13To understand past and future sea-level variability, it is important to know if during an interglacial the eustatic sea level 14 is constant or oscillates by several meters around an average value. Several field sites within and outside the tropics 15 have been interpreted to suggest such oscillations during Marine Oxygen Isotope Stage (MIS) 5e (129-116 ka). Here, 16 we present our analysis of one such non-tropical site, Hergla, where a facies succession indicates two foreshore deposits 17 above each other, previously interpreted as MIS 5e sea-level highstand amplified by a second rise. Our study, based on field, 18 microfacies, and optical age Bayesian statistics shows a sea-level rise forming the upper foreshore strata that 19 coincided with the global sea-level rise of the MIS 5a interstadial. The site does therefore not provide evidence for the MIS 5e 20 double peak. We conclude from our analysis that the facies-based proxy is insensitive to small-scale sea-level oscillation. 21 Likewise, uncertainties associated with age estimates are too large to robustly infer a short-term sea-level change. 22 23 120 and the ravinement surface is an erosional surface formed by 121 wave action while the sea level rises. 122 Because quartz OSL tends to underestimate ages beyond 123 ca. 80 ka (Shen and Mauz, 2011), extended dose-recovery test 124 were performed to semi-quantitatively assess potential age 125 underestimation caused by variable saturation doses of natural 126 sedimentary quartz. The ages of the OSL samples, indicated in 127 kilo years (ka) after sampling datum (AD2015), and key stra-128 tigraphic boundaries were modeled with a Bayesian approach. 129 This statistical model takes OSL age data as a probability 130 distribution alongside stratigraphic information to produce age 131 estimates of the OSL samples and of key points in the strati-132 graphy for which no age information is available (e.g., strati-133 graphic boundaries). The Agreement Index is calculated for 134 each sample and for the whole model to evaluate confidence of 135 the modeling results, where a value >60% corresponds to 136 about >95% probability of a chi-squared distribution (Bronk 137 Ramsey and Lee, 2013). To solve an apparent age inversion 138 (see Fig. 2), multiple Bayesian runs were performed until the 139 >60% Agreement Index was obtained (for details see 140 Supplementary Material). Ages of MIS 5 substages were 141 adopted from Martinson et al. (1987). 361 O'Leary, M.J., Hearty, P.J., Thompson, W.G., Raymo, M.E., 362 Mitrovica, J.X., Webster, J.M., 2013. Ice sheet collapse following 363 a prolonged period of stable sea level during the last interglacial. 364