Comparison of the Interpolation of Lead and Zinc Accumulation in the Soils around Zanjan Lead and Zinc Corporation Using laboratory and Landsat Data

Background: This aimed to compare the interpolation of lead and zinc concentration at the National Lead and Zinc Corporation using Landsat satellite and laboratory data to introduce an optimal interpolation method. Methods: After collecting the laboratory data, geostatistical approaches were applied to model the spatial distribution of lead and zinc, including radial basis functions, inverse distance weighting, and ordinary kriging (Gaussian, spherical, exponential, and circular). Estimation accuracy was evaluated by cross-validation and MAE, MBE, and RMSE diagnostic statistics. Results: The Gaussian model had the lowest error and was the optimal method for modeling lead and zinc. After investigating the correlations between the Landsat ۵ satellite bands and soil element concentrations, the spatial distribution of lead and zinc was re-zoned in the ArcGIS software. In both methods, estimation accuracy was evaluated by cross-validation and MAE, MBE, and RMSE diagnostic statistics. Conclusion: The MAE and RMSE of the satellite data of lead were ۳۸.۳۶ and ۹۱.۷۳, while they were ۵۲.۹۳ and ۷۴.۵۷ for zinc, respectively. The experimental data of lead were ۵۳.۰۴ and ۱۲۵.۱۸while they were ۱۰۸.۱۵ and ۲۳۹.۲۵ for zinc, respectively. The accuracy of the satellite data in the interpolation of the investigated elements had lower error and higher accuracy.

کلیدواژه ها:

Soil contamination ، Interpolation ، National lead and zinc co. ، Landsat satellite