Centennial variability of the Afro-Asian monsoon during the Holocene

The monsoon plays a major role in the Asian and African climate. Its variability exerts a strong control of water resources in lots of countries and its future evolution is of concern. While there is extensive knowledge of its mean-state evolution during the Holocene, its centennial variability has remained little explored. Such variability scale cannot be explored from the too short instrumental observation period, but high-resolution paleoclimate archives, such as speleothems, allows to access to indirect measurements of monsoon variability over long time scales.

In this work, we use data from an IPSL-CM6A-LR simulation to investigate this range of variability, both in spatial domains, using ordination techniques derived from Principal Component Analysis, and in frequency domains, using spectral analysis.

To assess the model correspondence to climate reconstructions, we first compare the simulated precipitation with speleothems δ¹⁸O records from the SISALv2 database¹, considering the long-term trends. The speleothems δ¹⁸O records constitute a composite proxy of temperature and precipitation. Following Parker et al. 2021², we applied a principal coordinate analysis (PCoA) to the δ¹⁸O and precipitation datasets in order to explore their spatial similarities. In both cases, a strongly predominant first coordinate is found. However, it explains more variance in the precipitation data (about 90%) than in the δ¹⁸O data (about 70%). This ordination technique also makes it possible to discuss similarities between regions by performing a clustering in the reduced PCoA space. A strong coherence is found in Asian monsoon variability, while the African monsoon is shown to be closer to the South American monsoon.

We then explore the centennial band of variability in the Fourier spectrum of precipitation time series (from simulation) for each tropical monsoon region. In this centennial band, most regions exhibit a white-noise spectrum, indicating that monsoon variability on these timescales has no memory. Significant peaks are identified in the East Asian monsoon.

References

1 Comas-Bru, L., Atsawawaranunt, K., Harrison, S., and SISALworking group members: SISAL (Speleothem Isotopes Synthesis and AnaLysis Working Group) database version 2.0, University of Reading [data set], https://doi.org/10.17864/1947.256, 2020a.

2 Parker, S. E., Harrison, S. P., Comas-Bru, L., Kaushal, N., LeGrande, A. N., & Werner, M. (2021). A data–model approach to interpreting speleothem oxygen isotope records from monsoon regions. Climate of the Past, 17(3), 1119-1138.