Volcanic stratospheric sulfur injections and aerosol optical depth from 500 BCE to 1900 CE

Abstract

Abstract. The injection of sulfur into the stratosphere by explosive volcanic eruptions is the cause of significant climate variability. Based on sulfate records from a suite of ice cores from Greenland and Antarctica, the eVolv2k database includes estimates of the magnitudes and approximate source latitudes of major volcanic stratospheric sulfur injection (VSSI) events from 500 BCE to 1900 CE, constituting an update of prior reconstructions and an extension of the record by 1000 years. The database incorporates improvements to the ice core records (in terms of synchronisation and dating) and refinements to the methods used to estimate VSSI from ice core records, and it includes first estimates of the random uncertainties in VSSI values. VSSI estimates for many of the largest eruptions, including Samalas (1257), Tambora (1815), and Laki (1783), are within 10 % of prior estimates. A number of strong events are included in eVolv2k which are largely underestimated or not included in earlier VSSI reconstructions, including events in 540, 574, 682, and 1108 CE. The long-term annual mean VSSI from major volcanic eruptions is estimated to be ∼ 0.5 Tg [S] yr −1 , ∼ 50 % greater than a prior reconstruction due to the identification of more events and an increase in the magnitude of many intermediate events. A long-term latitudinally and monthly resolved stratospheric aerosol optical depth (SAOD) time series is reconstructed from the eVolv2k VSSI estimates, and the resulting global mean SAOD is found to be similar (within 33 %) to a prior reconstruction for most of the largest eruptions. The long-term (500 BCE-1900 CE) average global mean SAOD estimated from the eVolv2k VSSI estimates including a constant "background" injection of stratospheric sulfur is ∼ 0.014, 30 % greater than a prior reconstruction. These new long-term reconstructions of past VSSI and SAOD variability give context to recent volcanic forcing, suggesting that the 20th century was a period of somewhat weaker than average volcanic forcing, with current best estimates of 20th century mean VSSI and SAOD values being 25 and 14 % less, respectively, than the mean of the 500 BCE to 1900 CE period. The reconstructed VSSI and SAOD data are available at https://doi.org/10.1594/WDCC/eVolv2k_v2.

A New Volcanic Stratospheric Sulfate Aerosol Forcing Emulator (EVA_H): Comparison With Interactive Stratospheric Aerosol Models

“…Consequently, an important question is whether using EVA_H would significantly affect forcing data sets used in VolMIP or PMIP. We test this hypothesis using the following: A Tambora (1815)‐like eruption with the same injections conditions as those used in Zanchettin et al, (; Figure ), that is, 60 Tg of SO at 0°N and 24 km altitude in April. An Eldgjá (939)‐like eruption with 32 Tg of SO (Toohey & Sigl, ) at 63.6°N and 12.5 km altitude (Moreland, ; 17.5 km for plume top which corresponds to 12.5 km for the umbrella cloud) in April. …”

“…A comprehensive reconstruction of volcanic forcing associated with all eruptions for which the mass, latitude and altitude of injection are constrained is beyond the scope of this paper but is the subject of ongoing work which will greatly benefit from recent efforts to better constrain eruption source parameters (Burke et al, ; Gautier et al, ; Hartman et al, ). However, the preliminary results shown in Figure reinforce the discussion of uncertainties given by Toohey and Sigl () and help to quantify the degree to which the recommended PMIP4 forcing represents an upper estimate for extratropical eruptions.…”

“…However, there is a large spread among the forcing predicted by these models for a specified volcanic injection (e.g., Zanchettin et al, ). This intermodel uncertainty adds to intramodel uncertainties as well as uncertainties related to constraining eruption source parameters, for example, the mass of SO and eruption latitude reconstructed from ice cores when investigating the impacts of past eruptions (Marshall et al, ; Toohey & Sigl, ). Given the computational cost of interactive stratospheric aerosol models, exploring how the propagation of model and source parameter uncertainties affect the assessment of the climate response to a volcanic eruption is challenging and requires significant efforts such as model intercomparison exercises (e.g., Timmreck et al, ; Zanchettin et al, ).…”

“…Another class of models consist of idealized models or “emulators” of volcanic aerosol evolution which have been developed to reproduce the spatiotemporal evolution of volcanic aerosol and associated perturbations of atmospheric optical properties, for example, using constraints from ice cores on the timing and mass of sulfur injected by past eruptions (e.g., Amman et al, ; Crowley & Unterman, ; Gao et al, ; Grieser & Schonwiese, ; Toohey & Sigl, ) or scenarios of future eruptions (Ammann & Naveau, ; Bethke et al, ). Grieser and Schonwiese (), Amman et al (), Gao et al (), and Toohey and Sigl () use emulators based on box models, where each box corresponds to a latitudinal region of the stratosphere. For a prescribed sulfur injection in one of the boxes, the evolution of the mass of sulfate aerosol is governed by time scale(s) for the following: (i) the production of sulfate from SO ; (ii) the mixing between the boxes; and (iii) the loss of aerosol to the troposphere.…”

Multi-model comparison of the volcanic sulfate deposition from the 1815 eruption of Mt. Tambora

“…Previous studies using only a few ice cores to reconstruct volcanic forcing histories may be biased (e.g. Zielinski, 1995Zielinski, , 1996Crowley, 2000), although it has been demonstrated that deposition fluxes derived from single ice cores at high-accumulation sites are representative of total ice sheet deposition (Toohey and Sigl, 2017). Gao et al (2007), who analysed 44 ice cores to investigate the spatial distribution of volcanic sulfate deposition during the last millennium, found larger average deposited sulfate on Greenland (mean deposition of 59 kg km −2 , using 22 ice cores) than on Antarctica (mean deposition of 51 kg km −2 , 17 ice cores) for the eruption of Mt.…”