V21E-01 INVITED
Total grain size distribution of particles fallen from the atmosphere following explosive ash eruptions
We have collected grain size distribution data (-6 to 12 phi) for tephra-fall from five recent eruptions (Fuego 14 Oct 1974; Mount St Helens 18 May 1980; El Chichón 4 April 1982; Crater Peak/Spurr 18 Aug 1992 and Crater Peak/Spurr 16-17 Sept 1992) where ash was collected immediately after extensive subarial fallout up to several hundreds of kilometers from the source volcano. Total grain size distributions for each eruption were estimated using previously reported deposit distribution maps and the new GSD data, which has improved measurements of fine particles. Results indicate that all ash-fall samples have multimodal size distributions, with fine (<50 microns in diameter) particles dominating in distal regions. Most can be broadly described as having a coarse mode, which could reflect fragmentation by explosive vesiculation, and a fine mode that probably reflects milling in pyroclastic flows or in the vent. Fine particle fallout in these eruptions was greatly enhanced relative to single particle terminal velocities and occurred in <12 hours. The abundance of water in volcanic clouds, satellite remote sensing retrievals of ice and observations of mammatus support hydrometeor involvement in fine ash particle settling. In general, tephra generated in explosive eruptions cannot be characterized by a simple lognormal GSD. Fallout consists of abundant fine particles, which have previously been underestimated because they occur in distal areas and are not well preserved in the geological record. The total GSD data presented here provides a new basis for source parameters in eruption models and also for hazard evaluations since fine particles are important in assessing both aircraft and health hazards.
V21E-02
Aerosol Measurements From Recent Alaskan Volcanic Eruptions: Implications for Volcanic Ash Transport Predictions
Size and time-resolved aerosol compositional measurements conducted during the 2006 Augustine Volcano and 2007 Pavlof Volcano eruptions provide ground-truth information for use in the validation of volcanic ash transport models. These measurements provide quantitative information on the size and concentration of the aerosol, which can be used to test the volcanic aerosol source profiles and transport characteristics used in volcanic ash transport models. Augustine Volcano is on an island in Cook Inlet in southern Alaska. For the 2006 Augustine Volcano eruption, the size and time-resolved aerosol measurements were made using an eight stage (35-5.0, 5.0-2.5, 2.5-1.15, 1.15- 0.75, 0.75-0.56, 0.56-0.34, 0.34-0.26 and 0.26-0.09 microns in aerodynamic diameter) DRUM aerosol impactor deployed in Homer, approximately 120 km northeast of the volcano. Aerosols from the volcano reached the sampler and showed that the size distribution of the volcanic emissions changed during the course of the eruption. For example, crustal elements were present in high concentrations in the largest size fraction (35-5.0 microns) but low concentrations in a smaller size fraction (0.75-0.56 microns) during the phreatomagmatic explosive events. However, during the magmatic emissions period, the concentrations of these elements in the large size fraction decreased, but greatly increased in the smaller size fraction. Pavlof Volcano is a volcano on the Alaska Peninsula in southwestern Alaska. During the 2007 Pavlof Volcano eruption, a network of four DRUM aerosol impactors was deployed downwind of the volcano in an attempt to characterize the change in aerosol size distribution and composition during transport away from the volcano. The samplers were located at Nelson Lagoon, approximately 80 km northeast of the volcano (eight stage DRUM impactor with a top cut point of approximately 12 microns), Sand Point approximately 90 km east of the volcano (three stage DRUM impactor with aerodynamic diameter size fractions of 2.5-1.15, 1.15-0.34 and 0.34-0.01 microns), Kodiak, approximately 550 km east-northeast of the volcano (three stage DRUM impactor), and Homer, approximately 680 km northeast (three stage DRUM impactor). The aerosol samples collected by the DRUM impactors were analyzed with 90 minute resolution for the entire duration of the six-week sample periods. The collected aerosol was analyzed for mass using a beta-gauge and elemental composition (28 selected elements between sodium and lead) using synchrotron x-ray fluorescence. The results from Pavlof Volcano will be compared to those of Augustine Volcano to determine the variability of aerosol emissions between the two eruptions. The Pavlof Volcano aerosol network will also be used to examine changes in size distribution and composition as the aerosols transport away from the volcano. The information gained from these analyses will be compared to volcanic ash transport model predictions to help quantify the limitations of the models and improve future volcanic ash predictions.
V21E-03 INVITED
3-D simulations of eruption column and umbrella cloud development during explosive volcanic eruptions
During an explosive volcanic eruption, a mixture of solid pyroclasts and volcanic gas released from the volcanic vent buoyantly rises as a turbulent plume (i.e., eruption column) and laterally spreads at the neutral buoyancy level as a gravity current (i.e., umbrella cloud). Woods [1988] proposed a steady 1-D plume model which predicts the heights of eruption columns and umbrella clouds and the volume fluxes of the eruption clouds as a function of the mass-discharge rate at the vent. Sparks et al. [1997] proposed a gravity current model which predicts the spreading rate of umbrella clouds for given volume fluxes of the eruption clouds. These models allow us to quickly estimate the conditions at the vent from the observations on the eruption clouds, or, inversely, to predict the behavior of the eruption clouds for given vent conditions. However, these simplified models contain empirical constants (entrainment coefficient of turbulent plume, k, and Froude number of gravity current, Fr) that should be justified. We have developed a 3-D numerical model which simulates the dynamics of eruption clouds, and determined the values of the empirical constants in the simplified models. We apply a pseudo-gas model to describe the injection of a mixture of solid pyroclasts and volcanic gas from a circular vent in a stationary atmosphere; the nonlinear density change of the ejected materials and air with variable mixing ratios is calculated by changing the effective gas constant of the mixture in the equation of state. In order to reproduce the quantitative features of turbulent mixing correctly, a third-order accuracy scheme with fine grid sizes is applied [Suzuki et al., JGR, 2005]. Our model simulates the fundamental features of eruption clouds including eruption columns, pyroclastic flows, co-ignimbrite ash clouds and umbrella clouds, and has quantitatively reproduced the behavior of the eruption cloud in the Pinatubo 1991 eruption (e.g., the total column height, the altitude and the spreading rate of the umbrella cloud for the observed mass discharge rate). Systematic comparison between the 3-D simulations and the simplified models shows that the altitude and the spreading rate of the umbrella cloud in the 3-D simulations are approximated by the 1-D plume model with k=0.1 plus the gravity current model with Fr=0.2 for eruptions in the tropical regions and Fr=0.1 for those in the midlatitude regions. On the other hand, the total column height in the 3-D simulations highly oscillates even for a constant mass discharge rate and its time-average can be substantially greater than the estimate of the column height from the 1-D plume model with k=0.1 particularly for large scale eruption clouds in the midlatitude regions. The knowledge from the 3-D simulations is found to be useful for assessment and refinement of the 1-D plume and gravity current models.
V21E-04
Computational models of overpressured volcanic plumes: Implications for ash dispersion
Although impractical for real-time hazards analysis, high-resolution computational fluid dynamics models provide detailed information about ash dispersion and often demonstrate flow behaviors that are not predicted by faster, parameterized plume models. Here, using computational models, we show that two eruptions with the same mass and heat flows at the vent have drastically different plume dynamics depending on whether the vent pressure is greater than or equal to atmospheric pressure. In the simulations shown here, an increase in vent pressure leads to a much higher gas-thrust region and the formation of a transitional column, which oscillates regularly between buoyant rise and collapse. In most treatments of plumes, however, vent pressure is assumed to be atmospheric or to have no effect on the flow dynamics of the greater column. The simulations shown here and others demonstrate that vent pressure is an important controlling parameter of plume dynamics that may have far reaching effects on ash dispersal.
V21E-05 INVITED
Volcanic Ash Clouds: Modelling for Aviation Safety
Ash clouds of volcanic origin pose a serious safety problem to aviation. Simplistic models often give inaccurate estimates of the dispersion height of these clouds and observational methods have problems if these clouds consist of a mixture of ice, water and ash, as is often the case in moist environments. In this talk results are presented of a sophisticated 3-D numerical model that show the effects of environmental conditions on plume rise and dispersion height for eruptions reaching from small sub-plinian to large co-ignimbrite eruptions. The model not only predicts injection heights, but also the rapid spread of the umbrella, mixing with ice and rain-out or freeze-out of the ash particles and gases. Also sedimentation is simulated realistically. These predictions may be useful for improved remote sensing and detection of hazardous ash plumes and lead to improved initial conditions for simpler dispersion models used at weather services providing ash warnings to the aviation industry.
V21E-06
Significant eruption source parameter(s) for operational volcanic ash cloud transport and dispersion models for Spurr, 1992, eruptions
The monitoring and forecasting of airborne volcanic ash clouds is a necessity for volcano observatories and volcanic ash advisory centres (VAAC) as volcanic ash is a major hazard for local and international aircraft as well as the local population. Ground observations and satellite remote sensing data are not always available in real time. Yet, volcanic ash tracking and dispersion (VATD) models are required to provide simulations/predictions of the ash clouds' movement for up to 18 hours. These VATD models require information on the volcanic eruption, such as eruption start time, plume height, grain size distribution, ash distribution in the plume and eruption rate. During operational real-time responses, these are often not available and therefore default values are used. As more information becomes available, the default parameters are substituted by the known eruption data. International work has been ongoing to determine improved eruption source parameters (ESP) as a function of volcano type, magma type and size of the eruption. These can then be used for VATD simulations before actual eruption data become available. Here, we use the CanERM, HYSPLIT and Puff VATD models, as used within the NOPAC region, for two well studied eruptions at Mount Spurr, Alaska in August and September 1992 to assess the most significant eruption source parameter(s). We show the model simulations using the actual eruption data collected and compare these to variations on this source data to get a sense of how significant the accuracy of each source parameter is on the results. The results from this study will be a first step in understanding which ESP are the most sensitive to changes for the operational VATD predictions.
V21E-07
Plumes and Wind I: Jetstream Interaction and Implications for Air Traffic
Volcanic plumes interact with the wind at all scales. On smaller scales, wind affects local eddy structure; on larger scales, wind shapes the entire plume trajectory. The polar jets or jetstreams are regions of high eastbound winds that span the globe from 30 to 60 degrees in latitude, centered at an altitude of about 10 km. They can be hundreds of kilometers wide, but as little as 1 km in thickness. Core windspeeds are up to 130 m/s. Modern transcontinental and transoceanic air routes are configured to take advantage of the jetstream. Eastbound jets can save both time and fuel by flying within it. Using both an integral model of plume motion that is formulated within a plume-centered coordinate system (BENT) as well as the Active Tracer High-resolution Atmospheric Model (ATHAM), we have calculated plume trajectories and rise heights under different wind conditions. Model plume trajectories compare well with the observed plume trajectory of the Sept 30/Oct 1, 1994, eruption of Kliuchevskoi Volcano, Kamchatka, Russia, for which measured maximum windspeed was 30--40 m/s at about 12 km. Tephra fall patterns for some prehistoric eruptions of Avachinskiy Volcano, Kamchatka, and Inyo Craters, CA, USA, are anomalously elongated and inconsistent with simple models of tephra dispersal in a windfield, but are modeled well by BENT. The numerical experiments suggest that over a wide range of mass eruption rates, from ~ 106 to ~ 108 kg/s, plumes rise between 9 and 11 km. In a still atmosphere, the rise height over this range of mass eruption rate increases from 17 to 33 km. This effect is caused in large part by the enhanced entrainment from the wind, as well as by plume bending. Two potentially useful observations can be made about air routes and volcanic eruption plumes under jetstream conditions. The first is that by taking advantage of the jetstream, aircraft are flying within an airspace that is also preferentially occupied by volcanic eruption clouds and particles. The second is that, because eruptions with highly variable mass eruption rate pump volcanic particles into the jetstream under these conditions, it is difficult to characterize the maximum size, grain size distribution and mass loading that might be present within a downwind volcanic plume or cloud that has interacted with the jetstream. Furthermore, anomalously large particles and high mass loadings could be present within the cloud, if it was in fact formed by an eruption with a high mass eruption rate.
V21E-08
An interdisciplinary effort to identify source parameters for models that predict eruption cloud transport and dispersion for aviation safety
During recent decades, dozens of commercial and military jets have inadvertently flown through volcanic ash clouds downwind from eruptions. These encounters have caused up to tens of millions of dollars (U.S.) in damage to each jet; a few nearly crashed when ceramitized deposits of ingested ash caused engines to fail. In order to avoid such encounters, the International Civil Aviation Organization (ICAO) has established Volcano Ash Advisory Centers (VAACs) throughout the world to detect eruptions using satellite imagery and to notify aircraft. VAACs also predict the paths of ash clouds using atmospheric transport models, using information on volcanic plume height, the mass rate of tephra entering the atmosphere, the vertical distribution of tephra, eruption duration, and grain size distribution as input parameters. In some cases, these "source parameters" must be estimated during an ongoing eruption with few or no observational constraints. In other cases, satellite or ground-based observations obtained during an eruption constrain plume height, umbrella cloud dimensions or growth rate, and other properties, allowing VAACs to refine source parameters, re-run models, and improve predictions. ICAOs International Airways Volcano Watch Operations Group has recommended improving methods of estimating eruption source parameters as a key step towards improving predictions of volcanic ash transport and dispersion. In the spring of 2007, members of the USGS; NOAA; the Universities of Michigan, Alaska, Pisa, and South Florida; the Air Force Weather Agency; the Australian Bureau of Meteorology; and the Canadian Meteorological Centre began an interagency effort to establish constraints on eruption source parameters. This effort has involved (1) assigning default parameters to volcanoes based on magma and eruption type; (2) improving characterization of total grain-size distribution and the size distribution in distal ash clouds; (3) refining empirical and model-based relationships between plume height and mass eruption rate to better constrain the latter from observations of the former; and (4) evaluating the sensitivity of models to source parameters using selected, well-characterized eruptions. Work thus far has found that ash-cloud trajectories can be highly sensitive to atmospheric temperature, humidity, and the three-dimensional wind field in the region, as well as to plume height, eruption rate, and height distribution of tephra. The fallout of fine tephra over a period of days cannot be accurately predicted without incorporating processes such as flocculation and scavenging by rain, which are not considered in current operational ash-cloud prediction models.