Modelling The Quality of Arabian Gulf Waters Based on Inherent Optical Properties (IOPs) Measured from Space

  • Arwa Abdulmawjood

Student thesis: Master's Thesis


Maintaining good-quality water resources is a growing challenge today due to increased human activities, eutrophication, and climate change. Therefore, a robust monitoring plan is needed to assess the emerging water quality issues and identify the possible sources of pollution. Monitoring water quality depends on the basic physical, chemical and biological properties of water. These properties could be assessed by the traditional monitoring methods which appear to be limited in terms of spatial coverage and temporal variability. Therefore, remote sensing could be an alternative invaluable source of water quality data. Modelling the water quality parameters using remote sensing can be based on the apparent optical properties (AOPs) and/or the inherent optical properties (IOPs). However, the AOPs are not enough to indicate the complex properties of such waters, therefore, considering the IOPs is important. The performance of IOPs models was found to be better in deep less turbid waters than in shallow coastal waters. The aim of this work was therefore to develop a satellite-based model to estimate and understand the bio-optical properties of such waters (backscattering coefficient (bb), chlorophyll a (Chla), and turbidity) since IOPs in the Arabian Gulf have never been studied. In-situ measurements have been collected in the Arabian gulf and Gulf of Oman over the period 2013-2016. This data was used to validate the performance of the developed model and make it robust. Radiometric, physical, chemical and biological properties of water were included in the in-situ data. By examining the spatial temporal variability of these properties, maps and vertical profiles were created to characterize the coastal waters of the Arabian Gulf and the Gulf of Oman. Furthermore, multiple probability distributions were applied to model Chla under various bb ranges. Based on the Kolmogorov-Smirnov (KS), Anderson-Darling (AD), and Chi-Square (χ2 ) tests, Lognormal and Weibull provided the best fit to the Chla under all bb ranges. Satellite data was then downloaded from NASA's Ocean Color website. VIIRS (Suomi-NPP) and Aqua/ Terra MODIS data products had been collected and processed for the same days of the field campaigns. The results showed that GIOP was the better model in comparison to QAA and GSM. All models, however, overestimated the Chla and bb retrievals. The key elements responsible for the bias between in-situ and satellite data were discovered to be turbidity-related factors (diffuse attenuation coefficients (Kd) and Secchi disk depth (SDD)). In order to reduce the effect of turbidity on the quality of the satellite data, developed models for both Chla and bb were built. These models performed well for Chla and two wavelengths of bb, with R2=0.53 for Chla, R2=0.75 for bb(440), and R2=0.63 for bb(532). The models were used also to create seasonal long-term maps for the study area for the period 2012-2020. bb varied seasonally and spatially in the Arabian Gulf and the Gulf of Oman due to many factors such as upwelling, nitrogen flux, and sediment-rich waters. The study's final phase was to test the established model in a variety of case studies in the waters of the study area as well as other waters around the world. The model was used again because VIIRS-derived Chla in these locations overestimated the actual values. The model performed well in both Case I and Case II waters, with the correction being largest in locations with the highest Kd values. Since few studies on modelling the Gulf waters based on remote sensing have been carried out. This work will thus make a substantial contribution to this region in addition to global coastal waters. Collaboration and open data sharing between research groups to achieve a global ocean coverage will help significantly in fully understanding the bio-physical property variability and optical property relationships.
Date of AwardDec 2021
Original languageAmerican English


  • Remote sensing
  • AOPs
  • IOPs
  • Chla
  • bb
  • Probability distributions
  • KS
  • AD
  • χ2
  • KD
  • SDD
  • Case I
  • Case II waters

Cite this