Improved Estimates of the Spatial Distribution and Temporal Trends of Water Quality Parameters Using Geostatistical Data Fusion Methods.

Contaminant;Sediment;Total Organic Carbon;Water Quality;Dissolved Oxygen;Weather Patterns;Hypoxia;Stratification;Geostatistics;Lake Erie;Chesapeak Bay;Civil and Environmental Engineering;Engineering;Environmental Engineering

Zhou, YuntaoMeadows, Guy A. ;

University of Michigan

关键词: Contaminant;    Sediment;    Total Organic Carbon;    Water Quality;    Dissolved Oxygen;    Weather Patterns;    Hypoxia;    Stratification;    Geostatistics;    Lake Erie;    Chesapeak Bay;    Civil and Environmental Engineering;    Engineering;    Environmental Engineering;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/98069/ytzhou_1.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Strategies aimed at reducing the degradation of water quality and predicting future changes in surface waters resulting from natural and anthropogenic forcing rely on the ability to track water quality changes, and to accurately quantify the distribution of water quality attributes. The three components of this dissertation focus on developing geostatistical data fusion methods that make optimal use of the available monitoring data in the Passaic River, Lake Erie, and the Chesapeake Bay, respectively. The first component presents a method for accurately estimating the spatial distribution of the total organic carbon in the sediments of the Passaic River using a dataset with non-uniform resolution. Estimating the spatial distribution of water sediment attributes at a uniform spatial resolution is often required for site characterizations and the design of appropriate risk-based remediation alternatives. Using a pseudodata example, a noval geostatisitical downscaling approach is shown to yield better estimates with a more accurate assessment of uncertainties, relative to traditional kriging methods. When applied to the estimation of the distribution of total organic carbon, geostatistical downscaling shows that the uncertainty associated with the spatial distribution of attribute is higher than would have been assumed if a kriging approach had been applied.The second and third components explore the degradation of water quality in time and space. Specifically, hypoxia (low dissolved oxygen) has been observed in Lake Erie and Chesapeake Bay since the early 1900s, leading to negative impacts such as ecosystem habitat degradation, altered migration patterns, and decreased fishery production. The interannual variability in hypoxic extent since mid-1980s in these two systems is quantified by combining spatially explicit auxiliary data with in situ dissolved oxygen measurements. The significance of nutrient loading, weather patterns, and stratification in explaining hypoxia in these systems is also explored. This research points to strong meteorological controls on hypoxia, through impacts on stratification and nutrient loading, in addition to the impact of anthropogenic activities.Overall, the developed geostatistical data fusion methods are shown to provide a means for producing reliable estimates of water quality attributes along with their associated uncertainties.