Publications

Li, JS; Gao, M; Feng, L; Zhao, HL; Shen, Q; Zhang, FF; Wang, SL; Zhang, B (2019). Estimation of Chlorophyll-a Concentrations in a Highly Turbid Eutrophic Lake Using a Classification-Based MODIS Land-Band Algorithm. IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 12(10), 3769-3783.

Abstract
Due to the coarse spatial resolution and saturation issues associated with the 1-km ocean bands of MODerate-resolution Imaging Spectrometer (MODIS) instruments, the higher resolution (250 and 500 m) land bands are tended to be used for water color applications in coastal and inland waters. However, these wide spectral bands provide limited spectral information; therefore, resolving the chlorophyll-a concentration (Chla) signal in highly turbid waters poses a significant challenge. In this study, we present a classification-based algorithm to estimate Chla in a highly turbid eutrophic lake, Taihu Lake in Eastern China, using four visible to near-infrared land bands of MODIS observations. A threshold segmentation method ofMODIS R-rs(555)/R-rs(645) was used to categorize the lake into two classes: Chla-dominated waters (Class 1) and suspended particulate matter (SPM)-dominated waters (Class 2). Then, a band ratio of R-rs(859)/R-rs (645) was applied to retrieve Chla in Class 1, and a newly proposed spectral index, theAnti-SPM Chlorophyll-a Index (ASCI), was used to estimate Chla in Class 2. Validation using the leave-one-out cross-validation (LOOCV) method showed that the average unbiased relative error (AURE) of the derived Chla is 44.4%, and the coefficient of determination (R-2) is 0.55. The algorithm was further applied toMODIS data of Taihu Lake between 2000 and 2015 to obtain Chla time series maps, whose spatial and temporal patterns agreed well with previous studies.

DOI:
10.1109/JSTARS.2019.2936403

ISSN:
1939-1404