<p>The carbonate skeletons of marine organisms are unique archives for high-resolution climate reconstructions. Well-preserved specimens potentially allow for seasonal to even daily scale variability reconstructions of climate and environment&#160; in deep time (pre-Quaternary), providing otherwise unavailable snapshots of climate variability during greenhouse periods (e.g. Steuber et al., 2005; Ivany et al., 2008; de Winter et al., 2017). However, uncertainties on past seawater compositions hamper use of the popular stable oxygen isotope ratio (&#948;<sup>18</sup>O) as proxy for paleotemperature reconstructions. The use of the independent carbonate clumped isotope (&#916;<sub>47</sub>) paleothermometer, which is insensitive to changes in seawater composition, on these promising fossil archives is complicated because of sample size limitations (Fernandez et al., 2017; Bernasconi et al., 2018).</p><p>In an attempt to circumvent these issues and use the &#948;<sup>18</sup>O and &#916;<sub>47</sub> measurements jointly for accurate seasonal reconstructions of temperature and seawater isotope composition, we present a novel data reduction approach that combines &#916;<sub>47</sub> measurements of small (~100 &#181;g) serially sampled aliquots to estimate summer and winter temperatures in mollusk shell records. When applied on &#916;<sub>47</sub> and &#948;<sup>18</sup>O measurements in the same specimens, combined with accurate shell chronologies, this approach reconstructs seasonal differences in temperature and seawater composition in a coastal site from the Campanian (Late Cretaceous) high-latitudes.</p><p>To test the robustness of these reconstructions, we apply different approaches of combining &#948;<sup>18</sup>O and &#916;<sub>47</sub> data on a wide range of simulated data representing various scenarios of variability in growth rate, temperature and sea water composition typical for the natural shallow marine environments of carbonate-producers. This approach tests how choices such as sampling resolution and the method of data collection and reduction influence the accuracy and reproducibility of (paleo)seasonality reconstructions in these scenarios.</p><p>Finally, we present preliminary data of &#948;<sup>18</sup>O and &#916;<sub>47</sub> analyses on bivalve specimens grown under controlled temperature conditions that allow us to calibrate the techniques above for temperature reconstructions. Together, these investigations pave the way for accurate, high-resolution climate reconstructions in deep time. These reconstructions provide valuable information on the dynamics of greenhouse climates, against which climate models can be compared to improve predictions of future climate.</p><p><strong>References</strong></p><p>Bernasconi, S. M., et al. Geochemistry, Geophysics, Geosystems, 19(9), 2895&#8211;2914, 2018.</p><p>Fernandez, A. et al. Geochemistry, Geophysics, Geosystems, 18(12), 4375&#8211;4386, 2017.</p><p>Ivany et al. Geological Society of America Bulletin, 120(5&#8211;6), 659&#8211;678, 2008.</p><p>Steuber, T. et al. Nature, 437(7063), 1341&#8211;1344, 2005.</p><p>de Winter, N. J. et al. Palaeogeography, Palaeoclimatology, Palaeoecology, 485, 740&#8211;760, 2017.</p>