Lori S. Glaze

Infrared image analysis of volcanic thermal features: Láscar Volcano, Chile, 1984–1992

Fifteen Landsat Thematic Mapper (TM) images of Lascar volcano (Chile), recorded between December 1984 and April 1992, document the evolution of a lava dome within the summit crater. Four of the scenes were acquired at night. In every image, the two short-wavelength infrared bands, 5 and 7, have detected thermal radiation from the volcano. As a consequence of the Planck distribution function, the relative response of these two channels depends on the proportions of very hot (> 600°C) surfaces occupying tiny pixel areas and broader regions at moderate temperatures (< 280°C). Intercomparison of bands 5 and 7 thereby provides a means for interpreting TM thermal anomalies even in the absence of ground observations. Pronounced changes in the configuration and intensity of the Lascar anomaly suggest that the volcano has experienced at least two cycles of lava dome activity since 1984. The first of these progressed through a “cooling” period, possibly reflecting a reduced flux of magmatic volatiles at the surface, and culminated in an explosive eruption on September 16, 1986, which appears to have completely destroyed the inferred lava dome. The TM data indicate that a new dome had been emplaced by November 1987, more than 15 months before it was first discovered by local observers. Lascars style of cyclical effusive and explosive activity is typical of many volcanoes, and the remote sensing techniques presented herein could be applied elsewhere.

Thermal radiance observations of an active lava flow during the June 1984 eruption of Mount Etna

The thermal budget of an active lava flow observed on 20 June 1984 from the Southeast crater of Mount Etna, Sicily, Italy, was analyzed from data taken by the Landsat Thematic Mapper. The Thematic Mapper images constitute one of the few satellite data sets of sufficient spatial and spectral resolution to allow calibrated measurements on the distribution and intensity of thermal radiation from active lava flows. Using radiance data from two reflective infrared channels, we can estimate the temperature and areas of the hottest parts of the active flow, which correspond to hot (>500 °C) fractures or zones at the flow surface. Using this technique, we estimate that only 10%-20% of the total radiated thermal power output is emitted by hot zones or fractures, which constitute less than 1% of the observed surface area. Generally, it seems that only where hot fractures or zones constitute greater than about 1% of the surface area of the flow will losses from such features significantly reduce internal flow temperatures. Using our radiance observations as boundary conditions for a multicomponent thermal model of flow interior temperature, we infer that, for the parts of this flow subject to analysis, the boundary layer and flow thickness effects dominate over radiant zones in controlling the depression of core temperature.

Transport of atmospheric water vapor by volcanic eruption columns

Contrary to assumptions often made in the literature, explosive volcanic eruptions are capable of transporting significant amounts of water into the stratosphere. In addition to the magmatic water component, atmospheric water vapor is entrained by the column at lower levels. A theoretical model for the conservation of mass, momentum, and thermal energy of four separate components (dry air, water vapor, liquid condensates, and solid particles) is used to determine the extent of atmospheric water redistribution. We examine the effects of water vapor condensation on dynamical characteristics and ambient water vapor transport. A simple technique is presented for deriving canonical forms for the complex system of ordinary differential equations governing the column components. Solutions of this model are presented that show the influence of different volcanic boundary conditions and a range of ambient water vapor distributions on transport of the buoyant column. We show that the water component (vapor + liquid) of small eruption columns rising through a wet atmosphere is dominated by entrained water, whereas larger columns are dominated by the magmatic water. This is due, in part, to the proportionately smaller entrainment surface area in relation to the control volume for the larger columns. We also show that a maintained column with an initial mass flux of 2.7 × 108 kg s−1 erupted into a wet atmosphere would inject 96 Mt of water vapor into the stratosphere over 24 hours, comparable to the annual input from methane oxidation or 100 midlatitude thunderstorms. This increase may accelerate the conversion of simultaneously erupted volcanic SO2 into sulfuric acid.

Sensitivity of buoyant plume heights to ambient atmospheric conditions: Implications for volcanic eruption columns

A theoretical model is developed to investigate the sensitivity of buoyant atmospheric plumes to a wide range of ambient atmospheric conditions, including the temperature gradient, the latitude of the source, and the season. The formulation highlights the compressibility of an ideal gas, internal consistency between the governing equations for the conservation of momentum and energy, and the explicit use of the equation of state. Specific results are presented for water vapor plumes and implications are developed for multicomponent (water vapor, silicate particles, and condensates) volcanic plumes. If plume cooling is due solely to adiabatic expansion and the entrainment and mixing of ambient air, then the atmospheric temperature gradient is shown to be a dominant influence on plume height. Changes in the atmospheric gradient of 10 K/km cause the height of a low-level plume to diifer by a factor of 2. We estimate the magnitude of this effect on volcanic plumes by considering water vapor erupted with equivalent heat fluxes. The sensitivity of plumes to ambient conditions is a result of the small density difference driving buoyancy. The plume density, in turn, is strongly controlled by the thermal energy of the system. Sensitivities associated with the thermal energy balance in the eruption column are also investigated. A modest thermal loss (1–2%/km) from the column by a process other than entrainment can result in a plume height significantly lower than one that cools by entrainment alone. Additional cooling of this magnitude could be caused by a variety of combinations of phenomena, including radiative heat loss and, possibly, the conversion of heat energy into turbulent rotational energy. For particle-laden plumes, there is the possibility of additional heat loss through the fallout of solids from the eruption column. To understand the details of the thermal energy balance in a plume, measurements must be made of the bulk plume temperature profile under known atmospheric conditions.

Ashfall dispersal for the 16 September 1986, eruption of Lascar, Chile, calculated by a turbulent diffusion model

A two dimensional model for the dispersal of volcanic ash, modified from that of Suzuki [1983], is applied to the 16 September 1986, eruption of Lascar, in northern Chile. The result is an isopach map predicting ground concentrations of ash. The map shows two local maxima along distinct dispersal axes and predicts that the deposit was detached from the source. Predicted concentrations are in agreement with a single reported thickness measurement.

Analysis of active volcanoes from the earth observing system

Abstract A study of volcanic activity and its effects on the atmosphere is one of 28 interdisciplinary investigations, for the Earth Observing System (EOS), due to be launched in 1997 and 1999. The volcanology investigation will include long- and short-term monitoring of selected volcanoes, the detection of precursory activity associated with unanticipated eruptions, and the detailed study of on-going eruptions. The data collected will allow us to address two aspects of volcanism: volcanic padforms and the atmospheric effects of eruptions. A variety of instruments on the two NASA EOS platforms, together with supplemental data from the Japanese and European platforms, will enable the study of local- to regional-scale thermal and deformational features of volcanoes, and the chemical and structural features of volcanic eruption plumes and aerosols. This investigation fits well within the overall goal of the EOS Project, which is to study the regional and global interrelationships between components of the Earth System, because it specifically investigates the links between volcanism, atmospheric chemistry and short-term (1–3 year) climate change.

A methodology for constraining lava flow rheologies with MOLA

Topography as measured by the Mars Orbiter Laser Altimeter (MOLA), when supplemented with imaging data, can be used to infer physical emplacement processes in lava flows on Mars with a level of detail analogous to what can be done with unobserved lava flow eruptions on Earth. MOLA, Viking Orbiter and Mars Orbiter Camera (MOC) data are used to develop new inferences regarding the rheology of a typical lava flow near Elysium Mons on Mars. We present a technique that uses MOLA Precision Experiment Data Records (PEDRs) to directly determine the longitudinal thickness profile of lava flows. This technique is preferable to using gridded topography derived from MOLA, particularly for features such as lava flows, with thickness variations at the same scale as their surroundings. Thickness profiles and underlying slope estimates can then be compared with results from rheologic models. The longitudinal thickness profile of the Elysium example discussed here exhibits a concave-up flow surface that is consistent with an exponential viscosity increase. The viscosity shows a relative increase of about 50 times over the length of flow examined when the density of the lava increases as a result of lava degassing.

New statistics for estimating the bulk rheology of active lava flows: Puu Oo examples

Downstream changes in lava rheology due to cooling, crystallization, and vesiculation have a strong influence on the final length and morphology of a lava flow. Three statistics are proposed to estimate the change in lava rheology with distance along the path of an active flow. These statistics correspond to three separate models of the volumetric flow rate dependence on the thickness of the flow. Each statistic is based on flow dimensions and topographic data that are often available from field measurements or remote sensing. One model assumes an elementary laminar Newtonian flow. A second empirical model often used to describe the flow of complex geologic materials, such as lahars, sediment-laden floods, and debris flows, is also investigated for comparison. A new volume-loss model is also proposed to account for stationary components such as levees and stagnant areas. The three statistics derived from these flow rate models are applied and interpreted for two well-defined lobes from episodes 2 and 18 of the 1983-1984 Puu Oo eruption. The power law dependence of the first two models results only in modest differences in the estimates of rheologic change along the flow path. However, the removal of lava from the active flow to construct stationary components produces significant differences in both the magnitude of computed rheologic changes and the ability to discern trends in rheologic changes along the path of the flow.

Stochastic‐ballistic eruption plumes on Io

Some active volcanoes on Io are associated with bright annular deposits. Here we characterize the dimensions of the annulus observed at Prometheus. Assuming that relative brightness in images is directly related to areal particle concentration on the surface, we develop a model describing emplacement of particles whose motion is controlled by stochastic processes near the vent and ballistic transport beyond. Stochastic processes are expressed as probability distributions for the important transport variables. By varying the distribution parameters, high particle concentration annuli on the surface come and go. For isotropic ejection from the stochastic region with a fixed energy, subsequent ballistic transport to the surface produces singularities in the areal concentration at r = 0 and r = rmax. This areal concentration of particles features peaks corresponding to the singularities. Truncation of the ejection cone such that particles with a single energy are ejected isotropically between 0 and some maximum angle θ0 increases the relative importance of the peak near rmax. Extrapolating the model with a narrow Gaussian energy distribution introduces enough dispersion in the areal concentrations to produce broad annuli. Varying combinations of the truncation angle and relative standard deviation for the energy distribution changes the shape and magnitude of the surface deposit. A truncation angle of 75° and a relative standard deviation of 0.08 produce a symmetric annulus closest in shape and size to that observed at Prometheus. From examination of the energetics associated with thermalized particles, we find that many molecular compositions are admissible as annulus constituents at Prometheus.

Influence of volatile loss on thickness and density profiles of active basaltic flow lobes

A bulk density increase due to degassing during emplacement may have a significant influence on the thickness of a lava flow and the rate at which it advances. We present a theoretical model of a lava flow that loses enough volatiles to cause density changes along the path of the flow. We assume that the flow is emplaced as a single, isolated unit and the bulk rheology (e.g., viscosity) is a function of distance from the vent. This type of model is applicable to solitary lobes of basaltic aa and isolated sheets of pahoehoe that advance as a fluid continuum with bulk lava density changes as a function of distance along the flow path. Equations for the flow thickness and the bulk density profiles are derived from mass and volume conservation. Formulas are tabulated for thickness and density profiles for various combinations of flow rates, rheologic changes, and degassing rate functions. We also tabulate formulas for estimating parameters associated with the form and rate of degassing from field data. The Mauna Loa 1984 “1 flow” is a typical example of a flow showing evidence of a bulk density increase and is used to estimate the model parameters. Thickness and density profiles are then computed for a range of plausible lava densities, two different rate functions for the loss of volatiles, and two different models of viscosity change. Results indicate that the thickness profile of a lava flow can be significantly affected when there is a large difference between the density at the vent and at the flow front. For relatively high rates of degassing, the flow profile has a maximum thickness located progressively closer to the vent as the rate of degassing increases. For depth-dependent degassing, an increase in viscosity acts to thicken the flow, which increases the rate of degassing, thus mitigating the thickening influence of the increasing viscosity. Degassing while a flow is active can increase the duration of emplacement by as much as 60%. We find that the flow thickness profiles are sensitive to the choice of flow rate and the initial density, regardless of the form of the degassing function. The nature of the flow rate can significantly affect the shape of the profile as well as the flow front thickness. When the rate of volatile loss depends on the flow thickness, the density profile depends explicitly on the way the rheology changes along the flow path. In all cases, density increases during emplacement counter the tendency of a flow to thicken due to increases in viscosity or resistance to flow with distance. Thus the parameters that define the rate of degassing, and the consequent density change along the path of a flow, emerge as important variables for a quantitative understanding of flow emplacement.