Assessing the Causes and Severity of Gulf of Mexico Hypoxia Using Geostatistical and Mechanistic Modeling.

Water Quality Modeling;Geostatistical Modeling;Bayesian Modeling;Gulf of Mexico Hypoxia;Civil and Environmental Engineering;Natural Resources and Environment;Engineering;Science;Natural Resources and Environment and Environmental Engineering

Obenour, Daniel R.Ibanez, Ines ;

University of Michigan

关键词: Water Quality Modeling;    Geostatistical Modeling;    Bayesian Modeling;    Gulf of Mexico Hypoxia;    Civil and Environmental Engineering;    Natural Resources and Environment;    Engineering;    Science;    Natural Resources and Environment and Environmental Engineering;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/100085/obenour_1.pdf?sequence=1&isAllowed=y

瑞士|英语

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

【 摘 要 】

Hypoxia, typically defined by dissolved oxygen levels below 2 mg L-1, is an environmental problem common to many coastal systems.One particularly severe example of hypoxia is the large ;;dead zone’ that forms nearly every summer on the Louisiana-Texas shelf of the northern Gulf of Mexico.While there is considerable agreement about the primary causes of hypoxia, there remains substantial uncertainty regarding its spatial and temporal variability, such that it is difficult to predict how hypoxia will respond to management actions and other environmental changes.This research focuses on improving our understanding of Gulf hypoxia through three types of quantitative modeling.First, a geostatistical regression is developed to empirically model how water column stratification (a primary driver of hypoxia) affects bottom water dissolved oxygen (BWDO) concentrations, and to also infer the importance of other primary drivers, such as nutrient loading.Second, a geostatistical spatial estimation model is developed to simulate BWDO and hypoxic layer thickness across the Gulf shelf, providing estimates of hypoxic zone area and volume for a 27-year study period.Third, a mechanistic model, driven by nutrient loading, flow, and weather conditions is developed to predict hypoxic severity, as determined from the geostatistical model.As with all environmental models, the models developed in this dissertation are approximations of reality, tuned to limited observational and experimental information, such that they contain significant uncertainty.Because of this, all models are developed within statistical frameworks that quantify uncertainty and allow results to be presented as ranges of likely values.Overall, this works suggests there has been considerable variability in the mid-summer hypoxic extent over the last few decades, and this variability is explained, in large part, by both nutrient loading and oceanographic conditions (i.e., stratification).Relatively parsimonious models that account for these two main drivers explain at least 70% of the year-to-year variability in hypoxic area and mean BWDO.Also, this work indicates that over the past few decades, the Gulf has not become increasingly susceptible to hypoxia formation (independent of the biophysical drivers considered), at least in terms of hypoxic area and mean BWDO.