Are Ground Water Models Reliable for Reconstructing Historical Contamination Levels?

A drinking water distribution system located within the U.S. Marine Corps Base Camp Lejeune  (CLJ) in North Carolina was contaminated by PCE and its biodegradation products.   In early 1982, a water quality survey completed at the site indicated the presence of several volatile organic compounds including PCE and TCE.   Further investigations revealed that a dry cleaning facility, located up-stream from the production wells, was the source of the contamination.   The dry cleaner used PCE since 1953 and disposed various forms of PCE-contaminated wastes in a septic tank and in shallow pits.  Therefore, the residents who lived in on-site family housing units had the potential to be exposed to these harmful environmental contaminants through the drinking water source.   In late 1980s, the concerns raised by CLJ public lead to an epidemiological study to evaluate the potential associations of utero and infant exposures to the VOCs and childhood cancers and birth defects.  The study included births occurring during the period of 1968-1985 to women who were pregnant while they resided at the base.  Since there was no monitoring data available for the period 1968-1982, the researchers used MODFLOW and MT3DMS codes to reconstruct the historical concentration levels.  In this presentation, I will first briefly review the details of these numerical models.  Later we will use this field scenario to investigate the following critical questions: 1) Are MODFLOW and MT3DMS models sufficiently reliable to reconstruct historical VOC contamination at field sites? 2) Are these model predications more reliable than common sense estimates derived from monitoring data?  3)  Can we use simple analytical tools to estimate these concentration levels?  and  4) Do we gain anything by adding further model complexities (e.g.,  multi-species reactive transport modeling)?   We will use this example to address the question—how much modeling is too much modeling?