Open Access Repository

Development of hyperspectral and RGB imaging systems for under-ice mapping of fine scale sea-ice biophysical properties

Downloads

Downloads per month over past year

Cimoli, E ORCID: 0000-0001-7964-2716 2020 , 'Development of hyperspectral and RGB imaging systems for under-ice mapping of fine scale sea-ice biophysical properties', PhD thesis, University of Tasmania.

[img] PDF (Whole thesis)
Cimoli_whole_th...pdf | Document not available for request/download
Full text restricted until 15 May 2021.

Abstract

Polar sea ice is one of the largest biomes on Earth covering up to 6.1% of the area of the global ocean. Within this biome, sea-ice algae constitute a large, yet poorly quantified fraction of biomass contributing to polar marine productivity and large-scale biogeochemical cycles. Ice algae support the foundation of polar marine food webs by sustaining pelagic fauna, seeding planktonic blooms, and exporting organic material to the benthos. Advancing the capability to capture spatio-temporal dynamics of ice algae and its drivers is highly desirable.
Modern understanding of ecology advocates that the analyses of large-scale ecological patterns require the acknowledgment and integration of small-scale processes and that complex interactions occur at multiple scales. In sea ice, traditional sampling methods have struggled to capture and quantify patterns in algal community distributions due to a lack of methods that do not destroy the community and address microscale patterns (<0.1 m). Compounded with the challenges of surveying in polar regions, datasets remain fragmentary and coarse, hampering a mechanistic understanding and an ability to extrapolate and predict responses to environmental change. This doctoral research thesis aims to fill this methodological gap. It deals with the development, application, and assessment of passive hyperspectral imaging (HI) and photogrammetric approaches for quantitative microscale mapping of key sea-ice biophysical properties. The research developed a novel in situ under-ice platform and a field-deployable ice-core scanner. Their integration allowed for the retrieval of bio-optical regression algorithms to map chlorophyll-a both on ice cores and in linear transects beneath Antarctic sea ice
The thesis provides an under-ice close-range remote sensing perspective to the multidisciplinary problem of mapping biophysical properties in sea ice (Chapter 2). It starts by reviewing the current understanding of ice algal biomass variability, its environmental drivers, and highlights the missing links, thereby establishing the need for research development in this field. Radiative transfer in sea ice and the possibility to establish bio-optical relationships are the theoretical foundation of the proposed approaches and are therefore elucidated. A compilation of studies employing bio-optical models to retrieve biomass in sea ice is presented while discussing caveats and potentials for improvement. Technical and logistical trade-offs to be considered in under-ice radiation transfer mapping are discussed and illustrated, together with advances in emerging marine technologies that are changing the spatial scales of the surveys.
Chapter 3 presents an initial assessment of pushbroom HI technology to capture the variability of ice algae at the microscale using an innovative inverted sea-ice simulation tank. Through artificial illumination and controlled concentrations of algal consortia, HI was tested for a range of key HI parameters (e.g., different spectral resolutions). Exploratory image analysis revealed proxies of biomass matching inoculated abundances at unprecedented scales (a 0.8 x 0.8 m area at sub-mm resolution). Considerations for sensor selection relevant to ice algal mapping were narrowed down through this assessment (e.g., ≤ 3.4 nm spectral resolution). The study laid out the fundamental steps for the deployment of in situ HI, whilst highlighting the suitability of artificial sea-ice tanks to mimic the under-ice light environment and for testing of key parameters to improve the methodology.
Chapter 4 details the development of a novel underwater sled system for capturing referenced and overlapping HI and RGB imagery of the under-ice habitat. The chapter focuses on the technical, logistical, and theoretical considerations that are behind the system development and design. Tested under fast ice off Cape Evans, Antarctica, the system proved to be capable of capturing proxies of ice algal biomass and under-ice topography for the first time in situ at sub-mm spatial resolution. A transect 20.1 m long with a 0.61 m swath was presented, and data quality over a 0.7 by 0.61 m subsample was assessed. Image acquisition parameters for meaningful data acquisition in a cold (-1.8 °C) and low-light (E\(_{d,}\) \(_{400−700 nm}\) = 0.35 ± 0.20 λ, W m\(^{-2}\)) environment were assessed. Overall this study established the foundation of an adaptable solution that unlocks many research opportunities for marine under-ice mapping. Potential scientific applications for the system, its limitations and future developments are discussed.
Chapter 5 describes a complementary, field-deployable, hyperspectral scanning set-up that enables spatially-explicit quantification of both the vertical and horizontal microspatial variability of chlorophyll-α proxies in sea-ice cores. It further enables the retrieval of bio-optical regression algorithms relating sampled chlorophyll-α to spectra. New spectral indices tailored to our test area were developed with this scanning system explaining up to 85% of variation in chlorophyll-α. The performance of novel indices is statistically validated and compared to traditional ones (e.g., NDIs). Chapter 5 presents a first attempt to apply the retrieved regression models applied to both the in situ and horizontal ice-core sections of hyperspectral images, yielding per-pixel chl-α in mg m\(^2\). The unique under-ice habitat patterns captured are discussed in a biophysical context.
Underwater HI is a very novel technology and can be expected to revolutionise close-range underwater remote sensing of marine biogeochemical systems. This thesis pioneered HI coupled with photogrammetric approaches for the first time in extremely challenging polar marine waters beneath the ice. The trials of this technology and associated methodology have shed new light onto undocumented features of the under-ice habitat, which may permit the development of new research questions to understand this important biome.

Item Type: Thesis - PhD
Authors/Creators:Cimoli, E
Keywords: sea ice, algae, hyperspectral imaging, under-ice, mapping, biomass, bio-optical, Antarctica
Copyright Information:

Copyright 2020 the author

Additional Information:

Chapter 2 appears to be the equivalent of a post-print version of an article published as: Cimoli, E., Meiners, K. M., Lund-Hansen, L. C., Lucieer, V., 2017. Spatial variability in sea-ice algal biomass: an under-ice remote sensing perspective, Advances in polar science, 28(4), 268–296

Chapter 3 appears to be the equivalent of a post-print version of an article published as: Cimoli, E., Lucieer, A., Meiners, K. M., Lund-Hansen, L. C., Kennedy, F., Martin, A., McMinn, A., Lucieer, V., 2017. Towards improved estimates of sea-ice algal biomass: experimental assessment of hyperspectral imaging cameras for under-ice studies. Annals of glaciology, 58(7), 68-77. © The author(s) 2017. It is is an Open Access article, distributed under the terms of the Creative Commons Attribution 4.0 International (CC BY 4.0) licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited

Chapter 4 appears to be the equivalent of a post-print version of an article published as: Cimoli, E., Meiners, K. M., Lucieer, A., Lucieer, V. 2019. An under-ice hyperspectral and RGB imaging system to capture fine-scale biophysical properties of sea ice, Remote sensing, 11(23), 2860. © 2019 by the authors. Licensee MDPI, Basel, Switzerland. The article is an open access article distributed under the terms and conditions of the Creative Commons Attribution 4.0 International (CC BY 4.0) license (http://creativecommons.org/licenses/by/4.0/)

Related URLs:
Item Statistics: View statistics for this item

Actions (login required)

Item Control Page Item Control Page
TOP