Christopher Pett BS Environmental Health Richard Stockton College Pomona NJ
Volcanic explosive events > VEI 3 (explosivity index) near the Arctic Circle if occurring during late fall and winter months (November through March) may lead to cyclogenesis. The mid-latitude Jetstream shifts further south and the TTL (top of troposphere level) is lower during winter months. A Vulcanian /Plinian plume (10 -25km) reaching into the Stratosphere is likely to cause a diversion of the mid-latitude Jetstream and a meridional flow of the Rossby waves. A specific series of events is hereby studied to demonstrate how a Vulcanian/Plinian eruption near the Arctic Circle may lead to cyclogenesis.
Cyclogenesis is a process by which warm and cold air masses meet and a cyclonic circulation develops in the atmosphere. Warm advection in lower atmosphere produces height rises amplifying the ridge aloft. Only temperature controls the thickness of the atmospheric layers as defined by the thermal wind relationship. The pressure gradient force is increased due to the increase of pressure differential between warm and cold air masses. The mid-latitude Jetstream is located above the polar front created by the uplift of warm air in the south beside the cold air in the north. A volcanic eruption of significant magnitude and duration such as an explosive event of VEI 3 or above will form a volcanic plume reaching into the stratosphere and disrupting the flow of the Jetstream. Such an event will produce a volcanic plume with initial upward thrust and as it rises the plume will rotate in a counterclockwise rotation due to wind shear and Coriolis force. The predominant wind will be eastward above the polar front due to high pressure warm winds in the south moving to lower pressure colder air to the north thus the spin of the earth will turn the winds to the right. The energy of the volcanic plume flowing upward will be directed into a cyclonic spin as it rises thus kinetic rotational energy. As the plume rises it gains a tremendous amount of mass and volume as it rises to the top of its column reaching into the stratosphere. This increase of mass and area of the plume equals the inertia of the plume at its height and radius. The vorticity of the plume equals the square of the tangential velocity divided by the radius. Torque, the force applied in rotating the plume around its axis, is the product of the inertia and the vorticity. Thus a volcanic explosive event produces a mesocyclone. This mesocyclone imparts its relative vorticity to the Jetstream and a positive vorticity advection ahead of the trough in the wave pattern. Convergence at the surface leads to divergence aloft at the point of positive vorticity advection. The pattern of the Jetstream goes from zonal to meridional flow. As a group the Rossby waves follow this pattern induced by the volcanic mesocyclone and the volcanic signal is amplified downstream with cyclogenesis on a larger scale forming.
In support of the mechanism of volcanic cyclogenesis, the following data is given for a specific volcanic eruption event. On January 11, 2006 the Augustine volcano located in the Cook Inlet of Alaska entered its explosive phase of eruption sending a column of ash and steam upwards 6.5km at 4:44am and another 10.2km into the stratosphere at 5:12am. Wallace et al. These two events are the preliminary round continuing through January 28th of thirteen events culminating in a cumulative erupted volume of 149 million cubic yards (magma) and inflated eruptive volume of 58 million cubic yards (tephra & pyroclastic flows). Coombs et al. Volume of tephra fall alone was 28 million cubic yards for all 13 events. Mass of tephra fall was based on mass eruption rate for event 9 on January 17th: 6.9 x10⁶kg/s. The eruption of event 9 sent a column of ash and steam 13.5km into the stratosphere on the morning of January 17th at 7:58 am. Schneider, D. J et al.. A volcanic explosivity index (VEI) of 3 was assigned by USGS scientists Wallace et al. based on plume height and tephra fall. The cumulative erupted volume is equivalent to 114 million cubic meters thus this eruption may be classified as Vulcanian/ Plinian. USGS Volcanic Explosivity Index. Newhall C.G., Self S.
STP Eruptive mass
1x10e9 kg Volume
1x 10e6 m^3 % mass density
Ash 1.73 1.73 95.3
DRE 0.665 2600
water 0.0813 0.081 4.7 1000
1 x 10e3 kg Volume m^3 Gas%mass
SO2 2.8 966 65 2.926
CO2 1.5 746 35 1.977
total 4.3 1712
1200K Volume m^3
subtotal 7532 >1%
8.1e7 kg 4.52E+08 0.18
total gas 4.52E+08
total volume Mass x density
1.8e9 kg 2.52e9m^3 1.4
mass Cp energy temp kJ Joules energy time power
kg kj/kgK kJ/K K joules seconds
steam 8.10E7 9.8 7.94E8 930 7.38E+11 7.38E14 738 TJ 250 3 TW
ash 1.73E9 1.62 2.80E9 930 2.60E+12 2.6 E15 2.6 TJ 10.4 TW
total 3.3 PJ 13.4 TW
Joules 3.3 PJ 13.5 TW
Watt 13.4 TW
Data was provided by McGee et al. Author performed calculations. It was found that water (steam) made up 95% mass of plume and nearly 100% of volume. Augustine volcano is an island in Cook Inlet thus groundwater supplies the water for steam and powers eruption into the atmosphere by its expansion and later condensation. Initial power of eruption at crater was calculated to be 13.4 TW, within 1% of Plumerise calculations.
PlumeRise, the Bristol model of volcanic plumes in a wind field was used to define event 9 of the 2006 Augustine eruption. Woodhouse et al. The momentum flux along the centerline trajectory in the plume (UV) of the 2006 Augustine eruption reached a maximum of 1.9e3kg /ms² at 9313 meters on Jan 17th. See Appendix: Data set 1: plumedata1. For explanation of variables of Plume data, see Appendix: Data set 2: Raw data from Plume. All calculation thereof are estimations from Plumerise model. The volume of entrained air in the plume was estimated 1.6e11m³at tophat (top of plume) with a power of 115 e12 Joules per second. The total power generated 225 meters (1477m asl) above the crater is 15 TW equivalent to the power generated by 5 “average Atlantic” hurricanes (3 TW). At 3148 meters asl., 1896m above the crater 30TW power generated is equivalent to a “Pacific super typhoon” Emanuel, K. A., (1999) The mass of steam, gases, ash above the crater is initially propelled at 5.3 times the speed of sound from the conduit. Petersen et al. The tremendous amount of heat contained in steam and ash entrains a massive amount of air from the ambient atmosphere, the volume of the plume increased by 352 times from crater to tophat. At 9.3 km asl, above the tropopause, momentum flux is at a peak and power generated is 101TW. At tophat 12.2 km power generated is 115TW equivalent to the power of 38 hurricanes or 3.8 super typhoons.
Volume of event 9 eruption plume was calculated using data derived from PlumeRise model and USGS reports.
1 Plume volume Initial volume 4.42e8m³
99.9% steam Steam 4.7% total mass
2 V=⅓πr²h V₁ =2.89e10m³ r=1776m h=8751m
V₂ =4.74e10m³ h=3250m r= 3730 Cone volume
V₃ =4.13e9m³ h=1250m r=1776m
Cs₁ = V₂- V₃ Cs₁=4.33e10m³ Conic section
V₄ =1.3e11m³ h=3450m r=6000m Cs₂= V₄- V₂
Cs₂ = 8.3e10m³ Tophat 12193m
Vt= V₁+ Cs₁ + Cs₂ = 1.55e11m³ 350 times initial volume =
3 At 3253m asl V=4.2e8m³ initial volume Ambient air entrained
115 Terrawatts power at plume top 13.5 TW at source
1. data derived from USGS reports, McGee et al. see Table 1 herein.
for initial volume of plume
2. see Figures 1& 2 in Appendix
V₁ height 10000m-1250m
V₂ height 12000m – 8750m
V₃ height 10000m- 8750m
V₄ height 12200m- 8750m
Refer to diagrams to calculate volume >10000m, find in Appendix: Figures 1& 2
3. data derived from Plumerise model, Woodhouse et al.
See Dataset1: plumedata1.xlsx
Volume of plume was calculated from a series of cones and conic sections. Volume data was put into Kineticenergy worksheet to calculate mass of plume at elevations and translational energy (Kt) upward. The upward thrust of energy, mass of plume times vertical wind squared Kt=½mu², accounted for less than 2% of total kinetic energy (KE) and a mean of less than 0.5%. Rotational kinetic energy (Kr) accounted for at least 98% at any elevation, inertia of mass times vorticity. Kr=½ Iω ² was calculated in steps. First, v= usinθ: tangential velocity at point on radius equals vertical velocity times sin of angle 360/2π. Second, angular velocity ω=v/r: tangential velocity/ radius. Third, vorticity of plume = angular acceleration squared: ω². Fourth, W/ω=τ: Watts/ angular velocity = torque. Fifth, I=τ /ω²: torque/ vorticity= inertia. Sixth, τ=Iω²: inertia x vorticity= torque. Skip step 6 & go to Kr=½ Iω ² or skip step 5 and go to Kr= τ/2. Total KE generated at ~300mb just below tropopause ~1.16PetaJoules, just above tropopause ~1.36PJ & ~3.7PJ at 200mb. The impact of an immense mass of air 8.6Megatonnes at 9253 m ~300mb deposits kinetic energy and puts a ripple in the Rossby wave. The ripple is amplified as it rises with less pressure in the atmosphere. See Appendix: Dataset 3: Kineticenergy.
Data generated by PlumeRise model was very useful in supporting a comparison between event 9 and MERRA data (Modern Era Retrospective-Analysis for Research and Applications). (NASA). Data for upper level of stratosphere, 10hPa for the period January 1st to March1st were downloaded and compared with PlumeRise1 data. MERRA data for the parameters of zonal wind velocity (U), heat flux (VT), momentum flux (UV), Z1 & Z2 waves of geopotential heights were plotted against daily mean values of the AO (arctic oscillation).
The Arctic Oscillation is a measure of sea level variations between the Arctic region (Thule) and the subtropics (Azores). In the positive phase higher pressure in midlatitudes and lower pressure in the arctic results in strong zonal winds in the polar region. In the negative phase cold arctic air escapes into midlatitudes when zonal winds weaken. Thompson et al.
Sample sets were selected for expected trends of sea level pressure versus MERRA data. Correlations were plotted to illustrate an AO response to a local disturbance notably a volcanic eruption near the Arctic Circle. Correlations were found between each set of MERRA variables and AO variables. Scatterplots were made with transformed data to illustrate slope of trend lines. Transformation did not change correlation coefficient or skew. Arctic oscillation decreased to negative values as Tmin, UV, VT, Z2 increased and U & Z1 decreased. See Appendix: MERRA1: MERRA10hPa. The momentum flux along the centerline trajectory in the plume (UV) of the 2006 Augustine eruption reached a maximum of 2.98e10kg/ms² at 9313 meters on Jan 17th. See Appendix: Dataset1 & Chart1. Two days later on Jan 19th MERRA data indicates momentum flux (UV) 45 -75°N, a vector quantity between meridional and zonal wind speeds and direction generally poleward, reached a maximum (270 m²/s²). See Appendix: Dataset4: 10hPaAugJ06. At 60°N momentum flux (m2/s2) reached a maximum in mid-January at the time of event 9 eruption, 90% above the mean. See Appendix: MERRA2: Momentum Flux. Zonal winds (U) in late December 2005 rose above 40m/s westerly then sharply dropped to zero then further changing direction on Jan 21st and reaching a maximum velocity 26 m/s easterly on Jan 26th. See Appendix: MERRA3: Zonal Wind.
Minimum temperatures (Tmin) at 10hPa (upper stratosphere) 50-90°N rose steadily following event 9 eruption from a low of 184.78K° on Jan 6th to a high of 206.69K° on Jan 27th, three days before SSW. In the beginning of January 2006 minimum temperatures reached below 186K° at 10hPa favorable for the formation of Polar Stratospheric Clouds, 90% below the mean. A sharp spike can be seen on the chart from this low point to 203.52K°, a nearly 19K°rise in ten days. After a drop to the mean temperature, it rose to maximum 206.69K° in 11 days for a rise of nearly 22K°, 90 % above the mean. See Appendix: MERRA4: Minimum Temperature.
Kinematic heat flux (VT) 45-75 N reached a peak 257 Km/s at 10hPa on Jan 20th, three days after event 9 eruption. True volcanic heat flux (Vt) from eruption at 9303m (270hPa) is Watts/m² 1.01e14 W/ 7.68e6 m² = 1.32e7 W/ m² derived from plumerisedata1. To evaluate the magnitude of the volcanic heat flux in comparison to kinematic heat flux involves the derivation of MERRA heat flux (Vk) into true heat flux (Vt). On Jan 21st heat flux reached maximum at 150hPa of 33.7 Km/s (VT) see Appendix: MERRA5: Heat Flux & Data Set 5. Assuming atmospheric values of specific heat (1006 J/K kg) and density (0.236 kg/m³) at 150hPa: Vk x Cp x β = Vt = 8.0e3 W/m².
Volcanic heat flux 1.64e14 W/ 1.16 m² = 1.41e7 W/m² at 150hPa, 1767 times greater than Vt; see Appendix: Dataset2: plumedata2. This data was used to approximate heat flux at higher elevation with input of initial vertical velocity of higher value (1750 m/s). Initial thrust of explosion of event 9 involved the opening of vent with removal of plug and subsequent burst of supersonic burst of matter from superheated steam. This is noted as an impulsive event by Petersen et al.
Heat flux (Km/s) at 60°N was 90% greater than mean during mid-January 2006. See Appendix: MERRA5: Heat Flux. Volcanic heat flux of this magnitude impacts the Jetstream and slows down zonal winds considerably. MERRA kinematic heat flux 45-75°N peaks at 257 K m/s on Jan 20th, three days after event 9 eruption. Heat flux decays as a log function with height by more than 2 orders of magnitude from 1 km above crater to tophat.Heat flux as a function of radius decays exponentially from 2km above crater to 11km. See Appendix: Datasets 6, 7 & 9. Torque (Nm) grows logarithmically by nearly 2 orders of magnitude from 1 km above the crater to a height of 11km. Heat flux decays logarithmically by more than 1 order of magnitude in same distance. As a function of radii, torque grows exponentially as heat flux decays. See Appendix: Datasets 9 & 10. Zonal wind (U) at 60°N slows on Jan 20th and goes negative (easterly) on Jan 21st, reaching a peak -26.15m/s five days before SSW on Jan30th, remaining easterly for twenty five consecutive days. See Appendix: Dataset4: MERRA10hPa.
The explosive phase of the eruption generated a series of upward propagating gravity waves into the stratosphere. The University of Alaska Infrasonic Array at Fairbanks detected a series of 12 infrasonic signals from the 2006 Augustine eruption occurring Jan 11 through the 28th. Olson et al. These signals were determined to have passed through the stratosphere and thermosphere on the way from Cook Inlet to Fairbanks, a distance of 675 km (sea level). The signals consisted of wave trains with period of three to ten minutes, in the range of gravity waves. Observed signal times correspond with explosive events on dates Jan 11, 13, 14, 17 and 28 as noted in Table 1 Olson et al & Table 4 of Chapter 9 USGS report Wallace et al. The shock waves of events 1-10 deposit kinetic energy and a ripple effect to the Rossby waves of the Jetstream. It is evident by examining the waves of the 200mb upper wind analysis that a volcanic signal of the 2006 explosive events has an impact on the Jetstream. An orange star is placed at the location of the Augustine eruption on each map. Spaces between isotachs widen (winds slow) and ripple where there is a star. The explosive events slow the winds aloft, amplify the low pressure at the surface and cause divergence aloft. The trough over North America to the east of the disturbance deepens strongly after event 9. Wave breaking can be seen over the Caspian Sea (1/13), Greece (1/14) and Europe (1/15). Wave blocking is evident over Kamchatka and Europe (1/27) & Siberia (1/30). See Appendix: Chart2: 200mb Rossby wave analysis heights/isotachs. Maps courtesy of NOAA.
A further analysis of vorticity on 500mb maps illustrates an increase in vorticity since the explosive phase of the Augustine eruption on January 11, 2006. See Appendix: Chart3: 500mbvorticity. A cyclone develops in the North Pacific on January 13th and is in full swing by the 15th. Deep troughs develop on January on the 13th and extend in the Pacific west of California and the Midwest. These troughs travel eastward and deepen over California and Florida on the 15th. These troughs reach into Mexico and the North Atlantic on the 18th as the Rossby waves travel eastward. Cyclones begin to generate in the Atlantic on the 19th and a storm is evident in the North Atlantic on the 23rd. Analysis of Upper Air data indicate very high potential vorticity generated at Alaska stations near the center of volcanic activity on January 17th. See Dataset11: results are tabulated in last columns (U,V) with symbols for anticyclonic (clockwise) and cyclonic (counterclockwise) movements. Potential vorticity anomalies existed at high values >25 PVU over the center of volcanic activity since the inception of the explosive phase of eruption and continued through the month of January. A baroclinic atmosphere was created in the troposphere in the vicinity of the eruption as can be shown by Upper Air data for Waves 1 &2 on Jan. 16th at station PAOM and Jan. 17th PAYA as noted by a sharp drop in the dew point. University of Wyoming. Widespread instability in the troposphere from stations across Alaska show a baroclinic drop in dewpoint and temperature in the lower atmosphere on Jan 17th to 18th. Top of troposphere layers range from 11km (PAYA) Yakutat, 10.6 km (PAOT) Kotzebue, 9.9 km (PAOM) Nome, 9.6 km (PAFA) Fairbanks to 8.9 km (PADQ) Kodiak near eruption, 8.6 km (PACD) Cold Bay. Map of stations is available on UWYO website. A hydrolapse is noted in Data set for each station showing sharp drop in dewpoint at designated height. Data indicates a layer of dry cold air at higher elevations over a layer of warmer moist air. As warm moist air rises over dry cold air, conditions of convective instability and turbulence develop in the upper atmosphere. The evidence of a widening between Jetstream isotachs above the region of volcanic disturbance points to these conditions were created locally by the vortex of the steam and ash ejected by the volcano. See Appendix: Data sets 12-18. Data is highlighted in documents to denote hydrolapse.
Calculation of wave speed of Rossby waves eastward demonstrates that an atmospheric wave at latitude 45N can travel from Alaska to the Northeast Coast in less than three days. See Data set 21.
11 days past the explosive phase of the Augustine eruption a cyclone of massive proportions developed over Northeast America from Newfoundland to North Carolina. Ncdc.noaa.gov. This was a snowfall blizzard with accumulations over 27 inches in NJ and CT commencing on February 11th to 13th. Moore P.D., Conditions in Alaska in February were warmer and wetter with record temperatures (+20F interior) and snowfall (+300% interior). NWS.
The evidence for cyclogenesis caused by volcanic explosive events is strong based on the facts presented herein. The shear mass (47 Megatonnes) and kinetic energy (176PJ kinetic energy) of the plume at 12.2 km of event 9 was sufficient to disturb the Jetstream and set the Rossby waves into a meridional flow pattern. The power generated by this single event was 115TW equivalent to nearly 4 super typhoons. The tremendous amount of heat and pressure generated within the magma chamber exploded out of the crater and thereby entrained the air around the plume to a much larger volume. At least 98% of the air was entrained by the cyclonic action upward of the plume as indicated by kinetic energy calculations. MERRA atmospheric data coincides with Plumerise data and the evidence of a volcanic explosive event as the cause of the disturbance. It is significant that momentum flux of the plume reached a maximum at Jetstream level locally above the volcano on the 17th. Two days later MERRA data indicated a maximum momentum flux over a much larger area 45N- 75N. On 200mb Jetstream maps, a clear disturbance can be seen on waves 1 & 2 on the 17th. The following two days show the trough of the Rossby waves deepening, a meridional flow. Supersonic shock waves were also noted by Olson et al., Caplan-Auerbach et al. and Medici et al. (Table 1).These events occurred on Jan. 11, 13 and 17. A ripple in waves 1 & 2 on 200mb maps can be seen on these dates coinciding with these volcanic explosive events. In addition upper air soundings indicate a baroclinic atmosphere created at the time and place of event 9. 500mb Vorticity maps and calculations indicate a PVA anomaly above the volcano at this time and place. These two conditions are favorable for the formation of a cyclone. Cyclones occur downstream of the volcanic events in Alaska as the meridional flow pattern increases and storm conditions develop in the Northeast. Thus a large meridional flow resulted in subtropical air and seawater going northward and cold Arctic air going southward as the Rossby waves traveled eastward.
The kinetic rotational energy of the volcanic plume (event 9) imparted a vorticity anomaly to the atmospheric Rossby waves. This can be observed on Jetstream maps 100mb through 500mb analysis. Wave 1 longitude 145W deviates from 65N to 73N, NVA. Wave 2 at 145W deviates from 60N to 52 N, PVA. Wave 3 at 145W deviates from 52N to 45N, PVA. Wave 4 at 145W deviates from 48N to 42N, PVA. At longitude 100 to 95W, Wave 1 & 2 are flat response, Wave 3 &4 deviate 7 & 8 degrees south respectfully. As Rossby waves travel east, vorticity imparts a meridional pattern in response to kinetic energy and vorticity of volcanic plume originating at Cook Inlet. This results in the development of cyclogenesis in the area of the Eastern seaboard of the US and Northern Atlantic Ocean. This can be observed on 500mb vorticity maps. Deep troughs in the Jetstream are evident on Jan 25th then cyclones form over the Ocean on the following days. On Feb. 2 through Feb.8, cyclones form in the Deep South and North Atlantic. By Feb 11th, an immense positive vorticity spreads from Colorado 105E south through Texas to the Flemish Cap 40E in the North Atlantic. This is the massive storm that hit the Eastern Seaboard on that date.
Microsoft Word Computer Programing for Word and Excel throughout document
Google and Mozilla Firefox for quick internet research and storage for documents
USGS Volcanic Explosivity Index http://volcanoes.usgs.gov/images/pglossary/vei.php
NASA, Annual Meteorological Statistics for the Northern Hemisphere http://acd-ext.gsfc.nasa.gov/Data_services/met/ann_data.html#ncep_clim_stats_nh
University of Wyoming, College of Engineering, Department of Atmospheric Sciences, http://weather.uwyo.edu/upperair/sounding.html
NOAA National Climatic Data Center, http://nomads.ncdc.noaa.gov/ncep/NCEP
Moore P.D., National Weather Service; Greer, SC; http://www.erh.noaa.gov/gsp/localdat/cases/2006/11-13Feb_Snow/FebSnow.html
NWS, Fairbanks Forecast Office of the National Weather Service and the Alaska Climate Research Center, with contributions by Ted Fathauer, Anton Prechtel, and Martha Shulski, February 2006 Summary; http://oldclimate.gi.alaska.edu/Statewide/2006/Feb06.html
Wallace, K.L., Neal, C.A., and McGimsey, R.G., 2010, Timing, distribution, and character of tephra fall from the 2005-2006 eruption of Augustine Volcano, chapter 9 of Power, J.A., Coombs, M.L., and Freymueller, J.T., eds., The 2006 eruption of Augustine Volcano, Alaska: U.S. Geological Survey Professional Paper 1769, p. 187-217 and spreadsheet. See Table 4.
Coombs, M.L., Bull, K.F., Vallance, J.W., Schneider, D.J., Thoms, E.E., Wessels, R.L., and McGimsey, R.G., 2010, Timing, distribution, and volume of proximal products of the 2006 eruption of Augustine Volcano, chapter 8 of Power, J.A., Coombs, M.L., and Freymueller, J.T., eds., The 2006 eruption of Augustine Volcano, Alaska: U.S. Geological Survey Professional Paper 1769, p. 145-185, 1 plate, scale 1:20,000, and GIS data.
Schneider, D. J.; Scott, C.; Wood, J.; Hall, T.; NEXRAD Weather Radar Observations of the 2006 Augustine Volcanic Eruption Clouds, AA(U.S. Geological Survey Alaska Science Center, Alaska Volcano Observatory, Anchorage, AK 99508 United States firstname.lastname@example.org), AB(National Weather Service, Anchorage Forecast Office, Anchorage, AK 99502 United States 😉 American Geophysical Union, Fall Meeting 2006, abstract #V51C-1686; 12/2006, 2006AGUFM.V51C1686S
Newhall C.G., Self S. The volcanic explosivity index (VEI) an estimate of explosive magnitude for historical volcanism; DOI:10.1029/JC087iC02p01231; Copyright 1982 by the American Geophysical Union.
Woodhouse, M.J., A.J. Hogg, J.C. Phillips, and R.S.J. Sparks (2013), Interaction between volcanic plumes and wind during the 2010 Eyjafjallajökull eruption, Iceland, Journal of Geophysical Research: Solid Earth 118, 92–109; DOI: 10.1029/2012JB009592
Emanuel, K.A., 1998: The power of a hurricane: An example of reckless driving on the information superhighway. Weather, 54, 107-108.
Petersen, T., De Angelis, S., Tytgat, G., and McNutt, S.R., 2006, Local infrasound observations of large ash explosions at Augustine Volcano, Alaska, during January 11-28, 2006: Geophysical Research Letters, v. 33, 5 p., doi:10.1029/2006GL026491
McGee, K.A., Doukas, M.P., McGimsey, R.G., Neal, C.A., and Wessels, R.L., 2010, Emission of SO2, CO2, and H2S from Augustine Volcano, 2002–2008, chapter 26 of Power, J.A., Coombs, M.L., and Freymueller, J.T., eds., The 2006 eruption of Augustine Volcano, Alaska: U.S. Geological Survey Professional Paper 1769, p. 616 Table 1 [http://pubs.usgs.gov/pp/1769/chapters/p1769_chapter26.pdf].
Thompson, D.W.J.,Wallace J.M., 1998: The Arctic Oscillation signature in the winter geopotential height and temperature fields, Geophys. Res. Lett., 25, 1297-1300. DOI: 10.1029/98GL00950
Olson, J. V.; Wilson, C. R.; McNutt, S.; Tytgat, G.; Infrasonic Wave Observations of the January 2006 Augustine Volcano Eruptions, American Geophysical Union, 2006AGUFM.V51C1685O
Caplan-Auerbach J., Bellesiles A., Fernandes J.K.; Estimates of eruption velocity and plume height from infrasonic recordings of the 2006 eruption of Augustine Volcano, Alaska; Western Washington University, Bellingham, WA, 98225, United States; Journal of Volcanology and Geothermal Research (Impact Factor: 2.19). 01/2010; 189(1):12-18. DOI:10.1016/j.jvolgeores.2009.10.002
E. F. Medici, Allen J.S.,Waite G.P.; Modeling shock waves generated by explosive volcanic eruptions, Article first published online: 23 JAN 2014, DOI: 10.1002/2013GL058340
Supporting documents in Appendix