Comparison Study on the Estimation of the Spatial Distribution of Regional Soil Metal(loid)s Pollution Based on Kriging Interpolation and BP Neural Network

Abstract

Soil pollution by metal(loid)s resulting from rapid economic development is a major concern. Accurately estimating the spatial distribution of soil metal(loid) pollution has great significance in preventing and controlling soil pollution. In this study, 126 topsoil samples were collected in Kunshan City and the geo-accumulation index was selected as a pollution index. We used Kriging interpolation and BP neural network methods to estimate the spatial distribution of arsenic (As) and cadmium (Cd) pollution in the study area. Additionally, we introduced a cross-validation method to measure the errors of the estimation results by the two interpolation methods and discussed the accuracy of the information contained in the estimation results. The conclusions are as follows: data distribution characteristics, spatial variability, and mean square errors (MSE) of the different methods showed large differences. Estimation results from BP neural network models have a higher accuracy, the MSE of As and Cd are 0.0661 and 0.1743, respectively. However, the interpolation results show significant skewed distribution, and spatial autocorrelation is strong. Using Kriging interpolation, the MSE of As and Cd are 0.0804 and 0.2983, respectively. The estimation results have poorer accuracy. Combining the two methods can improve the accuracy of the Kriging interpolation and more comprehensively represent the spatial distribution characteristics of metal(loid)s in regional soil. The study may provide a scientific basis and technical support for the regulation of soil metal(loid) pollution.

Research on Distributed 5G Signal Coverage Detection Algorithm Based on PSO-BP-Kriging

“…Chagas et al successfully applied neural network technology to positioning problems based on RSSI value estimation in [29,30]. Jia et al found that Kriging interpolation can better reflect the spatial distribution characteristics of the target region, but the accuracy of neural network interpolation is higher [31]. In [32], an improved model using BP neural network technology instead of Kriging global model is proposed, which is further extended by linear weighted aggregation method to improve the modeling accuracy.…”

The rapid increase of urban contaminated sites along China’s urbanization during the last 30 years

“… 33 To estimate the spatial distribution of heavy metals in soil more accurately, a backpropagation (BP) neural network was used, which reduced the error by 42% compared to Kriging interpolation. 34 In groundwater, the random forest (RF) model predicted groundwater fluoride contamination across India with an accuracy of 0.78% at a resolution of 1 km. 35 In addition, the increase in the number of site surveys in recent years has provided a large number of training samples for the machine-learning algorithm to identify contaminated sites.…”

Multi-Metal Distribution Patterns in Soils of the Sacramento River Floodplain and Their Controlling Factors

“…Metals in floodplain soils are not distributed homogeneously, but rather show spatial patterns in their distribution [16,17]. These spatial patterns result from transport, deposition, and accumulation processes that depend on environmental factors [18,19]. Factors that affect metal concentrations in soils are geology, land cover, hydrology, terrain, and emission sources [20][21][22].…”