Developing a Forward Model of Encrusting Coralline Algae

Friday, 19 December 2014
Jessica Ng1,2, Branwen Williams2, Diane M Thompson3 and Jochen Halfar4, (1)Organization Not Listed, Washington, DC, United States, (2)W.M. Keck Science Department, Claremont, CA, United States, (3)National Center for Atmospheric Research, Boulder, CO, United States, (4)University of Toronto, Earth Sciences, Toronto, ON, Canada
Climate proxy data has traditionally been interpreted through inverse models, which extract physical climate variables from proxy variables. This approach assumes stationarity of the proxy-climate relationship, typically reduces climate signal to a single variable, and requires extensive observational records. In contrast, forward models reverse the relationship, simulating proxy variables from physical climate variables for comparison to observed proxy variables. Since this approach accounts for multiple climate variables and avoids stationarity issues, forward models have been developed for several climate proxies, including tree ring width and oxygen stable isotopes (δ18O) of corals.

Here we develop a basic forward model for the climate archive coralline alga Clathromorphum sp.This long-lived alga grows in mid-to-high latitude regions and forms a solid calcite skeleton with annual growth bands similar to those of trees and tropical corals. Sub-annually resolved δ18O in annual growth bands (δ18Ospec) provide a record of local environmental and climatic factors, notably sea surface temperature (SST) and sea water oxygen stable isotopes (δ18Osw).

We model Clathromorphum δ18Ospec in the Aleutian islands from gridded SST and δ18Osw of the growing season from 1960 to 2004. The strongest climate signal is observed in July, likely due to suppressed growth in other months. Initial results suggest no influence of growth rate on the fractionation of oxygen isotopes and indicate that δ18Ospec anomalies are significantly correlated with summer SST anomalies. We run this forward model with observed SST and δ18Osw and compare the resulting simulated δ18Ospec with that measured in live-collected specimens. This foundational model may be adapted to other regions and modified to include other variables influencing coralline isotope records, such as light availability and ice coverage.