Elemental pigment (hyperspectral imaging) data of the varved sediment record (HZM19) from Holzmaar, Germany

Hyperspectral image (HSI) scanning of the composite record from Holzmaar (HZM19) was measured using a Specim PFD-CL-65-V10 E line scan camera (University of Bern, Switzerland). Data were processed using the ENVI software following the workflow of Butz et al. (2015, doi10.1117/1.JRS.9.096031): data were white-corrected, masked for cracks in the sediment surface and Relative Absorption Band Depths (RABDs) were computed for 2mm wide subsets. RABD671 (band depths from 640 to 702 nm) for Total Chloropigments-a (TChl-a), RABD845 (790 - 900 nm) for Bacteriopheopigments-a (Bphe-a), and RABD620 (600 - 640 nm) for Phycocyanin (PhyCy). To translate HSI indices into absolute concentrations, a pigment extraction was performed at the University of Bern using 23 samples covering the full range of RABD671 and RABD845 index values. Ca 1 g of wet sediment was treated with 100 % acetone following the method of Lami et al. (1994, doi:10.1007/BF00684032) and extractions were measured using a Shimadzu UV-1800 spectrophotometer to obtain bulk concentrations of TChl-a and Bphe-a in µg/g dry sediment using a molar extinction coefficient for TChl-a and Bphe-a. A proxy-proxy calibration was carried out using an ordinary least square regression. After all, only 1.42 % and 0.77 % of datapoints are outside of the calibration ranges for Chl-a (calibration range: 12.75 – 1202.68 µg/g, intercept = -4799.52, slope= 4756,45, r² = 0.8, p-val = 0.00, RMSEP 10-fold = 169.03, RMSEP % = 14.05) and Bphe-a (calibration range 0.38 – 345.12 µg/g, intercept = -1295,8, slope= 1319,7, r² = 0.94, p-val = 0.00, RMSEP 10-fold = 25.26, RMSEP % = 7.32). Ages refer to Birlo et al. (2023) and the related dataset is Model D available via doi:10.1594/PANGAEA.949292.