Premium
Ground‐Water Quality Variations in a Silty Alluvial Soil Aquifer, Oklahoma
Author(s) -
Hoyle Blythe L.
Publication year - 1989
Publication title -
groundwater
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.84
H-Index - 94
eISSN - 1745-6584
pISSN - 0017-467X
DOI - 10.1111/j.1745-6584.1989.tb01975.x
Subject(s) - hydrology (agriculture) , aquifer , carbon dioxide , alluvial plain , environmental science , groundwater , bicarbonate , alluvium , water quality , soil science , water well , geology , ecology , chemistry , geomorphology , biology , paleontology , geotechnical engineering , organic chemistry
Sixteen wells in a residential area in central Oklahoma were monitored from mid‐April 1986 to early February 1987 to characterize natural water quality variations in a shallow, fine‐grained alluvial soil aquifer. The wells are in four clusters, each with four wells, which range in depth from 8 to 14 feet. Two clusters are adjacent to a row of large deciduous trees and two are in grass‐covered areas. Considerable spatial and seasonal variations occurred in bicarbonate (the major ion in the ground water) and in electrical conductance. Both parameters were generally higher in wells adjacent to the trees than in the grassy areas. Carbon dioxide partial pressure values (Pco 2 ) calculated for ground‐water samples ranged from 10 −2·34 to 10 −0·42 atm. For a given pH, Pco 2 values were higher in the shallower wells under the trees than in grassy areas. These data suggest that more carbon dioxide was produced by tree respiration or associated microbial processes than by grass, resulting in higher bicarbonate concentrations and conductances. Statistical analyses of conductance and bicarbonate data indicated that sample variances were not always equal and that sample distributions were either normal, skewed, or polymodal. These results violate the assumptions of normal distribution and homogeneous variances underlying parametric statistical tests, such as Student's t test, which some regulatory agencies use to compare sample means. Results of this study suggest that more than one well is required to characterize background water quality, and that innovative statistical methods are needed to evaluate monitoring data.