Microscale Alkenone Heterogeneity and Replicability of Ultra-High-Resolution Temperature Records from Marine Sediments

Abstract. The alkenone-derived U^K₃₇ proxy is crucial for the reconstruction of past sea surface temperatures in marine sedimentary archives. Recent advances in mass spectrometry imaging (MSI) now allow to measure alkenone abundance at the micrometer scale. Such an approach can theoretically provide proxy records as highly resolved as observational records and hold the promise of continuously reconstructing climate variability from subseasonal or interannual to centennial and millennial timescales. However, due to processes occurring during and after deposition, as well as during sampling and measurement, it is unclear how much climate signal is preserved in the proxy signal at these small spatial scales. Here, we investigated this question on sediment records from the Santa Barbara Basin (SBB), off California. We performed replicated MSI measurements on sediments with varying degrees of lamination to analyze the spatial structure and spatial reproducibility of the alkenone signal. We find that alkenone distributions are spatially heterogeneous even within laminae but exhibit small scale clustering over the range of ~0.5–1 mm. Measurements along laminated horizons show longer ranges of similarity and less overall variability compared to measurements across depth. Signal to noise ratios (SNR), the amount of shared variance between proxy records derived from the replicates across varying sediment conditions range from ~1 SNR at interannual resolution to ~3 SNR at subdecadal timescales.

MSI-based U^K₃₇ records in the SBB, supported by careful estimation of noise and uncertainty, thus can capture subdecadal SST variability and provide an upper limit for the signal content of Holocene and late Pleistocene SST reconstructions.

The approach presented here can be used in other settings to infer optimal sampling and measurement resolution, as well as to provide uncertainty estimations for the proxy records.