ULF wave measurements onboard the Interball auroral probe

The IESP experiment implemented onboard the Interball auroral probe measures the six components (3B, 3E) of the waves in the ULF range: 0.1–10 Hz and from time to time 0–30 Hz. Two different kinds of waves have been observed in the auroral region at altitudes between 10 000 and 20 000 km: (1) electrostatic emissions which consist of quasi-monochromatic structures with frequencies above the oxygen gyrofrequency, superimposed on a wide band signal interpreted as a Doppler broadening, (2) electromagnetic wide band spectrum fluctuations. These emissions are interpreted as current-driven electromagnetic or electrostatic ion cyclotron waves. The electromagnetic/electrostatic character is controlled by the plasma parameter _i and by the O⁺ concentration.     

Influence of electromagnetic fields on the seismotectonic strain rate; relaxation and active monitoring of elastic stresses

We study the changes in the rate of seismotectonic strains in the Earth’s crust in two of the most seismically active Central Asian areas, the Garm region in Tadjikistan and the Northern Tien Shan, which are excited by high energy pulsed electromagnetic irradiation. It is shown that in both areas the average rates of strain release increase by a factor of 10–20. The increment of the seismotectonic strains initiated during the entire period of electromagnetic treatment is 1.3 × 10⁻⁶–6.3 × 10⁻⁶, which is comparable with the value of crustal strains observed during the preparation of strong earthquakes and amounts to 3–13% of the ultimate strains for rocks. Such a contribution to the process of quasi-plastic strains of the crust results in the additional relaxation of elastic stresses by 0.4–1.67 bar, which corresponds to approximately 1–17% of the stress released in the sources of strong earthquakes. The spatial variations in the changes of the strain rate during electromagnetic treatment are studied, and their correlation with the level of stresses in the Earth’s crust is revealed. It is shown that the change in the strain rate during the electromagnetic excitation can be used as the criterion for active monitoring of the stress state of the medium for the purposes of predicting the location, time, and magnitude of strong earthquakes.     

Research on amplitude spectrum of precursory electromagnetic radiation associated with the impending great earthquake of Gonghe in Qinghai Province

The high accuracy, automatic, digital MT measurement system with five electromagnetic components was firstly adopted by the author to the section MT monitor and measurement in the middle Qilian mountain in the Northwestern China, the more real precursory information of impending earthquake electromagnetic rediation of GongheM=7.0 earthquake (April, 1990) in Qinghai province was obtained by the system. The new complete quantification frequency analysis was firstly used in this paper in both time and space domains for the analysis of repeated measurement data at all sites. This paper concludes the precursory spectral characteristics of electromagnetic radiation before strong earthquake as following: 1. The spectra appears as the synchronous electric field interference type mainly of conduction current; 2. Narrow band (0.5 – 20s) of spectral period; 3. The singularity of amplitude (the pulse amplitudes of interference electric field is 10² times the normal values); 4. They show the directionality of preparing focus to some extent; 5. The intermittent of radiation (three peak periods appeared three days before the main shock). It is more significant that there existed a relative quiet period of about 48 hours between the peak period and the occurrence, observators can response quickly at observational sites and take it as one of the observational criterions for impending prediction. This paper also points out that the impending earthquake electromagnetic radiation obtained by the MT measurement system is mainly the conduction current, and the high-conductive layer existed commonly in the crust may serve as the better paths for the conduction current circuit and enables the far away MT measurement sites to receive it. The Chinese version of this paper appeared in the Chinese edition ofActa Seismologica Sinica,15, 186–193, 1993.     

The KVVN-7 multifunction digital measuring station for electromagnetic monitoring of seismoactive zones

The portable highly sensitive measuring station KVVN-7 has been designed. It enables us to perform frequency sounding with controllable sources and audio-magnetotelluric sounding in the field of natural variations in an electromagnetic field within one session. Signal recording is made by seven channels (three magnetic and four electric ones) in a broad frequency spectrum (0.1–2000 Hz) with elimination of frequencies on the edges of the set frequency range and in the odd harmonics of industrial frequency (up to the ninth harmonic). The station incorporates the system of band-stop filters and anti-alias filters at resistors having a low temperature coefficient of resistance (TCR is lower than 25 × 10⁻⁶ °C⁻¹) and capacitors having a low temperature coefficient of capacity (TCC is lower than ±30 × 10⁻⁶ °C⁻¹). Application of the KVVN-7 station allows the electric conductivity and fluid regime of the upper crust to be studied for both implementation of geological tasks and tasks related to electromagnetic monitoring of seismoactive zones in combination with seismic methods. The example of practical application of the KVVN-7 station in the Lovozero-Pulozero profile (Kola Peninsula) has been presented. A high effectiveness of a station has been demonstrated when study of a fluid-saturated layer with conductivity of a dilatancy-diffuse origin (“DD layer”) in the upper crust. The further perfection of the KVVN-7 measuring station is aimed at the design of a completely automated recording system through data recording to a built-in data medium (flash memory). Additionally, it is suggested to use a built-in analog-to-digital converter of high resolution (24 bit) for every channel in order to broaden the dynamical range of the station.     

Properties of electric turbulence in the polar cap ionosphere

Small-scale (scales of ∼0.5–256 km) electric fields in the polar cap ionosphere are studied on the basis of measurements of the Dynamics Explorer 2 (DE-2) low-altitude satellite with a polar orbit. Nineteen DE-2 passes through the high-latitude ionosphere from the morning side to the evening side are considered when the IMF z component was southward. A rather extensive polar cap, which could be identified using the ɛ-t spectrograms of precipitating particles with auroral energies, was formed during the analyzed events. It is shown that the logarithmic diagrams (LDs), constructed using the discrete wavelet transform of electric fields in the polar cap, are power law (μ ∼ s ^α). Here, μ is the variance of the detail coefficients of the signal discrete wavelet transform, s is the wavelet scale, and index α characterizes the LD slope. The probability density functions P(δE, s) of the electric field fluctuations δE observed on different scales s are non-Gaussian and have intensified wings. When the probability density functions are renormalized, that is constructed of δE/s ^γ, where γ is the scaling exponent, they lie near a single curve, which indicates that the studied fields are statistically self-similar. In spite of the fact that the amplitude of electric fluctuations in the polar cap is much smaller than in the auroral zone, the quantitative characteristics of field scaling in the two regions are similar. Two possible causes of the observed turbulent structure of the electric field in the polar cap are considered: (1) the structure is transferred from the solar wind, which is known to have turbulent properties, and (2) the structure is generated by convection velocity shears in the region of open magnetic field lines. The detected dependence of the characteristic distribution of turbulent electric fields over the polar cap region on IMF B _y and the correlation of the rms amplitudes of δE fluctuations with IMF B _z and the solar wind transfer function (B _y ² + B _z ²)^1/2sin(θ/2), where θ is the angle between the geomagnetic field and IMF reconnecting on the dayside magnetopause when IMF B _z < 0, together with the absence of dependence on the IMF variability are arguments for the second mechanism.     

The changing morphology of an open lava channel on Mt. Etna

An open channel lava flow on Mt. Etna (Sicily) was observed during May 30–31, 2001. Data collected using a forward looking infrared (FLIR) thermal camera and a Minolta-Land Cyclops 300 thermal infrared thermometer showed that the bulk volume flux of lava flowing in the channel varied greatly over time. Cyclic changes in the channel's volumetric flow rate occurred over several hours, with cycle durations of 113–190 min, and discharges peaking at 0.7 m³ s⁻¹ and waning to 0.1 m³ s⁻¹. Each cycle was characterized by a relatively short, high-volume flux phase during which a pulse of lava, with a well-defined flow front, would propagate down-channel, followed by a period of waning flow during which volume flux lowered. Pulses involved lava moving at relatively high velocities (up to 0.29 m s⁻¹) and were related to some change in the flow conditions occurring up-channel, possibly at the vent. They implied either a change in the dense rock effusion rate at the source vent and/or cyclic-variation in the vesicle content of the lava changing its bulk volume flux. Pulses would generally overspill the channel to emplace pāhoehoe overflows. During periods of waning flow, velocities fell to 0.05 m s^–1. Blockages forming during such phases caused lava to back up. Occasionally backup resulted in overflows of slow moving ‘a‘ā that would advance a few tens of meters down the levee flank. Compound levees were thus a symptom of unsteady flow, where overflow levees were emplaced as relatively fast moving pāhoehoe sheets during pulses, and as slow-moving ‘a‘ā units during backup. Small, localized fluctuations in channel volume flux also occurred on timescales of minutes. Volumes of lava backed up behind blockages that formed at constrictions in the channel. Blockage collapse and/or enhanced flow under/around the blockage would then feed short-lived, wave-like, down-channel surges. Real fluctuations in channel volume flux, due to pulses and surges, can lead to significant errors in effusion rate calculations. Editorial responsibility: A. Woods     

Statistical features of the seismic noise-field

We present a preliminary study of the dependence of the statistical features of the soil motion due to seismic noise on the near-surface geology in the frequency range from 1 Hz to ∼ 40 Hz. In detail, we have investigated the 3D average squared soil displacement 〈r²〉 and the distribution function of the displacement flctuations at different geological sites. The anomalous scaling of the average squared soil displacement 〈r²(τ)〉~τ_α, and the Gaussian shape of the probability distribution function of its fluctuations suggest that the soil motion under the influence of the seismic noise is consistent with a persistent fractional Brownian motion (fBm) characterized by a scaling exponent 1.5 < α < 2. Therefore, the seismic noise-field, thought as a stochastic process, shows a markovian character with a memory longer than a pure Brownian motion (α = 1/2). Moreover, a dependence of such persistent behavior of the noise-field dynamics on the near-surface local geology has been found and it is discussed.     

Investigation of the broadband ELF turbulence by observations of the FAST satellite

Scaling properties of variable electric fields in the topside ionosphere have been investigated on scales s from ∼30 m to 2 km by FAST electric field observations with sample rate of 512 s⁻¹, in sixteen events of the broadband ELF turbulence. It is shown that down to scales of a few hundred meters, the power of turbulent electric fluctuations is a power law, ∼s ^α. Scaling index α derived from the slope of logarithmic diagrams (LD) constructed by the discrete wavelet transform of data can be estimated as α = 2.2 ± 0.3, which is close to α estimate earlier reported for scales 1–30 km by electric field observations of the Dynamics Explorer 2 satellite. The behavior of α index is analyzed near the scale of the order of electron inertial length λ_e = c/ω₀ (ω₀ being the electron plasma frequency). At altitudes considered (700–2500 km), λ_e makes 100–900 m. We demonstrate that at scales ≤λ_e, a decrease of LD slope and deviation from the power law are typically observed. As pointed out in the discussion, this feature cannot be identified as a transition to the diffusion range, where dissipation of the turbulence occurs.     

Fine structure of Pi2-type geomagnetic pulsations

An analysis of the BEAR experiment and Polar satellite data showed that several ionospheric sources could act almost simultaneously in the area of Pi2-type geomagnetic pulsations. The aim of this study is to locate these sources and determine their coherence in the narrow-frequency band of (6–10) × 10⁻³ Hz. By using gradient analysis methods, we revealed the local sources of oscillations coherent (phased), that allows one to treat their contributions as manifestations of the fine structure of the field of Pi2 pulsations.     

Characteristics and quantitative of negative ion in salt aqueous solution by Raman spectroscopy at ?170°C

The results from Raman spectroscopy analysis of salt aqueous solutions at −170°C demonstrate that for those clearly sharp iron peaks whose Raman wavenumber is close to each other such as NO ₃ ⁻ and CO ₃ ²⁻ , their original shape could be restorable by the stripping technique, and that ice’s sharp characteristic peak 3090–3109 cm⁻¹) is steady, while the spectrum band of the complex compound (nCl⁻-[H⁺-OH⁻]_n) chlorine ion combined chemically with water molecule is 3401–3413 cm⁻¹. On the other hand, the research shows that the higher the negative iron concentration, the stronger its Raman characteristic peak intensity and the smaller the ice’s. Based on the number of data and theoretical work, the strong correlation of the molar concentration of negative ion with the S _i/S _{H 2}O band area ratio is built up. Moreover, the developed Raman method is successfully used in the component analysis of the field fluid inclusions from Silurian sandstone in Tarim basin.     

40Ar–39Ar dating of volcanogenic products from the AND-2A core (ANDRILL Southern McMurdo Sound Project, Antarctica): correlations with the Erebus Volcanic Province and implications for the age model of the core

The AND-2A drillcore (Antarctic Drilling Program—ANDRILL) was successfully completed in late 2007 on the Antarctic continental margin (Southern McMurdo Sound, Ross Sea) with the aim of tracking ice proximal to shallow marine environmental fluctuations and to document the 20-Ma evolution of the Erebus Volcanic Province. Lava clasts and tephra layers from the AND-2A drillcore were investigated from a petrographic and stratigraphic point of view and analyzed by the ⁴⁰Ar–³⁹Ar laser technique in order to constrain the age model of the core and to gain information on the style and nature of sediment deposition in the Victoria Land Basin since Early Miocene. Ten out of 17 samples yielded statistically robust ⁴⁰Ar–³⁹Ar ages, indicating that the AND-2A drillcore recovered ≤230 m of Middle Miocene (∼128–358 m below sea floor, ∼11.5–16.0 Ma) and >780 m of Early Miocene (∼358–1093 m below sea floor, ∼16.0–20.1 Ma). Results also highlight a nearly continuous stratigraphic record from at least 358 m below sea floor down hole, characterized by a mean sedimentation rate of ∼19 cm/ka, possible oscillations of no more than a few hundreds of ka and a break within ∼17.5–18.1 Ma. Comparison with available data from volcanic deposits on land, suggests that volcanic rocks within the AND-2A core were supplied from the south, possibly with source areas closer to the drill site for the upper core levels, and from 358 m below sea floor down hole, with the “proto-Mount Morning” as the main source.     

Variability of parameters of the <Emphasis Type="Italic">F</Emphasis>2-layer maximum in the quiet midlatitude ionosphere under low solar activity: 1. Statistical properties

Results of statistical analysis of the properties of variability of F2-layer maximum parameters (critical frequency foF2 and the height hmF2) in quiet midlatitude ionosphere under low solar activity in the daytime (1000–1500 LT) and nighttime (2200–0300 LT) hours are presented on the basis of Irkutsk station data for 2007–2008. It is found that the distribution density of δfoF2 could be presented as consisting of two distinctly different normal laws of this distribution, one of which corresponds to weak (|δfoF2| < 10%) fluctuations in foF2 and the other corresponds to strong (30% > |δfoF2| > 10%) fluctuations. Weak fluctuations in foF2 to a substantial degree are related to ionospheric variability at times less of than 1–3 h and determine the δfoF2 variability in the daytime hours. Strong fluctuations in foF2 are mainly related to day-to-day variability of the ionosphere at a fixed local time, the variability increasing by approximately a factor of 3 during the transition from day to night and determining the δfoF2 variability in the nighttime hours. The distribution density of ΔhmF2 is close to the normal distribution law. An interpretation of the different character of the distribution densities of δfoF2 and ΔhmF2 is given.     

Temporal variations of <Emphasis Type="Italic">δ</Emphasis><Superscript>18</Superscript>O of atmospheric water vapor at Delingha

Oxygen stable isotope of atmospheric water vapor is widely used to study the modern process of climate. Atmospheric water vapor samples were collected at Dlingha, northeast of Tibetan Plateau during the period from July 2005 to February 2006. The variation of δ ¹⁸O and the relationships between δ ¹⁸O and both the temperature and specific humidity are analyzed in this paper. Results show that the seasonal variation of δ ¹⁸O of atmospheric water vapor at Delingha is remarkable with higher δ ¹⁸O in summer and lower δ ¹⁸O in winter. The temporal variation of vapor δ ¹⁸O shows obvious fluctuations, with magnitude of over 37‰ The daily variation of the δ ¹⁸O is highly correlated with air temperature. The relationship between δ ¹⁸O and atmospheric water vapor content is complex. Study shows that δ ¹⁸O of atmospheric water vapor is positively correlated with specific humidity in winter in seasonal scale and inversely correlated with specific humidity in summer rainy period. The δ ¹⁸O values of atmospheric water vapor are lower than those of precipitation at Delingha, and the average difference is 10.7‰ Variations of δ ¹⁸O of atmospheric water vapor is also found to be affected by precipitation events, The model results show that the precipitation effect could have caused the vapor δ ¹⁸O in the raining season to lower by 7% in average in July and August. Supported by the National Natural Science Foundation of China (Grant Nos. 40671043, 40571039 and 40771048) and National Basic Research Program of China (Grant No. 2005CB422002)     

Registration of weak ULF/ELF oscillations of the surface electric field strength

Measurements of the atmospheric electric field strength made by an electrostatic fluxmeter with a unique threshold sensitivity for such devices (6 × 10^–2–10^–3 V m^–1 Hz^–1/2 in the 10^–3–25 Hz frequency range) and wide dynamic (120 dB) and spectral (0–25 Hz) ranges, are presented. The device parameters make it possible to observe the electric component of global electromagnetic Schumann resonances and long-period fluctuations in the atmospheric electric field strength.     

Light climate as the key factor controlling the spring dynamics of phytoplankton in Lake Zürich

During the spring seasons of 1983, 1986 and 1987 the development of phytoplankton in Lake Zürich was investigated (from February to May) using samples taken at short term intervals. The aim was to describe the effects of the short term dynamics of environmental factors on the algal growth. The results could then be used to discuss the existing theories to assess the start of phytoplankton growth pulses in spring. Only 7 to 10 days without wind driven vertical mixing were required in spring to start the first growth pulse, despite of a still very unstable water column (sometimes inverse thermal stratification). Mainly flagellates andStephanodiscus hantzschii increased their biomass and achieved net growth rates of 0.1 and up to 0.65 d⁻¹ respectively. During such a phase the mixing depth was always smaller than the euphotic depth. Later on, at the start of the spring bloom (=last growth pulse in spring before the clear water stage), the intensity of vertical mixing as well as the mixing depth were markedly reduced due to an increase in heat input and low wind. Then flagellates dominated (contribution up to 75.5% of the areal biomass reaching 60 g fresh weight m⁻²) and the growth rate rose to a maximum of 0.65 d⁻¹. Standard models of critical depth considers that there is only a biomass increase if the mixing depth is smaller than the depth of a water layer positive balanced between production and respiration. This model for determining the beginning of a phytoplankton growth pulse in spring takes no account of the favorable light conditions for phytoplankton cells at calm and sunny days in February and March. The newly developed threshold value model takes these situations into account: It assumes that the phytoplankton biomass increases when the calculated effective light climate is equal or greater than a previously fixed threshold. The calculations are based on the mean light intensity within the mixed layer at windy days or within the euphotic depth (z _eu) at calm days. In Lake Zürich a minimum of 0.2 10⁶ J m⁻²d⁻¹ (=0.9 mol quanta m⁻²d⁻¹) has to be reached or surpassed in at least 3 days before an exponential increase of algal biomass can occur. The value does not depend on short term fluctuations in neither radiation nor mixing depth. It seems that this value is rather low comparing with those of investigations in other water bodies (up to 0.8 10⁶ J m⁻² d⁻¹) but high related to values from algal cultures (0.02 10⁶ J m⁻²d⁻¹). As the weather can only be forecasted a few days ahead with any certainty the period for a more or less accurate prediction of an algal bloom is restricted to about 1 to 5 days.     

Small-scale heterogeneities below the Lanzhou CTBTO seismic array,from seismic wave field fluctuations

Seismic wave field fluctuations below Lanzhou Comprehensive Test Ban Treaty Organization seismic array are determined and interpreted to describe the scattering characteristics of the crust beneath the northeastern margin of Qinghai–Tibet plateau in China. The frequency-dependent intensities of the mean and fluctuation wave fields are analyzed for 21 deep (206–632 km depth) teleseismic events. The observed wave field fluctuations in the frequency range 0.5–2.5 Hz can be explained by the scattering of the teleseismic P wave front at random media-type structures. The stable and narrow range of the fluctuation parameters from these teleseismic events indicates evidence for scattered seismic phases which are generated inside the crust. The reasonable structural models contain heterogeneities with 1–3% velocity fluctuations and 2.4–8.2 km correlation lengths in the crust with 53 km thickness.     

Climatic variations in the past 140 ka recorded in core RM, east Qinghai-Xizang Plateau

The sequences of climatic evolution are reconstructed by the analyses of δ¹³C and δ¹⁸O of carbonate from core RM in the Zoige Basin since 140 kaB. P. During the Last Glaciation there existed at least seven warm climatic fluctuations and five cold events correlated with the records of ice core and deep sea, and during the preceding last interglacial period there were two cold climatic variations coinciding with the record of ice core GRIP. These results de-pict climatic instability in east Qinghai-Xizang Plateau over the last interglacial period. In addition, the environmental proxies of the carbonate content and pigments indicate the similar results to the stable isotope record from core RM. Project supported by the Chinese Climbing Project (85-029-02-01).     

The dynamics of a double-cell hydrothermal system in triggering seismicity at Somma-Vesuvius: results from a high-resolution radon survey (revisited)

Data collected at Somma-Vesuvius during the 1998–1999 radon surveys have been revisited and reinterpreted in light of recent geophysical and geochemical information. The duration of selected radon anomalies, together with the decay properties of radon, have been used to estimate the permeability and porosity of rocks of the deep hydrothermal system. The current local cyclic seismicity is explained by means of a double convective-cell model. Convective cells are separated by a low-permeability horizon located at about 2–2.5 km below sea level. Fluids convecting within the upper cells show temperatures ranging 300–350°C. Rock permeabilities in this sector are estimated on the order of 10⁻¹² m², for porosities (ϕ) of about 10⁻⁵ typical of a brittle environment where fluid velocities may reach ∼800 m/day. Fluid temperatures within the lower cells may be as high as 400–450°C, consistent with supercritical regimes. The hydrodynamic parameters for these cells are lower, with permeability k ∼ 10⁻¹⁵ m², and porosity ranging from 10⁻⁶ to 10⁻⁷. Here, fluid motion toward the surface is controlled by the fracture network within a porous medium approaching brittle–ductile behaviour, and fluid velocities may reach ∼1,800 m/day. The low-permeability horizon is a layer where upper and lower convecting cells converge. In this region, fluids (convecting both at upper and lower levels) percolate through the wallrock and release their brines. Due to self-sealing processes, permeability within this horizon reaches critical values to keep the fluid pressure near lithostatic pressure (for k ∼ 10⁻¹⁸ m²). Deep fluid pressure buildups precede the onset of hydrothermally induced earthquakes. Permeability distribution and rock strength do not exclude that the next eruption at Somma-Vesuvius could be preceded by a seismic crisis, eventually leading to a precursory phreatic explosion. The coupling of these mechanisms has the potential of inducing pervasive failure within rocks of the hydrothermal shell, and may be a prelude to a magmatic eruption. It is finally emphasised that the integrated analysis of seismic and geochemical data, including radon emissions, could be successfully used in testing temperature distributions and variations of porosity and permeability in active geothermal reservoirs.     

Carbon budgets of three temperate forest ecosystems in Dongling Mt.,Beijing,China

There is a general agreement that forest ecosystems in the Northern Hemisphere function as signifi-cant sinks for atmospheric CO2; however, their magnitude and distribution remain large uncertainties. In this paper, we report the carbon (C) stock and its change of vegetation, forest floor detritus, and mineral soil, annual net biomass increment and litterfall production, and respiration of vegetation and soils between 1992 to 1994, for three temperate forest ecosystems, birch (Betula platyphylla) forest, oak (Quercus liaotungensis) forest and pine (Pinus tabulaeformis) plantation in Mt. Dongling, Beijing, China. We then evaluate the C budgets of these forest ecosystems. Our results indicated that total C density (organic C per hectare) of these forests ranged from 250 to 300 t C ha-1, of which 35―54 t C ha-1 from vegetation biomass C and 209―244 t C ha-1 from soil organic C (1 m depth, including forest floor detritus). Biomass C of all three forests showed a net increase, with 1.33―3.55 t C ha-1 a-1 during the study period. Litterfall production, vegetation autotrophic respiration, and soil heterotrophic respira-tion were estimated at 1.63―2.34, 2.19―6.93, and 1.81―3.49 t C ha-1 a-1, respectively. Ecosystem gross primary production fluctuated between 5.39 and 12.82 t C ha-1 a-1, about half of which (46%―59%, 3.20―5.89 t C ha-1 a-1) was converted to net primary production. Our results suggested that pine forest fixed C of 4.08 t ha-1 a-1, whereas secondary forests (birch and oak forest) were nearly in balance in CO2 exchange between the atmosphere and ecosystems.     

Sulphur emissions to the stratosphere from explosive volcanic eruptions

 Two methods were used to quantify the flux of volcanic sulphur (as the equivalent mass of SO₂) to the stratosphere over different timescales during the Holocene. A combination of satellite-based measurements of sulphur yields from recent explosive volcanic eruptions with an appropriate rate of explosive volcanism for the past 200 years constrains the medium-term (∼10²years) flux of volcanic sulphur to the stratosphere to be ∼1 Mt a^–1, with lower and upper bounds of 0.3 and 3 Mt a^–1. The short-term (∼10- to 20-year) flux due to small magnitude (10¹⁰–10¹²kg) eruptions is of the order of 0.4 Mt a^–1. At any time the instantaneous levels of sulphur in the stratosphere are dominated by the most recent (0–3 years) volcanic events. The flux calculations do not attempt to address this very short timescale variability. Although there are significant errors associated with the raw sulphur emission data on which this analysis is based, the approach presented is general and may be readily modified as the quantity and quality of the data improve. Data from a Greenland ice core support these conclusions. Integration of the sulphate signals from presumed volcanic sources recorded in the GISP2 core provides a minimum estimate of the 10³–year volcanic SO₂ flux to the stratosphere of 0.5–1 Mt a^–1 over the past 9000 years. The short-term flux calculations do not account for the impact of rare, large events. The ice-core record does not fully account for the contribution from small, frequent events. Received: 27 September 1995 / Accepted: 13 December 1995