Research about Kent County, Michigan

USGS1998

Assessment of ground-water vulnerability to atrazine leaching in Kent County, Michigan; review, comparison of results of other studies and verification

Model assumptions and parameters used in an earlier study of the vulnerability of ground water in Kent County, Michigan, to atrazine contamination were reviewed and compared with other studies. The review indicated that model assumptions are consistent with those used in other models and that the parameters assigned in the Kent County model are within the broad range commonly found in the literature. Model assumptions most likely to limit the accuracy of the previous study include those of uniform transport and steady-state flow. Simulation results are more sensitive to parameter estimates for atrazine half life, organic-carbon content, and organic-carbon partition coefficient than to other model parameters. Potential atrazine detection probabilities and concentrations of atrazine reported in the previous Kent County Study were compared with data from other studies. Detection probabilities of 14 measurements of atrazine concentrations in ground water from shallow wells in the southern Lower Peninsula of Michigan were compared with detection probabilities based on potential atrazine concentrations computed for Kent County. Results indicate that the distribution of detection probabilities based on measured concentrations is similar to that based on adjusted potential concentrations. Potential concentrations were adjusted for effects of differences between sampling and modeling depths and for differences between the uniform application rate used for potential concentrations and the percentage of Kent County that is likely to be treated with atrazine. Potential concentrations of atrazine in the Kent County Study were within the wide range of concentrations measured in other states. A stratified random sampling strategy was developed to verify expected atrazine concentrations in ground water within Kent County. The strategy helps identify strata, determine the optimum allocation of ground-water samples within defined strata, and project the sampling error. Implementation of the strategy was illustrated by use of potential atrazine concentrations computed in the previous Kent County Study. Once detailed information on historical application rates of atrazine is developed, expected atrazine concentrations can be computed by use of the vulnerability model and used to implement the sampling strategy. Sampling results may be used to verify the effectiveness of the vulnerability model and local estimates of historical atrazine application rates by use of analysis of variance.

Michigan↗

USGS1997

Vulnerability of ground water to atrazine leaching in Kent County, Michigan

A steady-state model of pesticide leaching through the unsaturated zone was used with readily available hydrologic, lithologic, and pesticide characteristics to estimate the vulnerability of the near-surface aquifer to atrazine contamination from non-point sources in Kent County, Michigan. The modelcomputed fraction of atrazine remaining at the water table, RM , was used as the vulnerability criterion; time of travel to the water table also was computed. Model results indicate that the average fraction of atrazine remaining at the water table was 0.039 percent; the fraction ranged from 0 to 3.6 percent. Time of travel of atrazine from the soil surface to the water table averaged 17.7 years and ranged from 2.2 to 118 years. Three maps were generated to present three views of the same atrazine vulnerability characteristics using different metrics (nonlinear transformations of the computed fractions remaining). The metrics were chosen because of the highly (right) skewed distribution of computed fractions. The first metric, rm = RM λ (where λ was 0.0625), depicts a relatively uniform distribution of vulnerability across the county with localized areas of high and low vulnerability visible. The second metric, rm λ-0.5 , depicts about one-half the county at low vulnerability with discontinuous patterns of high vulnerability evident. In the third metric, rm λ-1.0 ( RM ), more than 95 percent of the county appears to have low vulnerability; small, distinct areas of high vulnerability are present. Aquifer vulnerability estimates in the RM metric were used with a steady-state, uniform atrazine application rate to compute a potential concentration of atrazine in leachate reaching the water table. The average estimated potential atrazine concentration in leachate at the water table was 0.16 μg/L (micrograms per liter) in the model area; estimated potential concentrations ranged from 0 to 26 μg/L. About 2 percent of the model area had estimated potential atrazine concentrations in leachate at the water table that exceeded the USEPA (U.S. Environmental Protection Agency) maximum contaminant level of 3 μg/L. Uncertainty analyses were used to assess effects of parameter uncertainty and spatial interpolation error on the variability of the estimated fractions of atrazine remaining at the water table. Results of Monte Carlo simulations indicate that parameter uncertainty is associated with a standard error of 0.0875 in the computed fractions (in the rm metric). Results of kriging analysis indicate that errors in spatial interpolation are associated with a standard error of 0.146 (in the rm metric). Thus, uncertainty in fractions remaining is primarily associated with spatial interpolation error, which can be reduced by increasing the density of points where the leaching model is applied. A sensitivity analysis indicated which of 13 hydrologic, lithologic, and pesticide characteristics were influential in determining fractions of atrazine remaining at the water table. Results indicate that fractions remaining are most sensitive to the unit changes in pesticide half life and in organic-carbon content in soils and unweathered rocks, and least sensitive to infiltration rates. The leaching model applied in this report provides an estimate of the vulnerability of the near-surface aquifer in Kent County to contamination by atrazine. The vulnerability estimate is related to water-quality criteria developed by the USEPA to help assess potential risks from atrazine to the near-surface aquifer. However, atrazine accounts for only 28 percent of the herbicide use in the county; additional potential for contamination exists from other pesticides and pesticide metabolites. Therefore, additional work is needed to develop a comprehensive understanding of the relative risks associated with specific pesticides. The modeling approach described in this report provides a technique for estimating relative vulnerabilities to specific pesticides and for helping to assess potential risks.

Michigan↗