A BOOTSTRAP MODEL FOR ESTIMATING THE CONCENTRATION OF SOME HEAVY METAL POLLUTANTS OF GROUND WATER FROM SOIL

Estimation of the concentration of some heavy metals in soil is very important in risk assessment study to determine contaminant risk to water from soil. An adaptive cluster sampling technique was employed and a model-assisted approach that incorporates models into design-based estimation was used. In this paper, a bootstrap procedure is proposed to approximate the distribution of the concentration of heavy metals in the soil. The proposed method overcomes the problems associated with methods requiring distributional assumptions the validity of which are often times extremely difficult to check. We conclude with a comparison of this procedure to the well known Horvitz-Thompson (HT) type estimator.