Auswertung von funktionellen Gruppen des Phytoplanktons aus hyperspektralen Satellitendaten und ihre Anwendung für die Untersuchung der Dynamik des Phytoplanktons in ausgewählten Meeresregionen.
|Other Titles:||Phytoplankton Functional Groups from Hyperspectral Satellite Data and its Application for Studying Phytoplankton Dynamics in Selected Oceanic Regions.||Authors:||Sadeghi, Alireza||Supervisor:||Bracher, Astrid||1. Expert:||Bracher, Astrid||2. Expert:||Burrows, John P.||Abstract:||
Phytoplankton play a unique role in the marine ecosystem as the basis of the marine food-web. They are the main drivers of the biogeochemical cycles in the ocean, as well as influencing the ocean-atmosphere exchanges of carbon dioxide and particular gases and particles. Based on these exchanges, phytoplankton influence the chemistry of atmosphere and the balance of global climate. Moreover, through interaction with light (absorption and scattering), phytoplankton have a significant impact on the underwater optics, being also responsible for the variations in ocean color. However, performing all these roles depends significantly on the type of phytoplankton, as indeed they comprise of a wide range of species and groups, with different capabilities and different distribution patterns in the World Ocean. Therefore, distinguishing between different types of phytoplankton is important to improve the knowledge of their actual roles in the ocean and climate system. As the spectral patterns of light absorption (essential for photosynthesis) vary among different groups of phytoplankton, the backscatter light from ocean preserves the spectral fingerprints of the inhabitant groups of phytoplankton. This feature can be used to determine remotely different types of phytoplankton. The purpose of this PhD-work was to improve a phytoplankton retrieval method, which was established to distinguish quantitatively major phytoplankton groups based on their absorption characteristics. The method, called PhytoDOAS, uses high spectrally resolved satellite data, provided by SCIAMACHY sensor. So far, by applying PhytoDOAS method to SCIAMACHY data, two main phytoplankton groups, diatoms and cyanobacteria, have been successfully distinguished. Through this work the method was improved to detect additionally coccolithophores, another important taxonomic group with significant biogeochemical functions. In this improvement, instead of the usual approach of the PhytoDOAS, which was based on single-target fitting, the simultaneous fitting of a certain set of phytoplankton groups was implemented within a wider wavelength window, thereby the new approach is called multi-target fit. Selection of the set of phytoplankton targets was according to the spectral analysis of absorption features of those groups that are most important with respect to the principal biogeochemical impacts, based on which marine microalgae are grouped as phytoplankton functional types, PFTs. The improved method was successfully tested through detecting independently reported blooms of coccolithophores, as well as by comparison of PhytoDOAS coccolithophores with global distributions of Particulate Inorganic Carbon (PIC), which is used as a proxy of coccolithophores. As the next step of this PhD-work, the results of the improved PhytoDOAS method were used to investigate temporal variations of coccolithophore blooms in selected regions. Eight years of SCIAMACHY data, from 2003 to 2010, were processed by the PhytoDOAS triple-target mode to monitor the biomass of coccolithophores in three oceanic regions, characterized by the frequent occurrence of large blooms. Then the PhytoDOAS results, as monthly mean time-series, were compared to appropriate satellite products, including the total phytoplankton biomass (total chl-a) from GlobColour data-set and the PIC distribution from MODIS-Aqua. To study the dynamics of coccolithophore blooms, the variations of coccolithophores, overall chl-a and PIC, as monthly mean time series, were investigated in the context of variations in the main oceanic geophysical parameters: sea-surface temperature (SST), mixed-layer depth (MLD) and surface wind speed. As a general result, it was observed that the inter-annual variations of the coccolithophore bloom cycles followed well the respective variations in the mentioned geophysical parameters, as they have been reported being associated with coccolithophore blooms. Observed anomalies were investigated based on the specific regional features of the geophysical conditions. Using the results of regional time series, the hypothesis that close coccolithophore blooms succeed the diatom blooms was roughly approved, suggesting, however, a weekly-based averaging of coccolithophores and diatoms for a more precise analysis. It has been frequently reported that high reflectance from surface waters in coccolithophore rich areas affects the performance of standard chl-a algorithms. The regional time series studies of this thesis indicated an underestimation of total chl-a by the standard algorithm during the time of occolithophore blooms. However, a comprehensive validation of the ocean color algorithms with in-situ phytoplankton data is needed to reach the final assessment of the short-comings.
|Keywords:||Phytoplankton, Coccolithophores, PhytoDOAS, SCIAMACHY, Chlorophyll-a, Ocean-color remote-sensing||Issue Date:||17-Apr-2012||Type:||Dissertation||URN:||urn:nbn:de:gbv:46-00102978-16||Institution:||Universität Bremen||Faculty:||FB1 Physik/Elektrotechnik|
|Appears in Collections:||Dissertationen|
checked on Jan 19, 2021
checked on Jan 19, 2021
Items in Media are protected by copyright, with all rights reserved, unless otherwise indicated.