Detecting trend and seasonal changes in bathymetry derived from HICO imagery: A case study of Shark Bay, Western Australia

Abstract

The Hyperspectral Imager for the Coastal Ocean (HICO) aboard the International Space Station has offered for the first time a dedicated space-borne hyperspectral sensor specifically designed for remote sensing of the coastal environment. However, several processing steps are required to convert calibrated top-of-atmosphere radiances to the desired geophysical parameter(s). These steps add various amounts of uncertainty that can cumulatively render the geophysical parameter imprecise and potentially unusable if the objective is to analyze trends and/or seasonal variability. This research presented here has focused on: (1) atmospheric correction of HICO imagery; (2) retrieval of bathymetry using an improved implementation of a shallow water inversion algorithm; (3) propagation of uncertainty due to environmental noise through the bathymetry retrieval process; (4) issues relating to consistent geo-location of HICO imagery necessary for time series analysis, and; (5) tide height corrections of the retrieved bathymetric dataset. The underlying question of

A semianalytical ocean color inversion algorithm with explicit water column depth and substrate reflectance parameterization

“…Within the SWIM algorithm, scene-wide fixed initial guesses of 0.2, 0.01, and 0.001 m 21 were used for P, G, and X, respectively. These initialization values are similar to those used previously when applying the HOPE algorithm and its variants [Garcia et al, 2014a;Klonowski et al, 2007] and are similar to initial guesses used in the default parameterization of GIOP [Werdell et al, 2013a]. It has been established that initial guess used in the L-M optimization can affect the retrieved IOPs, particularly if the starting guess is next to a local solution space minima [Garcia et al, 2014b].…”

“…Further, the magnitude of the IOP parameters (P, G, and X) in the L-M routine were constrained using upper and lower bounds following Garcia et al [2014a], 20:05a w ð443Þ < P < 5:0 ðm 21 Þ;…”

“…[2009]; BRUCE- Klonowski et al [2007]; SAMBUCA- Wettle and Brando [2006]) and CRISTAL have shown particular merit when applied to hyperspectral imagery (contiguous spectral bands, resolution $5 nm) Dekker et al, 2011;Fearns et al, 2011;Garcia et al, 2014a;Goodman and Ustin, 2007;Hedley et al, 2009;Klonowski et al, 2007;Lee et al, 1999;Lesser and Mobley, 2007]. However, these previous studies were designed primarily to demonstrate bathymetric retrieval and benthic classification capabilities for shallow waters, typically less than 10 m depth, with little emphasis on the derived IOP values and downstream geophysical products such as CHL, SPM, and water clarity measures.…”

A Method to Analyze the Potential of Optical Remote Sensing for Benthic Habitat Mapping

“…The atmospheric, sunglint and air-to-water interface corrections used on the HICO image of Shark Bay to convert top-of-atmosphere calibrated radiances to rrs are described in Garcia et al [44].…”

Progress in Southern Ocean Biologyfrom the Indian Sector: Half-Decadal (2009-13) overview

“…Surface Chl-a along 57.5°E was more uniform and the zonal variations across 60°S showed more productivity than the eastern margins (Pavithran et al, 2014b), which could be due to the availability of micronutrients because of proximity to the African continent. Phytoplankton diagnostic pigment-based analysis using CHEMTAX showed the cross-front distribution of phytoplankton communities and the relative contribution from major taxonomic groups to total biomass at STF and PF (Garcia et al, 2014;Mendes et al, 2015). Cyanobacteria and prochlorophytes dominated the total Chl-a in the STF, the haptophyes dominated the SAF region and diatoms were dominant at the PF.…”