V41B-01
Viscosity of Magmatic liquids: a model.
Here we present a multicomponent chemical model for predicting the viscosity of naturally-occurring, volatile- bearing silicate melts based on ~1542 experimental measurements of viscosity at T(K) on melts of known composition. Much, but not all, of this data stems from the work of our own laboratories in the past 20 years. The model incorporates 10 major and minor oxide components and the volatile components that are known to have a major influence at geological concentrations on the viscosity of melts; H2O and F. We use the oxide mole % as a chemical basis. We treat all iron as FeOTot. There is clear evidence that, in detail, that Fe2O3 and FeO play distinct roles in governing melt structure and that the redox state of very Fe-rich melts does measurably influence viscosity. Our redox simplification here is purely empirical, resulting from the fact that tests for model improvement, based on a redox factor, fail to produce an improvement of the fit. Fluorine is treated as mole % F2O-1. The model has the following attributes: i) its experimental basis spans virtually the entire compositional range found in naturally-occurring volcanic rocks, ii) the chemical model captures the effects of 10 major and minor oxide components and the volatile components H2O and F, iii) it is computationally continuous across the entire compositional and temperature spectrum of the database, iv) it is capable of accommodating both strong (near-Arrhenian T-dependence) and fragile (very non-Arrhenian T-dependence) behaviour of silicate melts, and v) it reproduces all empirically observed relationships between melt composition and transport properties such as glass transition temperature (Tg) and fragility (m). The model uses a total of 18 empirical coefficients.
V41B-02
Characteristics of the Upward Flow of Magmatic Volatiles in the Volcano Conduit Prior to an Explosion Derived from the Broad-band Seismic Records
A typical broad-band seismic record of Vulcanian explosion consists of two impulses, of low-frequency (LF) and high-frequency (HF) contents. We propose a conceptual model that considers the LF seismic signal as being generated by the vibration of the magma conduit during the upward flow of magmatic volatiles to the magma-free surface, while the HF seismic signal is generated by the explosion at the magma-free surface. We define the time D, representing the movement of magmatic volatiles, as the time from the beginning of LF impulse to the beginning of HF impulse and use the Fourier spectrum of the HF impulse to estimate the energy of explosion. Observations at two andesitic Mexican volcanoes: Volcan de Colima and Popocatepetl, indicated three depth levels for the beginning of the vertical movement of magmatic volatiles in the conduit at different stages of the eruption. The D values observed before significant Vulcanian explosions at both Mexican volcanoes are close, indicating similar depths of the beginning of vertical movement of magmatic volatiles. It is shown that there is a dependence of the energy of the explosions upon the duration of D.
V41B-03
Magma Ascent and Degassing at Volcán de Colima, Mexico: Monitoring the Increasing Activity
Volcán de Colima presents a great opportunity for the study of conduit and plume processes. Since 2001, the volcano has been in almost constant eruption with daily explosive activity and three significant effusive periods occurring since 1998. It is clear that small variations in certain factors, such as the magma ascent rate, temperature or volatile-contents can promote a change in eruption style. During 2005, various rapid switches were observed, with at times simultaneous occurrence of explosions and dome growth at different vents within the crater. The sequence of explosions included the largest observed since the last Plinian event in 1913. The monitoring strategy at Volcán de Colima is being revised as a response to the increase in activity during the last few years, which possibly signifies the start of a pre-Plinian phase. Traditional systems are being supplemented with thermal and infrasound monitoring, which are greatly increasing the data available to study the evolution of activity and facilitate hazard assessment. To highlight the critical parameters, models are required of the magma ascent and degassing processes. Vulcanian explosions from 2005 were characterized by variable ash-content, column ascent velocity and gas emission. The processes are being studied, integrating a variety of data sets. Following an explosion, rapid sealing of the system was often followed by brittle fracturing within the conduit. During subsequent pressurization, seismic swarms were detected as precursory signals to the following explosion. Increased fumarole temperatures have been measured remotely prior to this type of larger magnitude event some minutes or hours prior to their onset. In addition to the precursors, the relationship between different parameters is being studied, including a statistical study of the explosivity, temperature and ascent velocity of the column, and partitioning of acoustic and seismic energy. Resulting models will have an impact for the future monitoring strategy.
V41B-04
Infra-red Monitoring of Fumaroles at Volcán de Colima, México.
Infra-red cameras permit the remote measurement of temperatures in locations too dangerous to allow direct
measurement, which is currently in a state of unrest, producing multiple explosive events each day. Since January
2006, the temperatures of three fumarole areas upon the crater rim of Volcán de Colima have been measured
regularly from a distance of 5.8 km with a VarioCAM infra-red camera (wavelength 8-14 μm). Simple
mathematical models demonstrate that despite the large viewing distance, and consequent large pixel size and
reduced atmospheric transmissivity, changes in thermal output by the fumaroles can still be detected.
Sequences of thermal images, with images taken at five second intervals, were collected on an approximately
weekly basis, with the apparent temperatures of six regions of interest measured in each image. The resulting
graphs show changes in fumarole apparent temperatures during each visit. Trends within individual fumarole
temperature sequences are independent of the explosions. The data are then filtered to select only points from
nighttime, cloud-free images and mean temperatures for each of the regions of interest are calculated.
Comparison of daytime and nighttime fumarole and flank temperatures allows crude estimation of subpixel
temperatures. Between January and September 2006, the apparent pixel-averaged temperatures of the hottest
fumarole ranged from 30-40 °C, with maxima in mid-March and late-July. These trends are independent of
flank temperature. Between October and December apparent temperatures were lower (25-30 °C), then in
January 2007 they showed a sharp decline <15 °C. The trends remain when atmospheric corrections
are applied to the data. There is evidence that dome growth began within the crater at the beginning of 2007 and
the sharp decline in fumarole temperature may be related to associated changes in the plumbing of the upper
edifice.
http:www.ucol.mx/ciiv
V41B-05 INVITED
Daily Monitoring of Ecuadorian Volcanic Degassing From Space
Daily measurements of volcanic sulfur dioxide (SO2) emissions in Ecuador and southern Colombia since September 2004, derived from the Ozone Monitoring Instrument (OMI) on NASA's Aura satellite, are presented. We use automated processing to extract daily SO2 burdens and information on sources from the OMI datastream. Monthly average SO2 vertical columns are also used to illustrate variations in SO2 loading and to pinpoint degassing volcanoes. The dense concentration of active volcanoes in the northern Andes provides a test of OMI's ability to distinguish SO2 from multiple sources. Our analysis reveals that Tungurahua, Reventador and Galeras were responsible for the bulk of regional SO2 emissions in the timeframe of our study, with no significant SO2 discharge detected from Sangay. At Galeras and Reventador, we conclude that OMI can detect variations in SO2 release related to cycles of conduit sealing and degassing, which are a critical factor in hazard assessment. OMI SO2 data for Reventador are the longest sequence of degassing measurements yet available for this remote volcano. Comparisons between the Reventador SO2 data and coincident seismic data will be discussed. At Tungurahua, we observe a good qualitative agreement between OMI-derived SO2 burdens and coincident ground-based SO2 flux measurements, and note increasing emissions in the months prior to large explosive eruptions of the volcano in July and August 2006. Cumulative SO2 loadings yield a total of ~1.16 Tg SO2 emitted by the three source volcanoes between September 2004 and September 2006; predominantly via non-eruptive degassing. These measurements confirm OMI's potential as an effective, economical and risk-free tool for daily monitoring of SO2 emissions from hazardous volcanoes.
V41B-06
Seismoacoustic signatures of fluid oscillations in the volcanic plumbing system and eruption column
Portable arrays of broadband infrasound sensors collocated with broadband seismometers near Mount St. Helens (MSH), USA, and Tungurahua, Ecuador, have recorded many signals attributed to subsurface fluid oscillations, and strong signals associated with violent degassing. At MSH, long period (LP) seismic events modeled by a pressure transient in a subsurface resonant steam-filled crack, intermittently generate impulsive infrasonic pressure signals, while two dominantly phreatic eruptions were accompanied by strong infrasonic jet noises, and notably different seismicity. Tungurahua has produced countless explosions, LPs, gliding harmonic tremor, and jet noise lasting for several days at a time, indicative of alternating styles of degassing. We summarize the observations and then present preliminary numerical investigations of the coupling mechanism between seismic and acoustic LP events at MSH using a 3D finite difference representation of the elastodynamic and acoustic wave equations, including the effects of topography and wind. We also investigate the hypothesis that infrasonic signals observed during moderate to large eruptions at both volcanoes are generated by the same physical mechanisms underlying the generation of sonic jet noise, such as from the jet engines of flight vehicles.
V41B-07
Eruption dynamics at the active Santiaguito Dome inferred from a multidisciplinary geophysical experiment
Santiaguito Dome, which lies in the shadow of its parent stratovolcano (Santa Maria, Guatemala), has been
effusing and exploding for more than eight decades. Pyroclastic emissions from the currently active Caliente Vent
are characterized by ash and ballistic-rich emissions at least every hour, with vapor plumes that reach more than
a kilometer above the vent. Such activity is recently accompanied by slow effusion of a dacitic block lava flow,
which overtops the shallow 200-m-diameter crater and flows down the flanks of the cone. We are studying the
mechanics of this explosive activity with an arsenal of geophysical sensors because of the unparalleled view into
the crater from the Santa Maria summit vantage 1200 m above.
Data collected in January 2007 provides us with the opportunity to quantify the energy budget for small-magnitude
pyroclastic eruptions involving a silica-rich (dacitic) magma. Specifically, we are integrating seismic, infrasonic,
thermal imaging, gas flux (SO2) imaging, high resolution video, and doppler radar observations to assess
the temporal chronology of an eruption beginning with pre-eruption seismicity (apparent a few seconds to tens of
seconds before the eruption onset), followed by subsequent infrasound radiation, and multi-phase material flux
out through the vent. We are integrating these diverse data streams with the goal to better understand the
location of various elastic energy sources and to provide more insight into the complex geometry of the crater and
conduit system. Integrated measurements are used to assess the kinetic energy and thermal flux with an end
goal of using elastic wavefield studies to remotely quantify eruption intensity at this type of potentially hazardous
volcano.
http:earth.unh.edu/johnson/SANTIAGUITO/santiaguito.htm
V41B-08
Analysis of Seismicity Related to the December 2002 - February 2003 Eruptions at Popocatepetl Volcano, Mexico
Since its reactivation in 1994 Popocatepetl Volcano has undergone several activity stages, some of which comprise cycles of intrusion, extrusion and destruction of a lava dome. Popocatepetl has produced up to several tens of daily long-period (LP) events and several volcanic tremor episodes, along with a few volcano-tectonic earthquakes. A lava dome was extruded on May 2002 and subsequently destroyed by 19 explosions from December 2002 to February 2003. The seismicity accompanying the destruction of the lava dome was dominated by long-period (LP) events and a mix of harmonic and spasmodic tremor. We analysed the seismity related to this explosive phase and classified the majority of LP events into three families according to their frequency content. Most of the LP events are characterized by a dominant frequency near 2 Hz throughout the eruptive period. Many events also exhibited a low-amplitude short-period (SP) phase ocurring a few seconds before the LP signal. The SP signal was not observed without the LP event, although several events occurred in the absence of the precursory signal. We also found a notably decrease of LP events activity prior to December 18, 2002 and February 14 to 28, 2003 explosive episodes. Three different types of tremor lasting from a few minutes to several hours were also observed: spasmodic, harmonic and a very low-amplitude pulse-like type of tremor, the latter occurring on November 11, 12 and 30 with a duration of about 120 minutes each. Harmonic tremor episodes lasted from a few minutes to a few hours and there was no tremor occurrence days before the explosive events. Tremor spectra display a fundamental frequency which can vary in the range 0.6-1.3 Hz within a one-hour-long series of tremor. Similar observations are obtained for the peak spacing of the harmonic tremor spectra. These variations may reflect short-term fluctuations of the acoustic properties of the source while no long-term evolutions of the LP sources are detected prior the dome destruction. The seismic activity at Popocatepetl Volcano from December 2002 to February 2003 differs from that observed prior to the December 2000 dome destruction. In contrast, a similar behavior was observed prior to the eruptive activity of November and December, 1998 and that of January 2001. This analysis could be useful to determine and/or understand a the variability of the volcano behavior prior to intense activity and therefore, to forecast more accurately future possible eruptive scenarios.