Browsing by Author "Marshall, Lucy"

Now showing 1 - 2 of 2

A Clustering Approach to Hydrological Predictions in Ungauged Basins
(2013-03) Weber, Katelyn; Marshall, Lucy; Greenwood, Mark
In an effort to improve hydrologic analysis in areas with limited data, the International Association of Hydrological Sciences (IAHS) formulated the Predictions in Ungauged Basins (PUB) initiative. Hydrologists seek to link catchments in such a way that basins where little to no data collection occurs can be related to catchments that are gauged. Various metrics and methods have been proposed to identify such relationships, in the hope that “surrogate” catchments might provide information for those catchments that are hydrologically similar. To examine the relationship between a hydrological model and certain hydrological metrics, we first run the Dynamically Dimensioned Search (DDS) Algorithm [Tolson and Shoemaker, 2007] to calibrate six model parameters for the Probability Distributed Model (PDM) [Moore, et. al, 2007]. We then use hierarchical clustering based on Ward's Algorithm to link catchments based on these six calibrated parameters. Clustering has been used in multiple recent hydrologic studies [Hastie, et. al, 2009 and Sawicz, et. al, 2011], but catchments are often clustered based on physical characteristics alone. Usually there is little evidence to suggest that such “surrogate” data approaches provide sufficiently similar model predictions. Beginning with model parameters and working backwards, we hope to establish if there is a relationship between the model inputs and physical characteristics for improved model predictions in the ungauged catchment. To analyze relationships, we use a perMANOVA test [Anderson, 2001] to determine if our clusters of physical metrics show significant delineation, which provides evidence to suggest that the surrogate procedure does, in fact, result in similar hydrological model behavior. Further, we perform perMANOVAs to determine which hydrological and physical descriptors show significant differences among clusters, and follow this up with a sequence of pairwise perMANOVAs between clusters. This leaves us with a complicated structure of clusters that are different based on certain metrics in the study.
A quantitative approach for integrating multiple lines of evidence for the evaluation of environmental health risks
(2015-01) Schleier, Jerome Joseph III; Marshall, Lucy; Davis, Ryan; Peterson, Robert K. D.
Decision analysis often considers multiple lines of evidence during the decision making process. Researchers and government agencies have advocated for quantitative weight-of-evidence approaches in which multiple lines of evidence can be considered when estimating risk. Therefore, we utilized Bayesian Markov Chain Monte Carlo to integrate several human-health risk assessment, biomonitoring, and epidemiology studies that have been conducted for two common insecticides (malathion and permethrin) used for adult mosquito management to generate an overall estimate of risk quotient (RQ). The utility of the Bayesian inference for risk management is that the estimated risk represents a probability distribution from which the probability of exceeding a threshold can be estimated. The mean RQs after all studies were incorporated were 0.4386, with a variance of 0.0163 for malathion and 0.3281 with a variance of 0.0083 for permethrin. After taking into account all of the evidence available on the risks of ULV insecticides, the probability that malathion or permethrin would exceed a level of concern was less than 0.0001. Bayesian estimates can substantially improve decisions by allowing decision makers to estimate the probability that a risk will exceed a level of concern by considering seemingly disparate lines of evidence.