Estimating Dissolved Organic Carbon Export from Terrestrial Landscape to Coastal Water using Remote Sensing and GIS Technologies

Qian Yu

UMass, Amherst



The variability of dissolved organic carbon (DOC) affects the biological and optical characteristics of coastal and estuarine waters. Colored dissolved organic matter is the fraction of the DOC pool that absorbs light in UV and visible ranges. Due to the complex mixing of fresh and marine water in river plume regions, DOC and CDOM concentrations change dramatically over relatively small spatial scales. My research interest is to examine the sources and transport mechanisms of DOC and CDOM from terrestrial ecosystems to coastal oceans by integrating in situ CDOM measurement, remote sensing estimating, and GIS watershed modeling. A quantitative model was created based on discrete samples collected monthly in the last two years to estimate DOC sources. Daily sub-basin flows simulated by GIS- based SWAT (soil and water assessment tool) model are used to estimate the DOC flux. The study demonstrated that terrestrial sources of DOC to estuaries can be potentially associated with land cover/landuse, biophysical seasonal variation and climate conditions. CDOM concentrations from forest endmembers are about 30% higher than from residential endmembers, while wetland and industrial endmembers were 2-3 times higher than forest endmembers in the Neponset River Watershed. In situ hyperspectral reflectance data from the spectroradiometer were analyzed to predict CDOM measurements using linear functional analysis. Spectral reflectance as a function of wavelength can explain 86% of the variation of in situ and discrete CDOM measurements in the estuary of lower Hudson River. The study provides huge encouragement for us to make progress in estimating CDOM from Spaceborne hyperspectral images.