The Sr/Ca ratio of carbonate coral skeletons is commonly utilized as a paleothermometer to reconstruct sea surface temperatures (SST). Dome-shaped corals (e.g., Porites, Diploria, Siderastrea sp.,) grow all year long and build skeletons that can span more than 300 years into the past, and hence represent valuable archives to reconstruct past climate conditions. Fossil corals provide records of seasonal to interannual variability up to the last Interglacial. These paleo-temperature data sets are crucial for validating models that predict Earth’s future climate.

The Sr/Ca ratio recorded in corals is inversely correlated with the seawater temperature during mineralization. Traditional Sr/Ca analysis uses micro-drilled coral carbonate powder, that provides a spatial resolution of ~0.5 mm, enabling biweekly SST estimates. However, this approach is extremely labour-intensive and time-consuming given the sample numbers needed to reconstruct a single temperature record (~1000 samples / 100 yrs), not including any replication.

Direct in situ laser ablation presents significant potential for enhancing both throughput and spatial resolution. However, the quality of existing data from in situ Sr/Ca measurements using sequential instruments has proven inadequate for high-resolution temperature reconstructions. As a result, there is a strong demand for truly simultaneous multi-element detection.

To reconstruct weekly to biweekly SST in corals with high throughput (75 years/day) we have coupled the only commercially available ICP-OES instrument capable of simultaneous detection of elemental wavelength and adequate time-resolved acquisition of Sr/Ca ratios (SPECTRO Arcos II MV) with a commercial LA system (ESL imageGEO193, equipped with ESL TV2 and DCI). Calibration of intensity ratios is achieved using a series of coral-carbonate nanoparticulate powder pellets with varying Sr/Ca ratios. We will present the performance metrics of this setup for high-precision Sr/Ca analysis by LA-ICP-OES and compare it to traditional solution-based methods.