نوع مقاله : مقاله پژوهشی
نویسندگان
1 استادیار دانشکده منابع طبیعی دانشگاه ارومیه
2 استاد دانشکده منابع طبیعی دانشگاه تهران
3 دانشیار پژوهشی مرکز تحقیقات حفاظت خاک و آبخیزداری
4 عضو هیئت علمی دانشکده منابع طبیعی دانشگاه ارومیه
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
Owing to many drawbacks associated with traditional methodss for identifying sediment sources, fingerprinting techniques, based on physical, chemical and organic properties of sediment and source materials, are increasingly being used as an effective alternative approach to assembling such information. In this method, a suitable composite (set) of diagnostic properties and a multivariate mixing model were employed to estimate the relative contribution of sediment sources to sediments transported to basin outlet. In this study, using suitable composites of geochemical elements, radionuclides, organic Carbon, Nitrogen and Phosphorous, capable of discriminating differerent land uses of the study basin, and a multivariate mixing model was used to determine the contributions of these land uses to sediment yield. The suitable composite fingerprints (elements) were obtained using discriminant analysis.The study basin is the main subbasin of Pouldasht water spreading station basin, Margan, located in Makoo township, Western Azarbaijan province, Iran. The suitable composite fingerprints having capability to distinguish land uses consisted of 137Cs, OC, Pb, Be, Sn, P, Ti and Cu. Mean contributions from the three main land uses were in the order: rangelands with gully, channel and river bank erosions (65.85%), rangelands with sheet and rill erosions (13.81%), agricultural lands (20.34%). Low mean absolute errors (less than 22%) show high degree of correspondence between measured and predicted properties. High model efficiencies (greater than 0.999) confirm the goodness of fit of the mixing models. Also it is argued that fingerprinting estimates for sediment sources are consistent with field observations.
کلیدواژهها [English]