Receptor-Based Modeling of Groundwater ContaminationEPA Grant Number: U915324
Title: Receptor-Based Modeling of Groundwater Contamination
Investigators: Neupauer, Roseanna M.
Institution: New Mexico Institute of Mining and Technology
EPA Project Officer: Carleton, James N
Project Period: August 1, 1998 through August 1, 2001
Project Amount: $61,567
RFA: STAR Graduate Fellowships (1998) RFA Text | Recipients Lists
Research Category: Fellowship - Environmental Engineering , Engineering and Environmental Chemistry , Academic Fellowships
The objective of this research project is to develop a receptor-based modeling technique to improve characterization of known sources of groundwater contamination, and to identify previously unknown sources of groundwater contamination.
The governing equation for source-based mathematical models of contaminant transport in groundwater is the advection-dispersion equation (ADE), with the contaminant concentration as the state variable. The source-based ADE can be used to estimate the concentration of the contamination as a function of location and time after release from the source. Receptor-based models use a similar governing equation called the adjoint of the ADE. In the adjoint equation, location or travel time probability is the state variable; therefore, these probabilities are adjoint states of concentration. In the source-based model, the flow of information is downgradient, away from the source; in the receptor-based model, the flow of information is upgradient, away from the receptor and towards possible sources. To account for this reversal of flow, the sign on the advective term in the adjoint equation is reversed, and the receptor is a source of probability in the adjoint equation. The governing equation and boundary and initial conditions of the receptor-based model can be developed from adjoint theory. If contamination is detected at one receptor, the adjoint equation can be solved to obtain a distribution of location or travel time probability. If contamination is detected at multiple receptors, each receptor can be modeled as a source of probability in the adjoint equation; however, the equation does not account for the correlation between the detections. If the contamination parcels that are detected at the two receptors originated from the same source, the transport of the parcels is correlated. Therefore, the adjoint equation must be modified to account for the correlation of the two detections. With two correlated detections, the spread of the probability distributions should be smaller than the spread of the distribution with only one detection. The information from additional detections is expected to reduce the variance of the probability distribution, providing a better estimate of the source location.