the recent demise of the MEdium Resolution Imaging Spectrometer (MERIS), the use of sensors such as HICO has become critical for coastal ocean color research. Algorithms based on two-and three-band NIR-red models, which were previously used very successfully with MERIS data, were applied to HICO images. The two-and three-band NIR-red algorithms yielded accurate estimates of chl-a concentration, with mean absolute errors that were only 10.92% and 9.58%, respectively, of the total range of chl-a concentrations measured over a period of several months in 2012 and 2013 on the Taganrog Bay in Russia. Given the uncertainties in the radiometric calibration of HICO, the results illustrate the robustness of the NIR-red algorithms and validate the radiometric, spectral, and atmospheric corrections applied to HICO data as they relate to estimating chl-a concentration in productive coastal waters. Inherent limitations due to the characteristics of the sensor and its orbit prohibit HICO from providing anywhere near the level of frequent global coverage as provided by standard multispectral ocean color sensors. Nevertheless, the results demonstrate the utility of HICO as a tool for determining water quality in select coastal areas and the cross-sensor applicability of NIR-red models and provide an indication of what could be achieved with future spaceborne hyperspectral sensors in estimating coastal water quality.

Hyperspectral Shallow-Water Remote Sensing with an Enhanced Benthic Classifier

“…For satellite hyperspectral imagery this is more likely to occur over dark targets where there is a reduced signal-to-noise ratio with a higher proportion of atmospheric path radiance to the total at-sensor radiance. Here small errors in the sensor's radiometric calibration or inaccuracies in the atmospheric parameters used during atmospheric correction can significantly impact R sens rs (e.g., [59][60][61]). Bottom reflectance and benthic classification of hyperspectral imagery are products that can be generated using shallow water inversion models.…”

Sensor Capability and Atmospheric Correction in Ocean Colour Remote Sensing

“…Unfortunately, both of these sensors have been plagued with issues, such as unreliable spectral and radiometric calibration [100] and their low signal-to-noise ratio (SNR), especially over bodies of water, and this often leads to poor retrievals [101,102], e.g., the SNR of Hyperion ranges between 50:1 and 150:1 [103]. As HICO was specifically developed for ocean colour sensing, its 128 bands, which cover a spectral range from 380-960 nm, capture with a reasonable SNR (>200:1 for water-penetrating wavelengths and assuming 5% albedo) [18].…”

“…For HICO to achieve accurate retrievals over turbid waters, measurements of the aerosol optical depth [112] and aerosol type [113] needed to be either estimated or measured from other sources, such as AERONET. Other limitations include the fact that HICO suffered from spectral shifts (of up to 1 nm) and a low radiometric sensitivity (especially below 450 nm) [102]; it was not designed for regular global coverage [114], and its location on the International Space Station only permitted data capture within the latitude range ≈ +54°/−53°(which covered around 80% of the Earth's surface) [115]. Radiometric resolution is defined as the minimum amount of radiance that each spectral band of a sensor can reliably measure, and it is determined by the way the data are digitally stored [26].…”

“…Spectral inversion methods and constituent retrieval algorithms using hyperspectral data have shown benefits over using multispectral measurements [102,244]. Semi-analytical inversion methods attempt to derive inherent optical properties, such as absorption and scattering coefficients, from apparent optical properties, such as normalised water-leaving radiance and remote sensing reflectance [36,242,[245][246][247].…”

“…Although the majority of constituent retrieval algorithms for turbid waters have been developed for multispectral sensors and their estimates of chlorophyll-a are poor [25], a recent method has shown some success at retrieving the chlorophyll-a concentration in turbid waters [251]. The potential for hyperspectral sensors for monitoring chlorophyll-a and phytoplankton in turbid coastal and inland waters has been demonstrated in a number of studies [23,102,106]. Specific bands of HICO were selected by the three-band model in accordance with the optical properties of the water [23,102].…”