Study on the Spatial Distribution and Migration Characteristics of Petroleum Hydrocarbons in Soil Based on Artificial Neural Networks

Soil pollution due to petroleum hydrocarbons from industrialization and human activities has become an increasingly severe global issue. A comprehensive understanding of the spatial distribution and migration patterns of petroleum hydrocarbons in soil is essential for the precise prevention and control of soil pollution. This article utilizes soil samples taken from depths ranging from 0 to 8 m in a chemical plant, coupled with hydrogeological data, to model the spatial distribution of petroleum hydrocarbon content, and the prediction results are characterized by three-dimensional visualization. We delved into the migration characteristics of petroleum hydrocarbons in soil through the visualization perspective. The results show that the feed-forward neural network(FNN) has higher estimation accuracy than the traditional interpolation method; in the horizontal direction of the study area, the petroleum hydrocarbons content in the production area decreases as the distance from the pollution source increases, but the petroleum hydrocarbons content in the storage area increases and then decreases with the increase of the distance from the pollution source; Vertically, the petroleum hydrocarbon content shows a consistent rule of increasing and then decreasing with greater depth, and is mainly enriched in the lower part of the aeration zone and the upper part of the saturated zone within 4 m under the influence of groundwater. The corresponding soil layers were mainly miscellaneous fill and coarse sand. The prediction model proposed in this study provides a new means for the spatial distribution of pollutants in soil.