How precipitation intermittency sets an optimal sampling distance for temperature reconstructions from Antarctic ice cores
Many palaeoclimate proxies share one challenging property: they are not only driven by the climatic variable of interest, e.g., temperature, but they are also influenced by secondary effects which cause, among other things, increased variability, frequently termed <q>noise</q>. Noise in...
| Main Authors: | , , |
|---|---|
| Format: | Text |
| Language: | English |
| Published: |
2020
|
| Subjects: | |
| Online Access: | https://doi.org/10.5194/cp-2020-128 https://cp.copernicus.org/preprints/cp-2020-128/ |
| Summary: | Many palaeoclimate proxies share one challenging property: they are not only driven by the climatic variable of interest, e.g., temperature, but they are also influenced by secondary effects which cause, among other things, increased variability, frequently termed <q>noise</q>. Noise in individual proxy records can be reduced by averaging the records, but the effectiveness of this approach depends on the correlation of the noise between the records and therefore on the spatial scales of the noise-generating processes. Here, we review and apply this concept in the context of Antarctic ice-core isotope records to determine which core locations are best suited to reconstruct local-to-regional-scale temperatures. Using data from a past-millennium climate model simulation equipped with stable isotope diagnostics we intriguingly find that even for a local temperature reconstruction the optimal sampling strategy is to combine a local ice core with a more distant core ~ 500–1000 km away. A similarly large distance between cores is also optimal for reconstructions that average more than two isotope records. We show that these findings result from the interplay of the two spatial scales of the correlation structures associated with the temperature field and with the noise generated by precipitation intermittency. Our study helps to maximise the usability of existing Antarctic ice cores and to optimally plan future drilling campaigns. It also broadens our knowledge on the processes that shape the isotopic record and their typical correlation scales. Finally, the presented method can be directly extended to determine optimal sampling strategies for other palaeoclimate reconstruction problems. |
|---|