Positive Matrix Factorization (PMF) based on geochemical data for revising geological map in semi-desert covered terrains: A case study from a map sheer of Urat Rear Banner, Inner Mongolia, China

There is a significant deficiency for geological mapping in the Gobi Desert terrain covered by wind-blown sands with limited exposure of bedrock. Major and minor element data of stream sediments using Positive Matrix Factorization (PMF) analysis were employed to interpret the geological signatures from the Urat Rear Banner, a representative area of the Gobi Desert terrains. Six primary geochemical factors (M1, M2, M3, T1, T2, T3), corresponding to geological units were extracted from 8 major elements and 18 minor elements. Notably, the SiO₂-Al₂O₃-K₂O-and Pb-Ba-Th-Be-Y-Li-U combinations (primarily M1 and T1) are associated with felsic intrusive rocks, complex metamorphic rocks, and sandstones. The MgO-Fe₂O₃-TiO₂, Cu-Mn-Zn-Co combinations (primarily M2 and T2) indicate basic intrusive rocks, complexes. The Al₂O₃-Na₂O-CaO, Sr-Co-Mn-P-Cu-Zn-Ba (primarily M3 and T3) are related to neutral intrusive bodies (or complexes) and marble (or limestone). Additionally, the study utilizes standardized M1, M2, and M3 factor scores to refine the delineation of lithological boundaries of magmatic rocks. The findings highlight the effectiveness of the PMF method coupled with stream sediment geochemistry, as a powerful tool for extracting geological signatures. This approach offers valuable insights into geological information extraction in Gobi Desert terrains, particularly for identifying and delineating lithological boundaries within magmatic rock formations.