نوع مقاله : علمی -پژوهشی
نویسندگان
1 دانشگاه هرمزگان، بندرعباس، ایران.
2 سازمان انرژی اتمی، تهران، ایران.
3 استاد گروه جغرافیای طبیعی، دانشگاه اکستر، اکستر، انگلستان. Professor of Physical Geography, University of Exeter, Exeter, UK.
4 دانشگاه گنبد کاووس، گنبد، ایران
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
Todays, sediment fingerprinting have been increasingly used to determine source contributions of sediments, but there is a need to continue refining procedure especially with respect to uncertainty analysis during modeling. In this research, we applied a Monte Carlo fingerprinting framework to exploring of uncertainty associate to source contributions of sand dune sediments in the Jazmurian region, Kerman province. In order to, after selecting optimum composite fingerprints using a two-stage statistical procedure including Kruskal-Wallis H test and discriminant function analysis, Collins mixing model was solved 10000 times in the MATLAB with 95% confidence levels (percentiles 2.5% and 97.5%). Among of 10 tracers, four tracers including Cr, Ni, Li and Co were selected as optimum composite fingerprints and then they entered as modeling procedure. Full uncertainty (0-100%) calculated for some sources especially Qs and Qal, and also for some sediment samples, Qt recognized as main source. Variability on the source for sand dune sediment samples is due to winds with multiple directions. The effectiveness of this technique to sand dune sediment fingerprinting suggests that they will be useful in other desert regions with active dune fields for recognition of their sources, quantifying source contributions and studying uncertainty of source contributions.
کلیدواژهها [English]