Volcanoes and Climate, part 1.

Guest Lecturer: Ellen Thomas (Office 459, extension 2238; ethomas@wesleyan.edu)

Date: 15 November 2004


This is the first of two lectures on ‘Volcanoes and Climate’. This lecture will deal will with observed effects of volcanoes on climate outside the direct area of eruption, i.e., concentrate on historical eruptions and their global climate effects. The second lecture (Wednesday November 17) will concentrate on speculations on much larger-scale effects in the geological past (mass extinctions).

Readings for this lecture:

Additional on-line information

Main Points of today’s lecture


It has been suggested for a very long time that volcanoes may affect climate: 44AD, several Roman authors including Seneca, Plutarch: strange sunsets because of Etna eruption; 1783-1784: Ben Franklin remarked on eruption on Iceland, was Laki eruption; 1816: year without summer, Tambora eruption Indonesia1815; 1883: Krakatoa eruption Indonesia; 1963: Agung eruption Indonesia. Modern studies on climate effects of volcanic eruptions got a start with the 1982 El Chichon eruption (Mexico), and became more organized, involving many satellite observations, with the 1991 Pinatubo eruption, (Phillipines).

In order to work out what the effect is of volcanic eruptions on climate, we must include information on past volcanic eruptions and climate, and thus must include estimates of the magnitude of eruptions in the past. These data have come from:

  • Historical documents
  • Ice cores
  • Geological record (sediment cores)
  • Tree rings
  • Climate modeling (click here to download pdf file paper on modeling atmospheric impact of Laki eruption)

What determines how much effect a volcanic eruption has on climate?

  • Magnitude of eruption
  • Location of volcano
  • Chemistry of erupted material

Magnitude of eruption:

Volcanic eruptions have been assessed in various ways, but the size of eruptions is not easily quantified, especially if we want to quantify the size of past eruptions.

DVI: Lamb (1970) proposed to use the 'Dust Veil Index' while studying the effect of volcanoes on ‘surface weather, on lower and upper atmospheric temperatures, and on the large scale wind circulation. This parameter is defined by a rather complex (and partially subjective) combination of descriptions of historical reports of eruptions, optical phenomena, radiation measurements when available, temperature information, and estimates of the volume of ejecta. DVI values have been calculated/estimated for volcanic eruptions since 1500 AD and are available on line.

VEI: Newhall and Self described the ‘Volcanic Explosivity Index’, based on geological information (how much material was erupted; how high was the plume). This index estimates the size of volcanic eruptions of the past, without actual information on climate impact. Example of classification:


Plume Height







>103 m3





100-1000 m

>104 m3





1-5 km

>106 m3



Galeras 1992


3-15 km

>10.106 m3



Ruiz 1985


10-25 km

>0.1 km3


>10 yr

Galunggung 1982


> 25 km

>1 km3


>100 yr

St Helens 1981


> 25 km

>10 km3


>100 yr

Krakatoa 1883


> 25 km

>100 km3


>1000 yr

Tambora 1815


> 25 km

>1000 km3


>10,000 yr

Toba, 71 ka

Newhall and Self (J. Geophys. Research, v. 87, p. 1231&endash;1238, 1982).

There are other indices:

  • Mitchell’s index (~1970; mainly based on estimates of aerosol mass).
  • Sato’s index (1993), expressed as ‘optical depth’, i.e., a measure of the amount of extinction of radiation along a path through the atmosphere, proportional to the amount of material along that path (at wave length 0.55 mm). Values are based on Mitchell’s estimates for eruptions 1850-1883, optical data after 1882, satellite data after 1979.
  • Ice Core Volcanic Index: use data from ice cores, mainly based on sulfate levels,  estimate loss of radiative energy at wave length 0.55 mm; Robock and Free, 1995.

Location of volcano:

Does the plume penetrate the stratosphere (i.e., leave the troposphere and penetrate the tropopause)? Small particles in the troposphere are removed much more quickly (e.g., rain out) than particles in the stratosphere. Tropopause: ~16-17 km at equator, ~ 8 km at poles; rather sudden change in height at mid latitudes. Plumes from tropical volcanoes circle the earth in one or a few weeks, but plumes from high latitude eruptions usually do not penetrate to the other hemisphere.

Chemistry of erupted material:

Volcanic eruptions can put into the stratosphere teragrams (1015 grams; thousand of tons) of chemically active gases and small particles. These disturb the Earth’s radiation balance and the stratospheric chemical equilibrium.

What comes out of volcano in plume:

  • Ash (old material, blown to pieces)
  • Ash (juvenile material; molten drops)
  • Gases: H2O, CO2, SO2, H2S, H2, HCl, HF, CO, He, N2.

Note that the gases are not necessarily ‘juvenile’, i.e., derived from mantle sources; a large percentage of all gases reflect crustal recycling. Steam is usually the most abundant gas, followed by CO2, which is abundant even in emissions from non-active volcanoes (e.g., Lake Nyos).

Ash particles tend to fall out rapidly within the troposphere (minutes &endash; weeks), and have little to no influence outside the area of eruption. Gases are the main drivers of climate effects of large spatial and temporal scales. Note that some of the gases are greenhouse gases (and thus could be expected to cause warming), whereas others react to particles and cause cooling at least in some places in the atmosphere. In addition, there are indirect effects on atmospheric circulation. The total effects of volcanic eruptions on climate thus can be complex.

The major cooling effect of volcanic eruptions results from the emission of sulfur dioxide gas (SO2) into the stratosphere. Global sulfur emissions by volcanoes are estimated to be on average ~14% of total natural and anthropogenic emissions to the troposphere. Sulfur dioxide is converted into sulfate aerosol (mainly H2SO4 droplets) by reaction with OH- radicals and water, within a few weeks. Most particles have a diameter between 0.2-0.5 mm, i.e., similar to the wave-length of visible light, and they thus strongly interact with solar radiation by scattering light. Within the troposphere, most sulfate particles are removed (rain out) within a week, but they persists for much longer in the stratosphere.

If there is an aerosol cloud in the stratosphere, some short-wave solar radiation is back scattered (back into space); this is the dominant radiative effect at the Earth’s surface, causing cooling. The reflection of the setting sun from the bottom of the stratospheric volcanic aerosol  layers (commonly called ‘dust veils’) produces the well-documented red sunsets after a large volcanic eruptions.

The time that particles persist in the stratosphere is commonly expressed in e-folding decay time,  i.e. , the time that it takes for an anomaly to decay to e-1 of its initial value (time scale of a logarithmic process). This time is about 1 year for volcanic particles.

Some solar radiation is scattered forward, resulting in increased diffuse radiation (slightly counteracting the cooling), making the sky milky white rather than blue.  The stratosphere, however, in contrast to the troposphere, is heated, as the result of near infrared absorption of solar energy at the top of the aerosol cloud, and increased infra-red absorption of long-wave radiation from the Earth’s surface.

Eruptions cause local cooling for several days, as a direct consequence of the large masses of ash in the troposphere close to the eruption; the difference in temperature during day-night is eliminated for a few days. The effect of the radiative forcing resulting from back scattering of solar shortwave radiation results in cooling of the surface Earth (~0.1-0.2oC), globally after large (sulfur-rich) eruptions. 

The stratospheric warming in the region of the stratospheric cloud increases the latitudinal temperature gradient after an eruption at low latitudes, disturbing the stratospheric-troposphere circulation, increasing the difference in height of the troposphere between high and low latitudes, and increasing the strength of the jet stream (polar vortex, especially in the northern hemisphere). This leads to warming during the northern hemisphere winter following a tropical eruption, and this warming effect tends to be larger than the cooling effect described above.

There are other effects in the stratosphere, especially as to its chemistry. The large amounts of small particles (i.e., large surface area) serve as template for heterogeneous chemical reactions in the stratosphere, specifically the photochemical ozone-depletion reactions. These reactions take place in the presence of chlorine and other halides. Some of these are present in the volcanic cloud itself, but more have been supplied anthropogenically (CFCs). Ozone depletion after volcanic eruptions thus is enhanced as the result of anthropogenic activities.

In short:

  • Large volcanic eruptions inject sulfur gases into the stratosphere; the gases convert into submicron particles (aerosol) with an e-folding time scale of about 1 year.
  • The climate response to large eruptions (in historical times) lasts for several (2-3) years.
  • The aerosol cloud causes cooling at the Earth’s surface, warming in stratosphere
  • For tropical eruptions, stratospheric warming is much more severe at low latitudes than at high latitudes, thus latitudinal temperature gradient increases, leading to a stronger jet stream and northern hemisphere winter warming.
  • Volcanic aerosol particles serve as template for heterogeneous reactions in the stratosphere, thus enhance ozone destruction

The figure is Plate 1 in Robock, A., 2000, Volcanic Eruptions and Climate, Reviews of Geophysics, 38 (2), 191-219.

Figure: Schematic diagram of volcanic inputs to the atmosphere and their effects. This is an extended version of Figures 1 and 2 of Simarksi (1992), drawn by L. Walter and R. Turco.