Understanding spatiotemporal variations in aeolian sediment sources is vital for developing effective soil conservation strategies in arid environments. This study quantified sediment provenance in the Daranjir Playa (Kavır-e Dar Anjır), central Iran, using geochemical fingerprinting combined with a Bayesian mixing model. Potential sources, including old and young alluvial fans, orchards, salt pan areas, and streambanks, were identified through field surveys and wind analyses. Thirty-two source and twenty seasonal dune samples were analyzed for 25 geochemical elements (Al, As, B, Ba, Be, Ca, Co, Cr, Fe, Ga, K, La, Li, Mg, Mn, Na, Ni, P, Pb, S, Sr, Ti, V, Zn, Zr). Conservative tracers were used to select final tracers using the Kruskal–Wallis H-test and discriminant function analysis, yielding five tracers for spring (Al, B, Mg, Ni, Zn), four for summer (Al, Mg, S, Sr), and five for autumn and winter (Al, B, Mg, Na, Sr). Source contributions were quantified with the MixSIR Bayesian model and validated through virtual mixing experiments. Wind data showed strong seasonal variability, with the highest velocities in summer and dominant sand transport in winter. Model results indicated that salt pan areas were the main sediment source in spring, autumn, and winter (90.3%, 97.4%, and 90.5%), while streambanks dominated in summer (98.3%). Model validation yielded RMSE values ranging from 0.5 to 15.0%, MAE from 0.4 to 9.8%, and an index of agreement (d) between 0.13 and 1.00. This approach elucidates the seasonal dynamics of aeolian sediment sources, reflecting surface conditions and seasonal wind variations, and supports targeted land management in desert landscapes.

Full Text