Publications

Moradi, M; Kabiri, K (2019). Spatio-temporal variability of red-green chlorophyll-a index from MODIS data - Case study: Chabahar Bay, SE of Iran. CONTINENTAL SHELF RESEARCH, 184, 1-9.

Abstract
Chabahar Bay is a strategic and productive estuary in the south-east of Iran (north of Oman Sea). It is an optically complex bay, and its Chl-a content variations never studied. In this study, Red-Green Chlorophyll Index (RGCI) algorithm for estimating accurate Chl-a content was tested, validated and applied to Moderate Resolution Imaging Spectrometer (MODIS) data, using in situ bio-optical data collected seasonally from 2007 to 2015 in 35 stations. Chl-a concentrations varied from 7.7 to 31.3 mg m(-3) with a mean value of 16.4 +/- 5.9 mg m(-3). The mean absorption at 443 nm and 555 nm data showed that the detrital particles and CDOM contents is relatively low and the mean absorption is dominated by phytoplankton and non-living particles. The RGCI algorithm was tuned for green band center position of 555 nm, and showed improvement over the traditional blue-green band ratio algorithms (e.g. OC3) with mean relative error of 37.4% and RMSE of 73.2% for Chl-a ranging between 2 and 80 mg m(-3). The Wavelet Transform (WT) techniques were utilized to analyze the spatio-temporal stability and abnormality of MODIS Chl-a extracted using the tuned RGCI algorithm. Significant variability in time and space was observed, with higher Chl-a in the eastern segment and lower Chl-a in the middle of bay. The highest and lowest Chl-a concentrations were observed in summer and winter, respectively. WT components analysis and anomaly detection revealed the strong correlation of Chl-a concentrations and patterns with turbidity contents and adjacent river discharge. This study showed that the accuracy of RGCI algorithm depends on water body constituents and optimized green waveband position, and therefore the algorithm must be tuned regionally with in situ Chl-a data.

DOI:
10.1016/j.csr.2019.07.002

ISSN:
0278-4343