Kelp forest mapping by use of airborne hyperspectral imager

Abstract

We present an easy and efficient approach for remote sensing of ocean color, relevant for monitoring and management of kelp forest and bottom substrate with a cheap custom made hyperspectral imager. Remote sensing of ocean color was performed in the Kongsfjord, Spitsbergen (79° N and 12° E) from an airplane (2950 m altitude) equipped with a hyperspectral imager, giving monochromatic images (425-825 nm) using the push broom technique, captured with custom designed software in 5 nm steps. Synchronously in situ measurements of upwelling spectral irradiance, (E u (λ)) (λ = 350-950 nm) measured at 30 cm depth were performed as a reference for the remotely sensed images. Surface water samples were taken for enumeration and identification of organic (plankton), inorganic particles, and colored dissolved organic matter. For identification and classification of kelp and bottom substrate, Bayesian supervised classification and a differential histogram equalization technique were used and compared. Both techniques gave successful discrimination between kelp and bottom substrate in shallow water above the Secchi depth (<19 m). The imager could easily be implemented for other applications such as detection and monitoring of phytoplankton blooms, suspended matter, and colored dissolved organic matter in surface waters, especially in connection with environmental and aquaculture management.

Modelling the cultivation and bioremediation potential of the kelp Saccharina latissima in close proximity to an exposed salmon farm in Norway

“…The HOBO loggers were calibrated monthly by measuring the downwelling irradiance at 4 depths (surface, 1, 2 and 3 m) using a Ramses ACC spectroradiometer (Trios), measuring from 350 to 950 nm (Volent et al 2007), simultaneously with one of the HOBO loggers at the same depths. Diffuse attenuation coefficients were calculated from the Ramses data and used to calculate photosynthetically active radiation (PAR) values from the HOBO loggers at depths of 0, 0.5, 1, 1.5, 2, 2.5 and 3 m (see Fig.…”

Mapping Spatial Distribution and Biomass of Intertidal Ulva Blooms Using Machine Learning and Earth Observation

“…Different techniques such as image thresholding (Cavanaugh et al, 2010(Cavanaugh et al, , 2011Cui et al, 2012;Bell et al, 2015), visual interpretation (Donnellan and Foster, 1999;Gower et al, 2006;Pfister et al, 2017), supervised classification (Volent et al, 2007;Casal et al, 2011;Bermejo et al, 2020), and unsupervised classification (Fyfe et al, 1999;Duffy et al, 2018) have been used in vegetation mapping in coastal and estuarine areas. Among the thresholding techniques, band ratios, vegetation indices, density, and biomass are commonly used (Richards, 2013), whereas for supervised classification, spectral angle mapper and maximum likelihood classifications are more conventional approaches (Schroeder et al, 2019).…”

Dual-Mode Hyperspectral Bio-Imager With a Conjugated Camera for Quick Object-Selection and Focusing

“…Hyperspectral imaging, which combines spectroscopy and digital imaging, has been widely used in remote sensing, medicine, agriculture, food detection and other fields [1][2][3][4][5][6]. The most common working modes of hyperspectral imaging are point scan, push-broom and staring imaging.…”