ARSC afiliates participate in the American Geophysical Union's Fall Meeting

The following ARSC-affiliated scientists will be presenting seminars or posters (or both!) at the December meeting of the American Geophysical Union in San Francisco. Click on each name to read selected abstracts of their presentations.

11-15 December, 2006 San Francisco

Posters:

Use of Internet-Based Virtual Globes in the Earth Sciences I Posters

J Bailey, Arctic Region Supercomputing Center; M Nolan, University of Alaska Fairbanks DA: Thursday HR: 1340h

Use of Internet-Based Virtual Globes in the Earth Sciences II Posters

J E Bailey, Arctic Region Supercomputing Center; M Nolan, University of Alaska Fairbanks

SN: Geoscience Applications in Virtual Globes I Posters

J E Bailey, Arctic Region Supercomputing Center J Dehn, Alaska Volcano Observatory; L Blair, U.S. Geological Survey

Geoscience Applications in Virtual Globes II Posters

J E Bailey, Arctic Region Supercomputing Center J Dehn, Alaska Volcano Observatory; L Blair, U.S. Geological Survey

Presentations:

A Distributed Web-based Solution for Ionospheric Model Real-time Management, Monitoring, and Short-term Prediction

Kulchitsky, A, Arctic Region Supercomputing Center, : University of Alaska Fairbanks; Maurits, S, Arctic Region Supercomputing Center, University of Alaska Fairbanks; Watkins, B, University of Alaska Fairbanks, Geophysical Institute
ABSTRACT: With the widespread availability of the Internet today, many people can monitor various scientific research activities. It is important to accommodate this interest providing on-line access to dynamic and illustrative Web-resources, which could demonstrate different aspects of ongoing research. It is especially important to explain and these research activities for high school and undergraduate students, thereby providing more information for making decisions concerning their future studies. Such Web resources are also important to clarify scientific research for the general public, in order to achieve better awareness of research progress in various fields. Particularly rewarding is dissemination of information about ongoing projects within Universities and research centers to their local communities. The benefits of this type of scientific outreach are mutual, since development of Web-based automatic systems is prerequisite for many research projects targeting real-time monitoring and/or modeling of natural conditions. Continuous operation of such systems provide ongoing research opportunities for the statistically massive validation of the models, as well. We have developed a Web-based system to run the University of Alaska Fairbanks Polar Ionospheric Model in real-time. This model makes use of networking and computational resources at the Arctic Region Supercomputing Center. This system was designed to be portable among various operating systems and computational resources. Its components can be installed across different computers, separating Web servers and computational engines. The core of the system is a Real-Time Management module (RMM) written Python, which facilitates interactions of remote input data transfers, the ionospheric model runs, MySQL database filling, and PHP scripts for the Web-page preparations. The RMM downloads current geophysical inputs as soon as they become available at different on-line depositories. This information is processed to provide inputs for the next ionospheic model time step and then stored in a MySQL database as the first part of the time-specific record. The RMM then performs synchronization of the input times with the current model time, prepares a decision on initialization for the next model time step, and monitors its execution. Then, as soon as the model completes computations for the next time step, RMM visualizes the current model output into various short-term (about 1-2 hours) forecasting products and compares prior results with available ionospheric measurements. The RMM places prepared images into the MySQL database, which can be located on a different computer node, and then proceeds to the next time interval continuing the time-loop. The upper-level interface of this real-time system is the a PHP-based Web site (http://www.arsc.edu/SpaceWeather/new). This site provides general information about the Earth polar and adjacent mid-latitude ionosphere, allows for monitoring of the current developments and short-term forecasts, and facilitates access to the comparisons archive stored in the database.

A Solar Wind Short Distance Propagation Model Based on Magneto-Hydrodynamics

Kulchitsky, A, University of Alaska Fairbanks, Arctic Region Supercomputing Center
ABSTRACT: It is essential to know parameters of the Interplanetary Magnetic Field (IMF) and solar wind in the near Earth magnetosphere for research on the effects of the IMF in near-Earth space. Measurements of the IMF taken at the first Lagrange point (L1) by the ACE space satellite, about 230 Earth Radii away, are typically used for estimation of IMF near the Earth. In many geophysical applications, it is assumed that we can use a simple kinematic approach to map the parameters measured at the satellite to the Earth by shifting them in time. The simple formula is to calculate the time delay as the distance from the Earth to the satellite divided by the velocity of the solar wind, projected along the Earth-Sun line. There are two important problems to be addressed with this simple kinematic approach. The first problem is that solar wind is not uniform and particles that hit the satellite often miss the Earth. Typically, the satellite is shifted by 30 or more Earth radii from the Earth-Sun straight line, and the Earth is also shifting from this line during the solar wind propagation time. IMF parameters may differ between the satellite and the Earth-Sun straight line. In this work, we address a second important problem: solar wind and IMF parameters may change during it's travel from the satellite towards the Earth. It is suggested here that a Magneto-Hydrodynamics approach can be used to deal with this problem. A simplified 1D solar wind propagation model was derived for this case using mass and momentum conservation laws and Maxwell equations. The model is simple and fast enough to use even in real-time applications, yet it takes into account real conservation laws of solar wind motion. To validate this model, comparisons of the IMF data and solar wind parameters were performed for the ACE and WIND satellites. Two different events were used. The first was from April 30, 1999, a day on which both satellites were approximately on the same line with the Sun. In that case, shifting differences of solar wind should not be important. The second was October 21, 1998, when both satellites were far away from the Earth-Sun line. Comparisons of these measurements and calculations shows an improvement of IMF calculations compared to the simple kinematic delay method. The suggested method can be effectively used instead of the kinematic delay algorithm in many space physics applications.

Location of Ionospheric F-region Troughs Using a 3D Polar Model, and Validation Using Tomography Data

Watkins, B, University of Alaska Fairbanks, Geophysical Institute and Physics Dept; Kulchitsky, A, University of Alaska Fairbanks, Arctic Region Supercomputing Center; Maurits, S, University of Alaska Fairbanks, Arctic Region Supercomputing Center; Wright, J, Mercer University; Secan, J, North West Research Association
ABSTRACT: A 3D polar ionosphere model has been used with a variety of geographical conditions to determine the extent and location of F-region density troughs in the night-time auroral ionosphere. The results have been validated with tomography data obtained in the Alaskan region. Although the model uses statistical inputs, for quiet to moderately disturbed conditions, the trough locations are generally predicted by the model to be within 2 degrees of latitude compared to tomographic data.

Finite Difference Simulation of Seismic Scattering in Random Media Generated With the Karhunen-Loeve Transform

Thorne, M S, Arctic Region Supercomputing Center, Geophysical Institute, University of Alaska Fairbanks; Myers, S C, Lawrence Livermore National Laboratory; Harris, D B, Lawrence Livermore National Laboratory; Rodgers, A J, Lawrence Livermore National Laboratory
ABSTRACT: The scattering of seismic waves from small spatial variations of material properties (e.g., density and seismic wave velocity) affects all seismic observables including amplitudes and travel-times and also gives rise to seismic coda waves. Analysis of seismic scattering has provided a means to quantify small-scale seismic properties that cannot be determined through travel-time analysis or ray theoretical approaches. Numerical wave propagation techniques, such as Finite Difference (FD) techniques, have been utilized in analyzing the full waveform effects of the scattered wave field, although application of these techniques has been focused on studies in regional distance ranges. In order to simulate scattering in numerical schemes, random heterogeneity is added to a models seismic structure using a method based on the 2- or 3-D Fourier Transform (FT). The FT method is well-suited for introducing random perturbations into models on the regional scale in Cartesian geometries.Yet, numerical techniques for larger scale seismic simulation, e.g., global wave propagation, require computationally parallelized solutions and are generally not parameterized on a Cartesian grid. Both of these factors make use of the FT method problematic. The FT method is also restrictive in that constructing models with variable scale-length heterogeneity introduces a first-order discontinuity into the model space. We develop a new technique of generating models of random heterogeneity for numerical wave propagation by application of the Karhunen-Loève Transform (KLT). In contrast to the FT method, which computes the 2- or 3-D convolution of a correlation function with a set of random numbers to produce a realization of random media, the KLT method determines an orthogonal basis of a theoretical covariance matrix by calculating its eigenvectors and eigenvalues. This orthogonal basis is then used to construct a transform matrix by which the random media can be generated. The KLT is ideal as it allows one to construct the transform matrix with a minimum set of basis vectors. We demonstrate the following advantages of the KLT based method: (1) the technique works for both isotropic and anisotropic correlation structures on Cartesian and non-Cartesian grids, (2) the technique is readily parallelizable, and (3) the technique can be used to generate models with non-stationary correlation structures without introducing first-order discontinuities. Initial set-up of the KLT technique is in general slower than the FT technique; however multiple realizations of the random media may be rapidly generated and kriging interpolation can also be used to further accelerate the set-up. We compute 2-D FD synthetic seismograms for models with random heterogeneity and compare waveforms for models constructed with both the FT and KLT based techniques. We generate models with a change of correlation structure with depth, comparing predictions with first-order discontinuous and smoothly varying correlation structure. We also demonstrate the effects of scattering on the SH-wave field in global simulations using the axi-symmetric FD method, showing how the inclusion of random heterogeneity broadens the pulse width of teleseismic body wave arrivals and delays their peak arrival times. Coda wave energy is also generated which is observed as additional energy after prominent body wave arrivals.

Acceleration of the Arctic Water Cycle: evidence from the Lena Basin, Siberia

Cherry, J E, International Arctic Research Center, Arctic Region Supercomputing Center; Alexeev, V,International Arctic Research Center, P.O. Box 757335, Fairbanks; Liepert, B, Lamont-Doherty Earth Observatory of Columbia University; Groisman, P, National Climate Data Center; Romanovsky, V, Geophysical Institute, University of Alaska Fairbanks,
ABSTRACT: Strong positive climate feedbacks cause much of the Arctic to warm faster than the global average (IPCC, 2001; ACIA, 2004). The global hydrologic cycle is expected to accelerate in a warmer world because a warmer atmosphere can hold more water vapor. However, an important question that has not been adequately addressed is whether thawing of permafrost and deepening of the soil's active layer, which pulls moisture away from the surface into deeper reservoirs, will lead to a wetter or dryer Arctic climate (Vörösmarty et al., 2001). There is an apparent paradox in the Arctic between increasing annual precipitation trends (ACIA, 2004), increasing occurrence of forest fires (Kasischke et al., 1999; Korovin and Zukkert 2003), and the drying of surface lakes (Smith et al., 2005). Our investigation of hydroclimatological change in the Lena basin (Russia) points to an increasingly wet Arctic. Though much of the near-surface air temperature (SAT) warming is occurring when the ground is covered by snow, increases in frozen precipitation are also contributing to warmer soil temperatures by increasing soil insulation. A deeper active layer caused by spring and summer warming holds more soil moisture and is leading to increasing potential evapotranspiration (shown in the model), increasing hydrologic baseflow (modeled and observed), and increasing summer nighttime cloudiness (observed). Changes in summer cloud types are suppressing warming during the days, but warming the nights significantly even during the polar day (Groisman et al. 1996). Earlier onset of snow cover in autumn traps the spring and summer warming, a trend that leads to further deepening of the active layer. These observed and modeled feedbacks describe an Arctic hydroclimatological regime in which water storage and flow has increased and moved from the surface to the subsurface.

Interactive Volcano Studies and Education Using Virtual Globes

Dehn, J, Alaska Volcano Observatory, Geophysical Institute University of Alaska Fairbanks; Bailey, J E, Arctic Region Super Computing Center, University of Alaska Fairbanks; Webley, P,Arctic Region Super Computing Center, University of Alaska Fairbanks
ABSTRACT: Internet-based virtual globe programs such as Google Earth provide a spatial context for visualization of monitoring and geophysical data sets. At the Alaska Volcano Observatory, Google Earth is being used to integrate satellite imagery, modeling of volcanic eruption clouds and seismic data sets to build new monitoring and reporting tools. However, one of the most useful information sources for environmental monitoring is under utilized. Local populations, who have lived near volcanoes for decades are perhaps one of the best gauges for changes in activity. Much of the history of the volcanoes is only recorded through local legend. By utilizing the high level of internet connectivity in Alaska, and the interest of secondary education in environmental science and monitoring, it is proposed to build a network of observation nodes around local schools in Alaska and along the Aleutian Chain. A series of interactive web pages with observations on a volcano's condition, be it glow at night, puffs of ash, discolored snow, earthquakes, sounds, and even current weather conditions can be recorded, and the users will be able to see their reports in near real time. The database will create a KMZ file on the fly for upload into the virtual globe software. Past observations and legends could be entered to help put a volcano's long-term activity in perspective. Beyond the benefit to researchers and emergency managers, students and teachers in the rural areas will be involved in volcano monitoring, and gain an understanding of the processes and hazard mitigation efforts in their community. K-12 students will be exposed to the science, and encouraged to participate in projects at the university. Infrastructure at the university can be used by local teachers to augment their science programs, hopefully encouraging students to continue their education at the university level.

Volcano Monitoring Using Google Earth

Bailey, J E, Arctic Region Supercomputing Center, Alaska Volcano Observatory, Geophysical Institute ; Dehn, J, Alaska Volcano Observatory, Geophysical Institute; Webley, P, Arctic Region Supercomputing Center, Alaska Volcano Observatory/Geophysical Institute; Skoog, R,Alaska Volcano Observatory, Geophysical institute
ABSTRACT: At the Alaska Volcano Observatory (AVO), Google Earth is being used as a visualization tool for operational satellite monitoring of the region's volcanoes. Through the abilities of the Keyhole Markup Language (KML) utilized by Google Earth, different datasets have been integrated into this virtual globe browser. Examples include the ability to browse thermal satellite image overlays with dynamic control, to look for signs of volcanic activity. Webcams can also be viewed interactively through the Google Earth interface to confirm current activity. Other applications include monitoring the location and status of instrumentation; near real-time plotting of earthquake hypocenters; mapping of new volcanic deposits; and animated models of ash plumes within Google Earth, created by a combination of ash dispersion modeling and 3D visualization packages. The globe also provides an ideal interface for displaying near real-time information on detected thermal anomalies or "hotspot"; pixels in satellite images with elevated brightness temperatures relative to the background temperature. The Geophysical Institute at the University of Alaska collects AVHRR (Advanced Very High Resolution Radiometer) and MODIS (Moderate Resolution Imaging Spectroradiometer) through its own receiving station. The automated processing that follows includes application of algorithms that search for hotspots close to volcano location, flagging those that meet certain criteria. Further automated routines generate folders of KML placemarkers, which are linked to Google Earth through the network link function. Downloadable KML files have been created to provide links to various data products for different volcanoes and past eruptions, and to demonstrate examples of the monitoring tools developed. These KML files will be made accessible through a new website that will become publicly available in December 2006.

Three Dimensional Ash Dispersion Modeling within Google Earth : Past Eruptions and Operational Monitoring

Webley, P W, Arctic Region Super Computing Center (ARSC),Alaska Volcano Observatory (AVO)/Geophysical Institute (GI), University of Alaska Fairbanks (UAF) Bailey, J E,AF: Arctic Region Super Computing Center, Alaska Volcano Observatory (AVO)/Geophysical Institute (GI), University of Alaska Fairbanks (UAF); Dean, K, Alaska Volcano Observatory (AVO)/Geophysical Institute (GI), University of Alaska Fairbanks; Dehn, J, Alaska Volcano Observatory (AVO)/Geophysical Institute (GI), University of Alaska Fairbanks
ABSTRACT: Virtual Globes have become widely used for visualization in the scientific environment. They have become a tool for displaying two/three dimensional geophysical data operationally and retrospectively. There are over 100 active volcanoes in the North Pacific (NOPAC) Region which includes those on the Aleutian Islands, Alaska Peninsula, Alaska mainland, and the Kamchatka Peninsula and Kurile Islands, Russia. Volcanic ash is a major operational hazard and is a serious threat to human health and the aviation industry. The Alaska Volcano Observatory (AVO) monitors the volcanoes within the North Pacific (NOPAC) region and uses a volcanic ash dispersion model, Puff, to routinely track the ash clouds from volcanic eruptions. The model uses information such as event duration, size of ash plume and start time (from satellite or seismic data) to predict the movement of the ash cloud released. Puff allows the analyst to track a set number of particles, giving the location in space and time. In the recent past, Puff has been displayed in two dimensional maps of ash location, color coded by altitude and relative ash concentration. This is a useful tool for operational analysis but does not take full advantage of the three dimensional nature of the data. A virtual globe such as Google Earth allows the analyst to display markers at known locations. Given the three dimensionality of the Puff model, Google Earth becomes a tool to display these predictions of ash dispersion in various formats. Puff is a global ash dispersion model and the predicted ash cloud can be displayed quickly and automatically for any volcano. Here we show operational Puff predictions of the volcanic ash in three dimensions, both as iso- surfaces and particles, and study past eruptions to illustrate the capabilites that the Virtual Globes can provide.

Satellite Based Extrusion Rates for the 2006 Augustine Eruption

Dehn, J, Alaska Volcano Obsevatory, Geophysical Institute University of Alaska Fairbanks; Bailey, J E, Arctic Region Super Computing Center, University of Alaska Fairbanks; Dean, K G, Alaska Volcano Obsevatory, Geophysical Institute University of Alaska Fairbanks; Skoog, R,Alaska Volcano Obsevatory, Geophysical Institute University of Alaska Fairbanks; Valcic, L,Alaska Volcano Obsevatory, Geophysical Institute University of Alaska Fairbanks
ABSTRACT: : Extrusion rates were calculated from polar orbiting infrared satellite data for the 2006 eruption of Augustine Volcano, Alaska. The pixel integrated brightness temperatures from the satellite data were converted to estimates of ground temperature by making assumptions and using first hand observations about the geometry of the hot area (lava dome, flows and pyroclastic flow deposits) relative to the cold area in the kilometer scale pixels. Extrusion rate is calculated by assuming that at a given temperature, a lava emits an amount of radiation proportional to its volume. On ten occasions during the activity, helicopter based infrared imagers were used to validate the satellite observations. The pre-January 11 thermal activity was not significantly above background in satellite data. The first strong thermal anomalies were recorded during the first explosive phase on January 11. During successive explosive phases in January, bright thermal signals were observed, often saturating the sensors. Large areas (many km2) were observed to be warm in the satellite data, indicative of pyroclastic flows. Sometime during or after January 29, during a phase of sustained ash emission, the thermal signal became persistent, suggesting the beginning of lava effusion. The extrusion rates derived from satellite data varied from 0 to nearly 7 m3/s, giving an eruption rate of 2.7 m3/s. The extrusion event produced two blocky lava flows which moved down the north flank of the volcano. Extrusion occurred through at least March 15 (day 76) when a sharp drop in extrusion rate and thermal signal is observed. Based on the derived extrusion rates, it is estimated that 18 million m3 of lava was extruded during the course of the eruption. This value agreed well with photogrammetric measurements, but does not agree with volumes derived through subtraction of digital elevation models post- and pre- eruption. It should be noted that the thermal approach only works for hot lavas, and does not include pyroclastic deposits or ashfall, which are included in the DEM subtraction approach. However the pyroclastics should only account for a small amount of the extruded volume. In spite of its limitations, satellite based extrusion modeling provides a reasonable and safe method to monitor volcanoes and detect change in eruption style in near real time.

Visualizing Scientific Data Using Keyhole Markup Language (KML)

Valcic, L,Alaska Volcano Observatory, Geophysical Institute Bailey, J E, Alaska Volcano Observatory, Geophysical Institute,Arctic Region Supercomputing Center; Dehn, J Alaska Volcano Observatory, Geophysical Institute
ABSTRACT: Over the last five years there has been a proliferation in the development of virtual globe programs. Programs such as Google Earth, NASA World Wind, SkylineGlobe, Geofusion and ArcGIS Explorer each have their own strengths and weaknesses, and whether a market will remain for all tools will be determined by user application. This market is currently led by Google Earth, the release of which on 28 Jun 2005 helped spark a revolution in virtual globe technology, by bringing it into the public view and imagination. Many would argue that such a revolution was due, but it was certainly aided by the world-wide name recognition of Google, and the creation of a user-friendly interface. Google Earth is an updated version of a program originally called Earth Viewer, which was developed by Keyhole Inc. It was renamed after Google purchased Keyhole and their technology in 2001. In order to manage the geospatial data within these viewers, the developers created a new XML-based (Extensible Markup Language) called Keyhole Markup Language (KML). Through manipulation of KML scientists are finding increasingly creative and more visually appealing methods to display and manipulate their data. A measure of the success of Google Earth and KML is demonstrated by the fact that other virtual globes are now including various levels of KML compatibility. This presentation will display examples of how KML has been applied to scientific data. It will offer a forum for questions pertaining to how KML can be applied to a user's dataset. Interested parties are encouraged to bring examples of projects under development or being planned.

Monitoring Volcanic Eruptions Using Satellite Data in the North Pacific Region

Dean, K G, Alaska Volcano Observatory, Geophysical Institute, University of Alaska Fairbanks; Dehn, J, Alaska Volcano Observatory, Geophysical Institute, University of Alaska Fairbanks; Bailey, J E, Arctic Region Supercomputing Center and Alaska Volcano Observatory, University of Alaska Fairbanks; Horwood, C,Praxis Publishing Ltd, United Kingdom
ABSTRACT: A book entitled, "Volcanic Monitoring Volcanic Eruptions Using Satellite Data in the North Pacific Region", will be published next year by Springer/Praxis Press. The book will discuss how satellite data are used to monitor volcanoes and provide a consolidated but informative review of types and styles of eruptions and landforms as seen from space. Data from GOES, AVHRR, MODIS, Landsat, ASTER, and various SAR systems will be included. The book will consist of two parts. The first part will describe the state of the art of real-time volcano monitoring from space. The second part will be an atlas of images showing eruptions and volcanoes with a description of the activity or features observed. The atlas will include images from volcanoes around the world but will concentrate on the North Pacific (NOPAC) Region due to the variety of eruptive activity, frequency of eruptions, and the volume of data collected by the Alaska Volcano Observatory (AVO). A DVD with images and animations to show the dynamics of eruptions will be included with the book, and a web site will provide updated images of eruptions. Google Earth, as the most pervasive virtual globe software, can also be used to browse the atlas, giving a global spatial context to the data. A downloadable "KMZ" file will be shown that displays the atlas entries, organized by sensor, eruptive activity (thermal anomalies, plumes) and now through the use of this software, geographic location.

Using Google Maps to Access USGS Volcano Hazards Information

Venezky, D Y, U.S. Geological Survey; Snedigar, S, Alaska Volcano Observatory; Guffanti, M, U.S. Geological Survey; Bailey, J E, Arctic Region Supercomputing Center; Wall, B G, U.S. Geological Survey,
ABSTRACT: The U.S. Geological Survey (USGS) Volcano Hazard Program (VHP) is revising the information architecture of our website to provide data within a geospatial context for emergency managers, educators, landowners in volcanic areas, researchers, and the general public. Using a map-based interface for displaying hazard information provides a synoptic view of volcanic activity along with the ability to quickly ascertain where hazards are in relation to major population and infrastructure centers. At the same time, the map interface provides a gateway for educators and the public to find information about volcanoes in their geographic context. A plethora of data visualization solutions are available that are flexible, customizable, and can be run on individual websites. We are currently using a Google map interface because it can be accessed immediately from a website (a downloadable viewer is not required), and it provides simple features for moving around and zooming within the large map area that encompasses U.S. volcanism. A text interface will also be available. The new VHP website will serve as a portal to information for each volcano the USGS monitors with icons for alert levels and aviation color codes. When a volcano is clicked, a window will provide additional information including links to maps, images, and real-time data, thereby connecting information from individual observatories, the Smithsonian Institution, and our partner universities. In addition to the VHP home page, many observatories and partners have detailed graphical interfaces to data and images that include the activity pages for the Alaska Volcano Observatory, the Smithsonian Google Earth files, and Yellowstone Volcano Observatory pictures and data. Users with varied requests such as raw data, scientific papers, images, or brief overviews expect to be able to quickly access information for their specialized needs. Over the next few years we will be gathering, cleansing, reorganizing, and posting data in multiple formats to meet these needs.

The Use of High Resolution NWP data for Dispersion Modeling of Airborne Volcanic Ash and Tephra Fallout

Morton, D, Department of Computer Science, The University of Montana, Arctic Region Super Computing Center (ARSC); Webley, P W, Arctic Region Super Computing Center, Alaska Volcano Observatory (AVO)/Geophysical Institute (GI), University of Alaska Fairbanks (UAF); Dean, K, Alaska Volcano Observatory (AVO)/Geophysical Institute (GI), University of Alaska Fairbanks (UAF); Peterson, R, Department of Mechanical Engineering, University of Alaska Fairbanks
ABSTRACT: Ash dispersion models are routinely used to predict the movement of ash clouds from volcanic eruptions. The Puff dispersion model is used by the Alaska Volcano Observatory (AVO) in the North Pacific region as both an operational tool and for retrospective analysis of past events. The model requires some basic information to initialize a prediction including location of volcano, estimated plume height, particle size, initial distribution and the wind field to be used. Puff tracks the movement of a set number of hypothetical particles thereby predicting the transport of volcanic ash, and the location and relative amount of ash fallout. In the recent past, global, mesoscale and regional meteorological forecast models have been used as initialization for the Puff wind fields. These include the North American Mesoscale Model (NAM) and the Global Forecast System (GFS) model, at horizontal resolutions ranging from a few 10's to many 10's km. To use Puff as a tool for predicting ash/tephra fallout requires a much higher spatial resolution to resolve the low level wind patterns. Recently, high spatial resolution meteorological forecasts have been made possible using the Weather Research and Forecasting (WRF) model and incorporating its forecasts as initialization data for Puff. WRF can provide forecasts from sub km to 10's of km resolution and using nested grids can provide data at a several resolutions. WRF can be initialized using large scale operational models or re-analysis data for past events. Here we will show where WRF has been used as the initialization model for Puff at four eruptions within the NOPAC region (Mt. St Helens [1980], Mt. Spurr [1992], Mt. Redoubt [1989/90] and Mt. Augustine [2006]). In addition, the IAVCEI working group on modeling tephra fall hazards has outlined five eruptions (including Mt. St Helens) to study and we include those here as well, (El Chichon [1982], Cerro Negro [1995], Soufriere Hills [1997] and Mt. Etna [1998]). For this study, the Arctic Region Supercomputing Center (ARSC) is used to run the WRF model for custom regions around each of the selected volcanoes. We compare the high-resolution WRF-Puff predictions to those that are driven by low-resolution regional models to determine the advantages that the high resolution model can bring. In addition, we compare these model predictions of ashfall to isopachs data for each of these eruptions to assess accuracy.

Temperature-Dependent Pore Space of Sea Ice: X-ray Computed-Tomography and Dual Model Network Analysis

Pringle, D,Arctic Region Supercomputing Center, Geophysical Institute University Alaska Miner, J, Geophysical Institute University Alaska Fairbanks; Glantz, R, Department of Geography and Environmental Engineering, Johns Hopkins University; Hilpert, M, Department of Geography and Environmental Engineering, Johns Hopkins University; Eicken, H, Geophysical Institute University Alaska Fairbanks
ABSTRACT: Sea ice provides a mechanical and thermal barrier between the atmosphere and ocean at high latitudes and is an important component of the climate system. It is a composite material with temperature- and salinity- dependent volume fractions of ice, brine and air. Our present interest is the microstructural control of the (hydraulic) permeability, which affects the melt-season albedo (shortwave reflectance) of sea ice by enabling or restricting the drainage of surface meltwater, and is also relevant for other processes including nutrient delivery to microorganisms colonizing the lower layers of the ice sheet. At low temperatures, the relative brine volume (porosity) is small, the inclusions are disconnected, and the permeability is small. The relative brine volume increases with temperature, and the permeability increases strongly above a porosity of ~5%. We have imaged the pore space of natural and laboratory-grown sea ice with x-ray micro-computed tomography (CT) over a temperature range typical of that in the Arctic ( - 25°C to -3°C) and which spans corresponding changes in the permeability by several orders of magnitude. Based on our `Dual Model' approach, we identify pore bodies and throats from the segmented CT data to eventually derive a pore network. We characterize the temperature-dependence of the pore space connectivity and apply critical path analysis to investigate the dependence of permeability on pore microstructural evolution.

Arctic Region Supercomputing Center PO Box 756020, Fairbanks, AK 99775 | voice: 907-450-8600 | email: info@arsc.edu

Back to Top