Optimizing sampling strategies in high-resolution paleoclimate records

Abstract. The aim of paleoclimate studies resolving climate variability from noisy proxy records can in essence be reduced to a statistical problem. challenge is extract meaningful information about these by reducing measurement uncertainty through combining measurements for proxies while retaining the temporal resolution needed assess timing and duration variations parameters. In this study, we explore limits compromise testing different methods data (smoothing, binning, sample size optimization) on particularly challenging problem: seasonal stable isotope records. We test evaluate effects changes temperature hydrological cycle as well accretion rate archive parameters such sampling age model reliability seasonality reconstructions based clumped oxygen analyses 33 real virtual datasets. Our results show that strategic combinations significantly improve accuracy compared conventional analyses, especially settings which isotopic composition water poorly constrained. Smoothing using moving average often leads an apparent dampening cycle, reconstructions. A optimization protocol yields more precise than smoothing. However, most accurate are obtained monthly binning data, cases growth or cycles obscure cycle. analysis wide range natural situations reveals effect almost invariably exceeds composition. Thus, cases, allow reliable identification basis modeling absence independent chronological markers record. These specific findings us formulate general recommendations research have implications beyond reconstruction seasonality. briefly discuss our solving common problems paleoclimatology stratigraphy.