Multi-Compartment Pollution Assessment of Soil Metals and Air Pollutants in Lebanon’s Bekaa Valley: Integrating Geochemistry, Remote Sensing, and Multi-Scale Spatial Analysis

This study presents the first integrated multi-compartment assessment linking soil heavy- metal contamination and satellite-derived tropospheric air pollution in Lebanon’s Bekaa Valley. Six metals (Cd, Cr, Cu, Ni, Pb, Zn) were measured in 295 agricultural topsoils using DTPA extraction—which targets the labile, bioavailable pool most relevant to food-chain transfer in calcareous soils—and were compared with Sentinel-5P TROPOMI vertical columns of SO2, NO2 and HCHO extracted at point, 1000 m and 5000 m scales. Chromium was the dominant element, and composite pollution indices indicated overall good but transitional soil quality, with a majority of sites at warning level. Principal component analysis, spatial autocorrelation and land-use analysis converged on three plausible source groupings—predominantly geogenic (Cr, Ni), localized anthropogenic (Cd, Pb) and agricultural (Cu)—although, in the absence of isotopic or other direct source tracing, these assignments remain indicative rather than confirmed. Soil metals and atmospheric columns were largely uncorrelated (|r| < 0.3 for 17 of 18 pairs); this decoupling is consistent with the contrasting temporal integration and spatial support of the two compartments rather than with a single shared pathway. Multi-scale analysis supported a parsimonious dual-scale (point + 5000 m) sampling framework. The study offers a transferable assessment framework for other Mediterranean regions and underscores the need for locally calibrated, total-metal soil background values. Because these indices and threshold comparisons are computed from DTPA-extractable concentrations, they serve here as one-directional, bioavailability-based screening tools rather than regulatory determinations: an exceedance conservatively flags a site for confirmatory total-metal analysis, whereas a non-exceedance does not establish compliance.