Reconstruction of El Niño - Southern oscillation variability during the Holocene

Abstract

The El Niño – Southern Oscillation (ENSO) in the tropical Pacific constitutes the largest source of global climate variability on interannual timescales. Every 2-7 year the El Niño phenomenon causes altered Pacific circulation, leading to widespread droughts and floods. However, the exact mechanisms are still poorly understood and especially the expected behavior of ENSO dynamics in a greenhouse world under increased radiative forcing is heavily debated. A detailed reconstruction of past ENSO activity can provide insight into the spatial and temporal climate variability caused by the ENSO system on centennial to millennial time scales and, at the same time, help assessing the potential role of changes in radiative forcing as driving mechanism for ENSO dynamics. Investigation of terrestrial systems is essential, since the impacts of ENSO are most prominent on land. Botanical proxy records are increasingly used for climate reconstructions since plants respond immediately to changes such as drought or precipitation surplus. Plant remains buried in peat and lake deposits provide the source material utilized in proxy-based climate reconstructions.

In southern Florida wetlands, a strong well-documented ENSO teleconnection can be used for calibration of the botanical proxies. With improved AMS-radiocarbon chronologies it is shown that vegetation during the 20th century, reconstructed by pollen analysis, responded to both annual and decadal-scale changes in wetland hydrology. In addition, a new precipitation proxy based on the analysis of structural xeromorphic features in Quercus laurifolia leaves confims the pollen results. Furthermore, parallel stomatal frequency analysis performed on the fossil leaves reveals a distinct CO2 sensitivity of Q. laurifolia. Combined analysis of CO2 and drought stress signals provides a unique method to synchronously document changing ENSO-tied precipitation patterns and atmospheric CO2 concentrations beyond the period of instrumental measurements.

Spectral analysis and bandpass filtering of the near-annually resolved botanical-proxy records from Florida reveals significant variability, which is highly comparable to the measured ENSO periodicity. These results show that the ENSO signature can be detected in non annually-laminated sediments and allow direct testing of past ENSO dynamics in high-accumulation sediment sequences.

Pollen analyses of Holocene peat deposits from southwest Florida reveal a step-wise increase in wetland vegetation that points to an increased precipitation-driven fresh-water flow during the past 5, 000 years. The subtropical record documents ecosystem responsed to the onset of modern-day ENSO periodicities between ~7, 000 and 5, 000 years BP, known from tropical marine and terrestrial records, and subsequent ENSO intensification after 3, 500 years BP.

These changes are confirmed by pollen analysis from eastern Australia, a region with strong ENSO teleconnections that are anti-correlated with the ENSO impacts in Florida. At a highly precipitation-sensitive site in subtropical Queensland the vegetation strongly diversified into heterogeneous subtropical rainforest after 3 cal ky BP. These results are incorporated into a review of Holocene climate patterns in eastern Australia along a North-to-South transect. Early Holocene changes are divergent and asynchronous between sites, while middle to late Holocene conditions are characterized by a more synchronous change to arid and variable conditions. Based on the present-day Australian climate patterns and impact of ENSO, these results are in agreement with ENSO intensification.

Finally, an integrated overview of marine and terrestrial paleoclimatic data relevant to the detection of ENSO variability is presented. Analysis of proxy climate data indicate that, after a climate-state change at ~5 ky BP towards active ENSO cyclicity in the equatorial Pacific, ENSO-teleconnected regions are characterized by an increased amplitude of ENSO events only around 3 ky BP. Comparison with climate model scenarios shows that the generally accepted view that ENSO intensification results from summer Pacific trade wind reduction cannot completely explain the observed Holocene changes. An additional mechanism is proposed, involving increased Indo-Pacific Warm Pool (IPWP) heat charging, which could have resulted in a 3 ky BP increase of ENSO amplitude.