Interdependence of regolith density, moisture, and chemistry parameters, as well as their influence on natural gamma ray emissions, led us to investigate systematically whether statistical models could be inferred between gamma spectrometric data and regolith parameters. A series of soil sample parameters were analyzed vs. airborne gamma ray data through multiple linear regressions. Among the approximately 20 regolith parameters modeled (chemical, textural, and mineralogical), about 50% (equally distributed in each category) were found to be predictable, with acceptable error (Radj2 > 0.5 and p value < 5%). With an independent set of texture analyses, we validated two of the predicted parameters (sand and clay contents) with fairly low residuals, with standard deviations of 22 and 16%, respectively. Further statistical investigations revealed why a large number of soil parameters could successfully be modeled in this sedimentary environment. We showed that the main gamma emitters are hosted in weathering products and leached detrital materials. Two main mineral assemblages correlate with gamma variables: (i) fine-grained weathering clays (with correlated Al, Fe, Mn, Mg, Pb, and V elements), and (ii) residual K-rich minerals, interpreted as feldspar and/or muscovite (correlated with Na and Sr). Additionally, two chemical elements, Si and Ca, have specific behaviors and can scarcely be characterized by the gamma data: they apparently mitigate Ɣ signatures.