CHARACTERIZING AQUIFER HETEROGENEITY IN THE ABERJONA RIVER VALLEY, MASSACHUSETTS

The aquifer underlying the Aberjona River near Woburn has been studied in numerous ways. This work seeks to complement previous studies by geostatistically characterizing heterogeneity in the aquifer. The results from this work are compared to previous geostatistical studies from other areas and inferences are made for future remediation of such aquifer systems. Lithologic descriptions from borehole records were divided into two facies categories: mud and diamicton (facies m) and sand and gravel (facies s). The proportions of these two facies for the entire study area were determined to be 33% for facies m and 67% for facies s. In order to better model the aquifer, the study area was divided aerially and the proportions computed for each new division until the differences in proportions among divisions were maximized. After several iterations, it was found that the proportion facies m decreased significantly down the axis of the valley, but showed no trend across it. The mean thickness and standard deviation (s) of each facies were computed directly since data in the vertical direction are continuous. The mean thickness of facies m was 4.0 m (s = 4.1 m) while the mean thickness of facies s was 7.1 m (s = 6.6 m). The model for vertical spatial correlation was based on transition probability. In this model, the vertical length of facies m was 6.8 m and the length of facies s was 13.6 m. The model for lateral correlation was based on variography. In this model, the lateral length of facies m was 250 m and the lateral length of facies s was 500 m. Based on these results, the heterogeneity in the aquifer under the Aberjona River can be represented with the same geostatiscal model as was developed from studies of the Miami Valley and the White River aquifer systems. Representative transport simulations by Ritzi et al. (2000) demonstrated that there is a high probability that large contaminant plumes from past disposal will have encountered facies m regions and consequently have slow groups of mass with orders of magnitude greater residence time than the mass which has traveled faster along preferential flow pathways and is already under remediation. This suggests that remediation strategies should have a long-term perspective.