Intercomparison of proxy data and model simulations as key to understanding internal variability of the South American Monsoon System

The South American Monsoon System (SAMS) is the dominant mode of hydrologic variability in South America, and varies on interannual to centennial timescales. As anthropogenic forcings continue to influence the climate, understanding the natural variability of this system is an essential basis for diagnosing sensitivity to these external forcings. This research compares the historical variability of the monsoon captured by a network of δ18O records from isotopic proxies (including speleothems, ice cores and lake sediments) across the SAMS domain to monsoon simulations from a global climate model.

Initial results from comparisons between proxy records and precipitation simulated by the NCAR Community Earth System Model Last Millennium Ensemble (CESM-LME) suggest the model represents temporal variability (of monsoon strength) reasonably well across the region. Additionally, both data sources isolate the monsoon core and South Atlantic Convergence Zone as leading modes of variability.

Using the CESM-LME single-forcing runs, we have examined the influence of distinct forcings on SAMS variability. Applying a Monte Carlo Principal Component Analysis technique to the isotopic records and corresponding locations within the model space, we decompose spatiotemporal modes of monsoon variability. The comparison between model simulations and proxy data allows us to better understand the underlying internal and external forcing mechanisms that influence SAMS variability on different timescales. Looking forward, we anticipate that these results may help us constrain our understanding of how changes to external forcings will influence the SAMS in the future.