Uncertainty analysis of remote sensing of colored dissolved organic matter: Evaluations and comparisons for three rivers in North America

• Published in: ISPRS journal of photogrammetry and remote sensing Vol. 84; pp. 12 - 22
• Main Authors: Zhu, Weining, Yu, Qian, Tian, Yong Q.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.10.2013; Elsevier
• Subjects: absorption; algorithms; Animal, plant and microbial ecology; Applied geophysics; Biological and medical sciences; CDOM; chlorophyll; Coastal; coasts; color; Dissolved organic matter; Earth sciences; Earth, ocean, space; EO-1 Hyperion; Estuaries; Exact sciences and technology; Fundamental and applied biological sciences. Psychology; General aspects. Techniques; Internal geophysics; Inversions; model uncertainty; QAA-CDOM; reflectance; Remote sensing; Remote sensing inversion; River systems; Rivers; sediments; Teledetection and vegetation maps; Uncertainty; Uncertainty analysis; Mississippi; American River; California; United States; ASW, USA, Mississippi; CDOM; River systems; EO-1 Hyperion; QAA-CDOM; Remote sensing inversion; Uncertainty analysis; algorithms; spectra; color; accuracy; Image; North America; coastal zone; uncertainties; high resolution; models; inverse problem; absorption; reflectance; streams; chlorophyll; remote sensing; concentration; biogeochemistry; in situ; surface waves; correlation; sediments
• ISSN: 0924-2716; 1872-8235
• DOI: 10.1016/j.isprsjprs.2013.07.005

The uncertainties involved in remote sensing inversion of CDOM (Colored Dissolved Organic Matter) were analyzed in estuarine and coastal regions of three North American rivers: Mississippi, Hudson, and Neponset. Water optical and biogeochemical properties, including CDOM absorption and above-surface spectra, were collected in very high resolution. CDOM’s concentrations (ag(440), absorption coefficient at 440nm) were inverted from EO-1 Hyperion images, using a quasi-analytical algorithm for CDOM (QAA-CDOM). Uncertainties are classified to five levels, in which the underwater measurement uncertainty (level 1), image preprocessing uncertainty (level 4) and inverse model uncertainty (level 5) were evaluated. Results indicate that at level 1, in situ CDOM measurement is significant with 0.1 in the unit of QSU and 0.01 in the unit of ag(440) (m−1). At level 4, surface wave is a potential uncertainty source for high-resolution images in estuarine and coastal regions. The remote sensing reflectance of wavy water is about 10 times of the truth. At level 5, the overall uncertainty of QAA-CDOM inversion is 0.006m−1, with accuracy R2=0.77, k=1.1 and RMSElog=0.33m−1. The correlations between uncertainties and other water properties indicate that the large uncertainty in some rivers, such as the Neponset and Atchafalaya, might be caused by high-concentration chlorophyll or sediments. The relationships among the three level uncertainties show that the level 1 uncertainty generally does not propagate into level 4 and 5, but the large uncertainty at level 4 usually introduce large uncertainty at level 5.