Extracting Climate Signals from Coral Skeletal Geochemistry: Improving Climate Reconstructions from Reef-Building Corals

Abstract

The tropical oceans are the engines of global climate, serving as a critical source of heat and moisture that is distributed poleward. Therefore, climate variability and change in this region have global impacts. However, our understanding of tropical climate variability is limited by sparse and short observational records in the tropical oceans, especially in the Pacific basin. Such short records make it difficult to identify the full scope of natural variability and detect anthropogenic (human-caused) trends. Geochemical records derived from reef-building coral skeletons can help address this knowledge gap. Such records enable the reconstruction of key climate parameters, such as surface ocean temperatures and salinity, over time scales that pre-date the instrumental record by decades or even centuries. However, the process of extracting a climate signal from coral skeletal material can involve conflating factors, such as changes in coral growth in response to environmental extremes, that can introduce biases to the climate reconstruction. In this dissertation, I and my colleagues develop records of coral geochemistry and growth, with an emphasis on the tropical Pacific Ocean. We assess various causes of uncertainties in coral paleoclimate records, and apply this information to improve reconstructions of past tropical ocean climate. In the first study, we present paired records of geochemistry and coral growth from 18th century and modern Galápagos Island corals. The growth metrics include extension rate (i.e., the width of an annual growth band), skeletal density, and calcification rate (the product of extension and density). We find that skeletal density---a less commonly measured component of coral growth---identifies geochemical biases caused by suboptimal sampling of skeletal growth features that are undetected using other metrics. Removing these biased portions of the geochemical records improve sea-surface temperature reconstructions by up to 20%. This study lays the foundation for future reconstructions of eastern tropical Pacific climate over the past four centuries from modern and pre-industrial era Galápagos corals. Anthropogenic changes in the tropical hydrological cycle ("hydroclimate"), such as precipitation and evaporation, can affect drought and flood risk worldwide. Ocean salinity can provide insight into hydroclimate where instrumental data is sparse, and can be reconstructed by coral skeletal geochemistry. The second study leverages a network of such records from across the global tropics to reconstruct spatial and temporal changes in salinity over the 20th century. We compare reconstructions generated from several commonly used salinity data sets. Results show that salinity data sets disagree on the dominant patterns of salinity variability over the late 20th century, and these discrepancies introduce considerable uncertainty into the final coral reconstructions. Nonetheless, these reconstructions reveal long-term trends that are likely associated with hydrological cycle intensification and changes in atmospheric circulation associated with anthropogenic climate change.

The final study investigates the hydroclimate of the western and central Pacific, which is closely linked to the position of the tropical rain belt. We generate synthetic coral records (“pseudocorals”) using a range of assumptions in order to identify which coral sites strongly reflect salinity. We then use these results to develop a network of nine hydroclimate-sensitive coral sites, including two new coral geochemical records from the Marshall Islands. These new records provide key northerly endpoints that enable the reconstruction of the tropical rain belt's location over the mid-to-late 20th century. This reconstruction shows interannual and decadal variability related to meridional (north-south) temperature gradients, and a weak but consistent southward trend in the location of the rain belt over this period. Together, these studies address sources of uncertainty in coral paleoclimate reconstructions. Applying these lessons, these studies further extend the observational record, providing new insights into the climate history of our tropical oceans.