Improvement of sampling strategies for randomly distributed hotspots in soil applying a computerized simulation considering the concept of uncertainty

Abstract

Introduction

The pollution of soil and environment as a result of human activity is a major problem. Nowadays, the determination of local contaminations is of interest for environmental remediation. These hotspots can have various toxic effects on plants, animals, humans, and the whole ecological system. However, economical and juridical consequences are also possible, e.g., high costs for remediation measures.

Materials and methods

In this study three sampling strategies (simple random sampling, stratified sampling, and systematic sampling) were applied on randomly distributed hotspot contaminations to prove their efficiency in term of finding hotspots. The results were used for the validation of a computerized simulation.

Results and conclusion

This application can simulate the contamination on a field, the sampling pattern, and a virtual sampling. A constant hit rate showed that none of the sampling patterns could reach better results than others. Furthermore, the uncertainty associated with the results is described by confidence intervals. It is to be considered that the uncertainty during sampling is enormous and will decrease slightly, even the number of samples applied was increased to an unreasonable amount. It is hardly possible to identify the exact number of randomly distributed hotspot contaminations by statistical sampling. But a range of possible results could be calculated. Depending on various parameters such as shape and size of the area, number of hotspots, and sample quantity, optimal sampling strategies could be derived. Furthermore, an estimation of bias arising from sampling methodology is possible. The developed computerized simulation is an innovative tool for optimizing sampling strategies in terrestrial compartments for hotspot distributions.