Statistical interpretation of the relation between geochemical and mineralogical data obtained from lake and floodplain sediments (Amuq Valley / Turkey)


Creative Commons License

Aykasım D.

XVII International Clay Conference, İstanbul, Türkiye, 25 - 29 Temmuz 2022, ss.63

  • Yayın Türü: Bildiri / Özet Bildiri
  • Basıldığı Şehir: İstanbul
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.63
  • Orta Doğu Teknik Üniversitesi Adresli: Evet

Özet

Amuq Valley region, that includes Amuq Lake and floodplains of Orontes, Afrin, and Karasu River, and its

surroundings have always been a focus of interest for scientists as they provide a unique data collection that

preserves both geological and archaeological records of environmental and social turbulences experienced

during the Holocene epoch. In the Amuq Valley region, mineralogical and elemental geochemical investigations

were carried out along 149 samples from undisturbed sediment cores that were taken from the Amuq Lake

and from the vicinity of the archaeological sites of Tell Kurdu, Tell Atchana, and Tell Tayinat. Mineralogical

content of the sediments along these cores were examined by X-Ray Diffraction (XRD) and SEM-EDS.

Inductively Coupled Plasma – Optical Emission Spectrometry / Mass Spectrometry (ICP-OES/MS) analyses

were conducted on the samples to detect quantitative concentrations of major, minor and trace elements.

A method was developed by MapInfo software to convert intensities of bulk powder and oriented airdried

clay fraction X-ray diffractogram peaks to numerical values. Numerical values obtained from XRD

and ICP-OES/MS analyses were correlated by Correlation Matrix and Principal Component Analysis. Hence,

mineral-element relationships were statistically identified. The main purpose of this study is to reveal

statistical correlations between geochemical and mineralogical data obtained and characterized from the

lake and floodplain sediments of Amuq Valley and uncovering the relationship between geochemistry and

mineralogy of the lake and floodplain samples by providing some statistical evidences.

Statistical evaluation of the data revealed that quartz, serpentine, feldspar group minerals and clay minerals

have higher correlation coefficients in t he floodplain samples wh ile ca lcite an d ar agonite ha ve higher

correlation coefficients in the lake samples. A significant effect of environmental difference was determined

on aragonite correlations. It was also examined that the clay mineral peaks before 20° 2θ on bulk powder Xray

diffractograms have stronger correlations with quartz, serpentine, and feldspar group minerals in contrast

to the other clay mineral peaks. Moreover, element vs. mineral correlations revealed possible presence of

impurities like rutile in quartz due to high correlation coefficients of Si, Al, Fe, Mn, Cu, Ti, Hf, Zr, and Nb,

while other minerals generally correlate with their chemical constituents except clay minerals which also

correlate with trace elements including rare earth elements such as Hf, ,Y La, Ce, Pr, Nd, Sm, Eu, Tb, Dy,

Th, Ga, Er, Yb, Lu, Zr, and Nb. These results also showed that application of this statistical analysis

technique is an efficient and timesaving approach on large data sets.