Climate response to the Toba super-eruption: Regional changes

Abstract

Climatic consequences of the Young Toba Tuff (YTT) eruption about 73 ka are a crucial argument in the current discussion about the fate of modern humans, especially in Africa and Asia. Earth system model (ESM) simulations of the YTT eruption are used to investigate its regional climate impacts, in particular focusing on areas relevant to human evolutionary issues during that time. Uncertainties concerning the stratospheric sulphur emission for the YTT eruption are addressed by comparing ESM simulations of a 100 times Pinatubo-like eruption as an upper and a 3 times Pinatubo-like (Tambora) eruption as a lower estimate. Information about transient changes in vegetation types after the YTT eruption are obtained by forcing an offline dynamical global vegetation model with the climate anomalies simulated by the ESM under both glacial and interglacial background climate conditions. The simulated temperature changes in those areas that were inhabited by humans suggest thermal discomfort, but not a real challenge for survival. Precipitation is reduced in all regions during the first two years but recovers quickly thereafter. Some catchments in these regions (Ganges/Brahmaputra, Nile), experience an over-compensation in precipitation during the third to fifth post-eruption years which is also reflected in anomalously strong river runoffs. Change in vegetation composition may have created the biggest pressure on humans, who had to adapt to more open space with fewer trees and more grasses for some decades especially in the African regions. The strongest environmental impacts of the YTT eruption are simulated under interglacial background conditions suggesting that the climate effects of the YTT eruption did not impact humans on a major scale and for a period long enough to have dramatic consequences for their survival.

Introduction

The Young Toba Tuff (YTT), eruption in northern Sumatra is considered to represent the most recent “super eruption” on Earth based on the amount of solid ejecta making this eruption a magnitude 8¹ event (2500–3000 km³, Rose and Chesner, 1987, Chesner et al., 1991). The timing of the event near the stadial climate prevailing about 73 to 72 ka (Ninkovich et al., 1978, Chesner et al., 1991) initially led to the suggestion that the massive eruption accelerated the long-term cooling trend to the stadial (Rampino and Self, 1992). Zielinski et al. (1996) reported that cooling already had begun before the sulphate peak in the Greenland GISP2 ice core at 71 ± 5 ka, which they attributed to the YTT eruption. Higher resolution oxygen isotope stratigraphy by Stuiver and Grootes (2000) places the sulphate peak near the end of the cooling phase. Schulz et al. (2002) dated the YTT ash layer in their high-resolution benthic and planktonic δ¹⁸O records slightly younger than the Marine Isotope Stage 5–4 boundary, leading to a YTT eruption age of the Toba volcano of about 70 ka, matching Greenland (GISP2 and GRIP) ice-core chronologies. The exact timing of the YTT eruption suffers from large uncertainties in ice core chronologies and other proxy data for the era. No traces of YTT ash have been found in ice cores that would allow unambiguous identification of the origin of isotope peaks. Sediment cores also have small sampling rates and may suffer from reworking. Hence, at best, it can currently be stated that the YTT eruption occurred at some time during the cooling towards the Marine Isotope Stage 5-4 boundary. The exact timing of the eruption is important, as different responses of the climate system would be expected, especially of vegetation cover, to the strong volcanic forcing under different climate background conditions.

Besides the timing problems, there are also other uncertainties of the effects of the YTT eruption. Although the total amount of ejected ash and its sediment pattern is reasonably well documented (e.g. Rose and Chesner, 1987, Chesner et al., 1991, Acharyya and Basu, 1993, Westgate et al., 1998, Jones, 2010), there are considerable uncertainties regarding the sulphur yield of the eruption, which is the parameter mainly determining the global climatic consequences of the eruption. Until the publication of sulphate concentrations in the Antarctic EPICA-Dome C ice core (Traversi et al., 2009), there was no evidence of any sulphate peak around the time of the YTT eruption over Antarctica. Although the data of Traversi et al. (2009) show sulphate concentrations in reasonable agreement with the GISP2 core (∼1200 ppb SO₄²⁺ vs. ∼2000 ppb SO₄²⁺), uncertainty in the dating accuracy (ranging between 1.5 ka and 3 ka), next to the missing direct identification of the ash, leaves it an open question whether or not both spikes result from the same source. If they were indeed a result of the YTT eruption, the estimated sulphur emission would be 60–100 times (510–850 Mt S) that from Mount Pinatubo (June 1991). This estimate is in contrast with results from Scaillet et al. (1998) and Chesner and Luhr (2010) suggesting a sulphur yield of not more than 2–5 times the one from Pinatubo for the YTT eruption, i.e. similar to the 1815 eruption of Tambora. Hence, the main parameter determining the climatic consequences of the YTT eruption is highly uncertain. Its estimates vary by more than one order of magnitude. From the publications currently available, the commonly assumed stratospheric sulphur injection of 100 times Pinatubo is certainly an upper limit. Recent simulations with coupled atmosphere-ocean models (Jones et al., 2005, Robock et al., 2009) yielded a decade of severe cooling by up to 10 K (global mean) for a YTT eruption comparable to 100 times Pinatubo: however, there was no further sustained cooling. Earth System Model (ESM) simulations (Timmreck et al., 2010), also using 100 times Pinatubo sulphur emission but, for the first time, also considering microphysical processes as well, indicate that global mean temperature anomalies are three times smaller than suggested by Jones et al. (2005) and Robock et al. (2009). The reduced response mainly arises from a more complete treatment of stratospheric aerosol formation and growth, leading to much weaker radiative forcing due to larger particle sizes and a faster removal rate.

Climatic consequences of the YTT eruption are a crucial argument in the current discussion about the fate of modern humans especially in India and for African refugees around 70 ka (e.g. Ambrose, 1998, Williams et al., 2009, Haslam and Petraglia, 2010). Of particular interest is temperature, precipitation, runoff, and food supply (e.g. vegetation) changes in potential regions of human refugees, which directly impact the evolutionary aspects (see Fig. 1 for an overview of relevant YTT eruption climate effects). Following Timmreck et al. (2010) who have already analyzed the simulated temperature and precipitation anomalies after the YTT eruption over the Indian subcontinent, the focus of ESM analysis in this study is extended to more regions relevant to human evolutionary issues during that time (see Fig. 2): most importantly, Southern and Eastern Africa (Scholz et al., 2007, Jacobs et al., 2008, Drake et al., 2011 and others), but also South East Asia where evidence was found of a high survival rate of mammalian mega fauna during the time of the YTT eruption (Louys, 2007, Louys, 2012 this volume). The sensitivity of the ESM model results are tested with respect to the assumed sulphur amount of a Tambora-like eruption which is considered to be the lower limit for the YTT eruption. Robock et al. (2009) already investigated the uncertainty in the sulphur emission for the YTT eruption, considering values in the range of 33–900 times the Pinatubo injection, and obtained global maximum cooling between 8 and 17 K. However they did not take into account the temporal evolution of the aerosol size distribution, which turned out to be an important diminishing factor for the volcanic climate signal (Timmreck et al., 2010). Furthermore, the hydrological cycle is investigated in detail by analyzing large river catchments within or close to the regions of interest: Ganges/Brahmaputra for the northern and Krishna for the southern part of India, Orange for Southern Africa, Nile for North East Africa, and Mekong for South East Asia. Vegetation plays the dominant role in the food chain, and therefore the climatic disturbances simulated by the ESM are used as drivers of a dynamical global vegetation model (DGVM) to estimate the lasting effects of the transient climate anomalies on vegetation. In order to account for the uncertainties in the timing of the YTT eruption, the DGVM was forced with climate anomalies under glacial and interglacial background conditions. This allows for an estimate of the upper and lower bounds of vegetation anomalies.