Facultad de Ciencias Agrotecnológicas, Universidad Autónoma de Chihuahua, Avenida Pascual Orozco s/n, Campus I, Chihuahua, Chihuahua 31200, México.
Facultad de Zootecnia y Ecología, Universidad Autónoma de Chihuahua, Periférico Francisco R. Almada Km 1, Chihuahua, Chihuahua 31453, México.
Int J Environ Res Public Health. 2019 Jan 29;16(3):375. doi: 10.3390/ijerph16030375.
Mining is a major source for metals and metalloids pollution, which could pose a risk for human health. In San Guillermo, Chihuahua, Mexico mining wastes are found adjacent to a residential area. A soil-surface sampling was performed, collecting 88 samples for arsenic determination by atomic absorption. Arsenic concentration data set was interpolated using the ArcGis models: inverse distance weighting (IDW), ordinary kriging (OK), and radial basis function (RBF). For method validation purposes, a set of the data was selected and two tests were performed (P1 and P2). In P1 the models were processed without the validation data; in P2 the validation data were removed one by one, models were processed every time that a data point was removed. An arsenic concentration range of 22.7 to 2190 mg/kg was reported. The 39% of data set was classified as contaminated soil and 61% as industrial land use. In P1 the method of interpolation with the lowest RMSE was RBF (0.80), the highest coefficient of E was RBF (46.25), and the highest value was with RBF (0.48). In P2 the method with the lowest RMSE was OK (0.76), the highest E value was 50.65 with OK, and the reported the highest value with OK (0.52). The high arsenic contamination in soil of the site indicates an abundant dispersion of this metalloid. Furthermore, the difference between the models was not very wide. The incorporation of more parameters would be of interest to observe the behavior of interpolation methods.
采矿是金属和类金属污染的主要来源，这可能对人类健康构成威胁。在墨西哥奇瓦瓦州的圣吉列尔莫，采矿废物被发现毗邻一个居民区。进行了土壤表面采样，采集了 88 个样本，用于通过原子吸收法测定砷。使用 ArcGis 模型对砷浓度数据集进行了插值：反距离加权（IDW）、普通克里金（OK）和径向基函数（RBF）。为了验证方法，选择了一组数据并进行了两项测试（P1 和 P2）。在 P1 中，模型在没有验证数据的情况下进行处理；在 P2 中，每次删除一个数据点时，都会删除验证数据，并处理模型。报告的砷浓度范围为 22.7 至 2190 毫克/千克。数据集的 39%被归类为污染土壤，61%为工业土地用途。在 P1 中，RMSE 最低的插值方法是 RBF（0.80），E 最高的是 RBF（46.25），而 值最高的是 RBF（0.48）。在 P2 中，RMSE 最低的方法是 OK（0.76），E 值最高的是 OK（50.65），而 值最高的是 OK（0.52）。该地点土壤中的高砷污染表明该类金属大量分散。此外，模型之间的差异不是很大。纳入更多参数将有助于观察插值方法的行为。
