Risk Evaluation of Chemical Clogging of Irrigation Emitters via Geostatistics and Multivariate Analysis in the Northern Region of Minas Gerais, Brazil

Abstract

In this study, we analyzed the hydrogeochemistry of 350 underground wells in the northern region of the state of Minas Gerais, Brazil, for water-chemical parameters that may contribute to the chemical clogging of emitters in drip irrigation systems. Risk class maps were generated for each parameter, and the area was classified based on the water characteristics, considering the degree of water-use restriction in micro-irrigation (i.e., no, moderate, and severe restriction). Inverse distance-weighted, random forest, and ordinary kriging methods were used as interpolation methods. Moreover, a multivariate analysis was conducted to analyze the results. Pearson’s correlation coefficient showed a strong and significant correlation between pH and carbonates, hardness, total dissolved solids (TDS), and electrical conductivity (EC) and between TDS and EC. Principal component analysis revealed that most of the variations in the water quality of the wells could be explained by water–rock interactions with the consequent dissolution of minerals. The principal components were natural sources of ionic salt groups, dissolution of minerals rich in alkaline cations, chemical weathering of iron–magnesium minerals, and increased pH with the conversion of bicarbonates into carbonates. In the parameter cluster analysis, three possible mechanisms that contribute to emitter clogging in the study area were identified: precipitation of calcium and magnesium salts; oxidation of iron and manganese ions forming oxides and insoluble hydroxides; an increase in pH, which converts bicarbonates into carbonates. Clustering analysis revealed the wells that are susceptible to clogging with the exact cause.