PhD Defense: Neranga Kaluappuwa Hannadige
Location
Physics : 401
Date & Time
November 9, 2023, 10:00 am – 1:00 pm
Description
ADVISOR: Dr. Pengwang Zhai
TITLE: Remote Sensing of Aerosols and Ocean Color with Multi-Angle Polarimeters and Spectro-Radiometers
ABSTRACT: The synoptic scale global coverage by satellite remote sensing of aerosols and ocean color can provide comprehensive details regarding air quality, ocean ecology, global biogeochemical cycles, and climate change. Atmospheric correction (AC), the process of extracting the water-leaving signal from the sensor measurement, is an essential step in ocean color remote sensing. The heritage AC processes applied to Moderate Resolution Imaging Spectroradiometer (MODIS)-like single-view spectrometers work well over open ocean waters, though its performance degrades over optically complex scenes in the presence of absorbing aerosols and/or coastal waters. The multi-angle polarimetric (MAP) measurements contain rich information on aerosol microphysical properties, making MAP a powerful instrument to characterize aerosols and perform AC over optically complex scenes, owing to its ability to perform measurements at multiple viewing angles and different polarimetric states. The MAP measurements in synergy with the spectrometers such as in NASA's Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) and ESA's Meteorological Operational-Second Generation (MetOp-SG) future satellite missions are examples of the synergistic use of MAPs to improve the AC of the multi- or hyper-spectral radiometers. The first part of this dissertation presents an AC scheme, "Polynomial-based Atmospheric Correction (POLYAC)", for hyperspectral radiometers that utilizes aerosol and surface information retrieved from collocated MAPs. POLYAC is expected to provide a robust and computationally efficient AC scheme for future satellite missions whenever the traditional algorithms fail.
Joint retrieval algorithms have been developed to perform retrievals of the MAP measurements. These algorithms iteratively fit the sensor measurements with a forward model simulation by perturbing the state vector containing parameters that represent the optical properties of the atmosphere and ocean system. The forward model includes sub-models to represent the optics of the atmosphere, ocean water surface, and ocean body. The retrieval performances of joint retrieval algorithms partially depend on the trade-off between how accurately the sub-models represent an observed scene and the number of parameters that consist in each sub-model. A large number of parameters has the ability to closely resemble an observed scene while it becomes computationally demanding and mathematically unstable for optimization algorithms. The information content of the remote sensing measurements further constrains the number of parameters that could be adopted in a particular model.
The impact of ocean color bio-optical models that characterize the water leaving signal in terms of spectral inherent optical properties (IOPs) of the water body is important. There is no universal bio-optical model and they differ with respect to the type of waters they represent. Typically a bio-optical model with more than a single parameter is required to represent the optical properties of optically complex waters such as coastal waters. Up to date, the number of optimal parameters that are used in a coastal water bio-optical model is arbitrary and ranges between 3 to 7. The second part of this dissertation presents the evaluation of the information content of the water-leaving signal under discrete wavelengths by using the principal component analysis. Furthermore, I have also studied the impact of the number of bio-optical model parameters and their model parametrizations on the semi-analytical algorithm (SAA) based inversion of the water-leaving signal. This work is expected to provide constraints on the size of the parameter space of ocean bio-optical models, that would be used in future ocean color retrieval algorithms.
TITLE: Remote Sensing of Aerosols and Ocean Color with Multi-Angle Polarimeters and Spectro-Radiometers
ABSTRACT: The synoptic scale global coverage by satellite remote sensing of aerosols and ocean color can provide comprehensive details regarding air quality, ocean ecology, global biogeochemical cycles, and climate change. Atmospheric correction (AC), the process of extracting the water-leaving signal from the sensor measurement, is an essential step in ocean color remote sensing. The heritage AC processes applied to Moderate Resolution Imaging Spectroradiometer (MODIS)-like single-view spectrometers work well over open ocean waters, though its performance degrades over optically complex scenes in the presence of absorbing aerosols and/or coastal waters. The multi-angle polarimetric (MAP) measurements contain rich information on aerosol microphysical properties, making MAP a powerful instrument to characterize aerosols and perform AC over optically complex scenes, owing to its ability to perform measurements at multiple viewing angles and different polarimetric states. The MAP measurements in synergy with the spectrometers such as in NASA's Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) and ESA's Meteorological Operational-Second Generation (MetOp-SG) future satellite missions are examples of the synergistic use of MAPs to improve the AC of the multi- or hyper-spectral radiometers. The first part of this dissertation presents an AC scheme, "Polynomial-based Atmospheric Correction (POLYAC)", for hyperspectral radiometers that utilizes aerosol and surface information retrieved from collocated MAPs. POLYAC is expected to provide a robust and computationally efficient AC scheme for future satellite missions whenever the traditional algorithms fail.
Joint retrieval algorithms have been developed to perform retrievals of the MAP measurements. These algorithms iteratively fit the sensor measurements with a forward model simulation by perturbing the state vector containing parameters that represent the optical properties of the atmosphere and ocean system. The forward model includes sub-models to represent the optics of the atmosphere, ocean water surface, and ocean body. The retrieval performances of joint retrieval algorithms partially depend on the trade-off between how accurately the sub-models represent an observed scene and the number of parameters that consist in each sub-model. A large number of parameters has the ability to closely resemble an observed scene while it becomes computationally demanding and mathematically unstable for optimization algorithms. The information content of the remote sensing measurements further constrains the number of parameters that could be adopted in a particular model.
The impact of ocean color bio-optical models that characterize the water leaving signal in terms of spectral inherent optical properties (IOPs) of the water body is important. There is no universal bio-optical model and they differ with respect to the type of waters they represent. Typically a bio-optical model with more than a single parameter is required to represent the optical properties of optically complex waters such as coastal waters. Up to date, the number of optimal parameters that are used in a coastal water bio-optical model is arbitrary and ranges between 3 to 7. The second part of this dissertation presents the evaluation of the information content of the water-leaving signal under discrete wavelengths by using the principal component analysis. Furthermore, I have also studied the impact of the number of bio-optical model parameters and their model parametrizations on the semi-analytical algorithm (SAA) based inversion of the water-leaving signal. This work is expected to provide constraints on the size of the parameter space of ocean bio-optical models, that would be used in future ocean color retrieval algorithms.
