Soil salinity hazard monitoring with portable X-ray fluorescence spectrometry and electrical conductivity meters

Soil salinization is a major form of land degradation that diminishes soil fertility and health, thereby severely impacting the economic development and livelihood improvement of agrarian countries worldwide. Salt-affected soils generally include both saline soils, characterized by excess soluble salts that hinder plant growth, and sodic soils, where high exchangeable sodium levels disrupt soil structure and permeability. Frequent and accurate monitoring of farmland salinity is a vital action for the timely management and control of salinization. Thus, this study was carried out in irrigated fields of the Tassaout region of Morocco, where irrigation has been practiced for centuries, with the objective of evaluating the effectiveness of using portable X-ray fluorescence (pXRF) for direct soil salinity measurement by comparing the results with EC measured values in soil-to-water (S:W) extracts at various ratios, namely 1:1, 1:2.5, and 1:5. In addition, pertinent soil physico-chemical properties, including pH and organic matter content, were determined for correlation analysis. The study proved that pXRF can be a reliable, cost-effective, and quick option for the electrical conductivity of saturated extract (EC e ) measurement both in the laboratory and in situ . Furthermore, the study developed predictive models for soil EC e estimation by complementing the pXRF technique, EC meter, and machine learning algorithms. The models were trained on 75% of the dataset using k-fold cross-validation and the remaining 25% for validation. The models performances were significantly better for EC 1:1 (R 2 = 0.94), EC 1:2.5 (R 2 = 0.93), and EC 1:5 (R 2 = 0.97). In conclusion, pXRF can be a reliable, cost-effective, and quick option for direct EC measurement both in the laboratory and in situ . In addition, the predictive models developed are promising tools for accurately inferring EC soil-to-water extract (EC S:W ) values either and EC e from pXRF or EC-meter readings. The study recommends using the pXRF technique and the predictive models developed for large-scale salinity monitoring.