Luminescence dating of glacigenic deposits from northern Germany: A comparison of multigrain aliquots and single grains K-feldspar methods

Glacigenic sediments are challenging to date with luminescence dating technique, particularly with regard to resetting of the luminescence signal before sediment deposition. In this study, six samples from four different ice-marginal positions in northern Germany were investigated to test different luminescence dating methods that account for heterogeneous bleaching of meltwater deposits. Multigrain measurements exhibited agreement between fading corrected infrared stimulated luminescence (IR) and fading corrected post-infrared IR measured at 225 °C (pIRIR₂₂₅) ages, suggesting the studied samples were likely well bleached, as these signals bleach at different rates. Single-grain measurements showed a high proportion of saturated grains (up to ∼33 %) in these samples. The single-grain results were therefore assessed using conventional single-grain analyses and the L_nT_n method. Accurate estimation of the overdispersion (OD) of a well-bleached grain population (σ_b) is essential for applying statistical age models. Because no suitable well-bleached analogue sample with comparable burial doses was available, σ_b was estimated by combining intrinsic OD (from dose recovery tests) and extrinsic OD in quadrature. Both D_e and L_n/T_n OD values exceeded the derived σ_b values, confirming that all studied samples are poorly bleached and that the Minimum Age Model (MAM) is required. MAM ages derived from conventional single-grain and L_nT_n approaches are generally consistent; however, standard MAM ages tend to be slightly younger than L_nT_n MAM ages, likely due to truncation of the D_e distribution through the exclusion of saturated grains. These finding demonstrate that (1) consistent IR₅₀ and pIRIR₂₂₅ ages cannot be considered a reliable indicator of well-bleached samples, and (2) σ_b and overdispersion values should be determined and interpreted carefully, especially for old samples with a large number of saturated grains. We recommend applying the L_nT_n method at the single-grain level when dating partially-bleached samples with grains close to saturation, to obtain more reliable luminescence ages.