Master Thesis

REMOTE ESTIMATION AND SEASONAL VARIATION OF CHLOROPHYLL-A AND TOTAL SUSPENDED MATTER IN THE GULF OF GUINEA NIGERIA LOCATION

Andrew Ehiabhi Akhighu (03/2023-08/2023)

Support: Carl Beierkuhnlein

The Gulf of Guinea (GoG) is a coastal, shallow, turbid sedimentary basin characterized by complex optical properties. Interestingly, no previous research has compared the actual data from the GoG with the effectiveness and accuracy of satellite-based ocean colour information. This study attempts for the first time, to employ satellite images from Landsat 8 Operational Land Imager (OLI) for monitoring two key water quality variables - Chlorophyll-a (Chl-a) and Suspended Particulate Matters (SPM) concentrations in the GoG Nigeria location. Generic remote sensing algorithms—Normalized Difference Chlorophyll Index (NDCI) and Nechad et al., 2010 Suspended Particulate Matters (SPM) — were employed to monitor Chl-a and SPM concentrations from 10 coastal-ocean sampling locations during the dry and rainy seasons of 2014 and 2015 representatively. Achieving this objective requires minimal error reflectance values, generated from the use of an appropriate Atmospheric Correction (AC) model, which necessitated the evaluation of two widely used atmospheric correction models namely - Fastline of Sight Atmospheric Analysis of Hypercubes (FLAASH) and the Atmospheric Correction of OLI Lite (ACOLITE). The evaluation of the ACOLITE and FLAASH models generated an average root mean square error (RMSE) percentage of 6.66% and 14.49%, respectively. ACOLITE, designed specifically for inland, coastal, and ocean water bodies, demonstrates higher accuracy than FLAASH. Statistical analysis further indicates that Chl-a concentrations are highest during the dry season due to the depletion of nutrients by phytoplankton. The lowest estimated Chl-a concentration (NDCI value of approximately - 0.6 5 mg m−3, corresponding to a Chl-a range of 7.5 mg m−3; was found in the Lekki ocean sampling locations during the rainy season. On the other hand, the highest NDCI Chl-a concentrations ranging from 0.3 to 0.35 5 mg m−3, with a Chl-a range of 33-50 mg m−3 are observed in Ogun, Ondo, Delta, and Lekki ocean sampling locations, in decreasing order, during the dry season. Bar-Beach and Badagary ocean sampling locations consistently exhibit lower concentrations (ranging from - 0.05 to -0.2 5 mg m−3, with a Chl-a range of 7.5 mg m−3) across all seasons. Similarly, statistical analysis reveals that the highest SPM concentrations are found in the Ogun and Lekki ocean sampling locations (98.8 mg l-1 and 68.3 mg l-1, respectively). The lowest concentrations, approximately 15.03 mg/l, are estimated in the Badagary and Bar-beach ocean sampling locations. Although no consistent pattern is identified in the trend analysis of SPM, we observed higher concentrations echoing more during the dry season at the Arcuate Delta and Strand coast, while at the Lagoon and Mahin mud coast a higher concentration echoes more in the rainy seasons. An oil spill detection analysis was performed in a section of the study area using Sentinel-1 images from the European Space Agency to further explain the resultant spatial variability or anomalies observed of Chl-a and SPM in the area under investigation. This analysis confirms the presence of significant hydrocarbon (crude oil) on surface waterbodies in the Arcuate Delta and Strand coast (southeast section) of the GoG Nigeria location, which supports the anomalies identified in Chl-a and SPM concentration values. Despite the challenges posed by the low signal-to-noise ratio (SNR) and the widened wavelength of L8 OLI, which can affect the sensor’s capacity to accurately measure water leaving radiance or reflectance, coupled with the absence of specific regional algorithms for the Gulf of Guinea, this study represents the first attempt to utilize remote sensing tools in this water body. Consequently, this study draws attention to alternative data sources for retrieving regional essential ocean variables (EOV) in a dynamic environment such as the GoG. It also highlights the need for developing new sensor platforms for cloud-contaminated regions and specific regional remote sensing algorithms, particularly for Chl-a and SPM.