Investigating ENSO‐Related Temperature Variability in Equatorial Pacific Core‐Tops Using Mg/Ca in Individual Planktic Foraminifera

Abstract

El Niño Southern Oscillation (ENSO) is the largest source of interannual climate variability on Earth today; however, future ENSO remains difficult to predict. Evaluation of paleo‐ENSO may help improve our basic understanding of the phenomenon and help resolve discrepancies among models tasked with simulating future climate. Individual foraminifera analysis allows continuous down‐core records of ENSO‐related temperature variability through the construction and comparison of paleotemperature distributions; however, there has been little focus on calibrating this technique to modern conditions. Here, we present data from individual measurements of Mg/Ca in two species of planktic foraminifera, surface dwelling Globigerinoides ruber and thermocline dwelling Neogloboquadrina dutertrei, from nine core tops across the equatorial Pacific (n ≈ 70 per core for each species). Population variance, kernel probability density functions, and quantile‐quantile analyses are used to evaluate the shape of each Mg/Ca‐temperature distribution and to compare them to modern conditions using monthly temperatures from the Simple Ocean Data Assimilation. We show that populations of individual Mg/Ca measurements in both G. ruber and N. dutertrei reflect site‐specific temperature distribution shapes and variances across the equatorial Pacific when accounting for regional differences in depth habitats. Individual measurements of both taxa capture zonal increases in population variance from the western equatorial Pacific to the central equatorial Pacific and a spatially heterogeneous eastern equatorial Pacific, consistent with modern conditions. Lastly, we show that populations of individual Mg/Ca measurements are able to recover meaningful differences in temperature variability between sites within the eastern equatorial Pacific, lending support to this tool's application for paleo‐ENSO reconstructions.

Modulation of late Pleistocene ENSO strength by the tropical Pacific thermocline

“…The total variability and shape of this distribution is determined by the annual cycle, interannual variability (including ENSO), and decadal and longer variability. Statistical modeling suggests individual foraminifera population variability is diagnostic of ENSO change in the CEP 27 , and core-top calibration has demonstrated that individual foraminifera Mg/Ca captures population variability related to ENSO 29 . Quantile-Quantile (Q–Q) analysis has been used to assess ENSO change in individual foraminifera populations 9 , 10 , and supported by with dispersion statistics provide a framework for determining ENSO change during specific sample intervals and under particular sets of climate boundary conditions.…”

Modeling the impact of bioturbation and species abundance upon discrete-depth individual foraminifera analysis used in ENSO-type climate reconstructions.

“…This noisy under-or oversampling of the distribution tails by IFA also translates directly to sample Q-Q plots (panels c and f in Fig. 5-8), which are commonly used in IFA studies to investigate the population distribution (Ford et al, 2015;Rongstad et al, 2020). This level of noise in the tails increases substantially in the case of increased analytical error, i.e.…”

“…Many IFA studies have gone beyond studying a discrete depth's 1σ SST value and have branched into more forensic studies of a discrete depth's IFA-derived SST distribution. These studies have focussed on analysing the shape of said distribution using various statistical tools, including histograms to infer distribution skewness analysis of histograms (Khider et al, 2011;Leduc et al, 2009), as well as quantilequantile (Q-Q) plots (Ford et al, 2015;Rongstad et al, 2020;Thirumalai et al, 2019;White et al, 2018;White & Ravelo, 2020). Such analysis can reveal apparent shifts in the shape of the downcore, IFAderived SST distribution, which the aforementioned studies have attributed to changes in ENSO-type climate variability.…”

Relationship between individual chamber and whole shell Mg/Ca ratios in Trilobatus sacculifer and implications for individual foraminifera palaeoceanographic reconstructions

“…The general assumption of individual foraminifera analysis is that the measured whole specimen Mg/Ca integrates the entire life history of the individual, providing 2-4 week 'snapshots' of ocean conditions. The correlation of such whole specimen Mg/Ca values from core tops with surface ocean temperature and variability suggests that this value is a representative proxy of past ocean temperatures 15,23 . Differences in population means, variances, and distributions between whole specimen Mg/Ca and individual-chamber (f0, f1, f2) Mg/Ca values demonstrate that individual-chambers do not fully capture the relevant whole specimen population statistics.…”

“…Paleoclimatic interpretation of individual foraminifera Mg/ Ca data involves analysis of dispersion statistics (e.g., variance, standard deviation, kernel density functions or median absolute deviation) and differences in population distributions [7][8][9]16,23 in order to infer the changes in past conditions. Studies using Mg/Ca from individual foraminifera analysis have used both ICP-MS 23 , ICP-OES 16 , and LA-ICPMS [7][8][9]11 analytical techniques. While differing analytical methods may provide similar population mean Mg/Ca results, this is no guarantee that the population distributions are similar.…”