An Investigation of the Chemical Composition of Bottom Sediments from Dried Lake Gavur of Türkiye by Using XRF and Multivariate Data Analysis

Abstract

Lake sediments offer valuable information about the geographical, climatic and environmental changes. This work examined the chemical elements of eleven sediment samples from the old bottom (0-5m; 0.5m increments) of the dried Lake Gavur in Türkiye by applying XRF along with PCA and PLSR data analyses. The highest but negative correlations were observed for Sr (-0.89), S+SO3 (-0.74) and Zr+ZrO2 (-0.61) with sampling depths meaning that their concentrations were lower in deeper locations (4.0-5.0 m). In contrast, Ti+TiO2 (0.60), Al+Al2O3 (0.53) and Cu+CuO (0.51) had high and positive correlations and their quantities increased as the depth increased. The age of two samples was determined as 4752±33 BP years (2.5 m) and 5470±35 BP (5.0 m). Ca+CaO content was found as lower and Ti+TiO was higher in the deeper sediment (5.0 m) indicating rainy periods. PCA clustered the samples into two groups as deeper samples (4.0-5.0 m) and other samples (0.0-3.5m) while PLSR grouped them into three clusters as deeper (4.0-5.0 m), medium-depth (1.5-3.5 m) and shallower (0.0-1.0m) samples. PLSR model had a good performance to estimate the sampling depth from the chemical elements (R2cal=0.95; R2val=0.67) which means that the variations of the chemicals were highly dependent on the sampling depth.