Zoning Arsenic Presence Possibility in Calcareous Soils of Qorveh Plain Using Logistic Regression

Arsenic is a toxic metal whose presence in environment may cause serious damages to human and other forms of life. Existence of arsenic containing minerals among parent materials is a major factor contributing to arsenic pollution in soils. The objective of this research was the zoning of arsenic presence possibility and also recognition of the parameters affecting its concentration and solubility in calcareous soils of Qorveh plain in Kurdistan province، in Iran، using logistic regression model. Therefore، systematic soil sampling was conducted for the depth of 0-10cm surface soil in 107 sampling points across the eastern sub-basin of Qorveh plain following a grid distribution having intervals of 0. 9×0. 9km. Physico-chemical properties of the collected samples including acidity، electrical conductivity، organic matter، cation exchangeable capacity، soil texture، free iron oxides and anions of carbonates، phosphates، and sulfates were investigated as the major variables affecting concentration and solubility of arsenic in soil. The results showed that there were significant relationships between free iron oxides، phosphates، cation exchangeable capacity and geological characteristics with soil arsenic across the study area. Analysis of the results using the Predicted Percentage Correct Test (PPCT) and the index of Receiver Operating Characteristic (ROC) yielded values of 89. 7% and 95. 4%، respectively، which approved reliability of the logistic regression model to recognize influential parameters on zoning، concentration، and solubility of soil arsenic. Evaluation procedure to verify the map of arsenic presence possibility using Success Rate Curve (SRC) method yielded a value of 84. 18%، which emphasized on more arsenic concentration in extreme east and west parts of the study area.