Groundwater contamination

Algorithm

Groundwater flow

Ecology

Biology

Identification (biology)

Pollution

Geotechnical engineering

Groundwater pollution

Computer science

Botany

Environmental remediation

Geology

Environmental engineering

Groundwater

MODFLOW

Genetic algorithm

Hydrology (agriculture)

Environmental chemistry

Contamination

Environmental science

Chemistry

Water pollution

Aquifer

Machine learning

Journal of Hydrology

The accurate identification of groundwater pollution sources is the key to site remediation and management, which directly relates to the cost and effect of remediation during later periods. Studying the identification of some small-area site pollution sources by using a small amount of groundwater monitoring data is difficult and is currently a hot topic in the field of groundwater management. In this study, a method combining the Advection-Dispersion Equation (ADE) of contaminants in groundwater with Genetic Algorithm (GA) was proposed to identify groundwater pollution sources, that is, the location of the pollution source, the time of pollution release, and the intensity of pollution release. A sandbox experiment for the migration and transformation of deterministic petroleum hydrocarbons in groundwater is designed to verify the method. The results show that the method code program has basically the same identification parameters as the deterministic experimental pollution source parameters. The ADE-GA method was applied to an actual site polluted by a gas station in Beijing, China. The parameters of the pollution source obtained from the detailed site investigation are basically consistent with the calculated values. Therefore, the source code established in this work is a reliable tool for identifying the parameters of groundwater pollution sources in small areas.

Application of a genetic algorithm to groundwater pollution source identification