Logo link to homepage

Report on Atmospheric Effects (1995-2001) — April 1999

Bulletin of the Global Volcanism Network, vol. 24, no. 4 (April 1999)
Managing Editor: Richard Wunderman.

Atmospheric Effects (1995-2001) Tracing recent ash by satellite-borne sensors and ground-based lidar

Please cite this report as:

Global Volcanism Program, 1999. Report on Atmospheric Effects (1995-2001). In: Wunderman, R. (ed.), Bulletin of the Global Volcanism Network, 24:4. Smithsonian Institution.



Atmospheric Effects (1995-2001)

All times are local (unless otherwise noted)


Observers at the Alaska Volcano Observatory initially inferred that the 19 April Shishaldin plume reached ~13-14 km altitude based on what appeared to be as the most reliable pilot reports (see above and Bulletin v. 24, no. 3). Yet, one pilot reported the plume to 18.3 km altitude and satellite data suggested similar altitudes. Through at least late May, scientists continued to detect and track stratospheric aerosols. At the time of this writing we have learned of successful satellite detection by GOES 10, the Total Ozone Mapping Spectrometer (TOMS), Stratospheric Aerosol and Gas Experiment (SAGE II), and the Polar Ozone and Aerosol Measurement (POAM). Ground-based lidar also detected presumed Shishaldin aerosol layers far from the source.

GOES observations. GOES 10 data portrayed early images of the plume (figure 6). According to Dave Schneider, thermal split-window imagery showed curiously little evidence of the plume in the stratosphere. Detection conditions were non-ideal: a warmer stratospheric cloud (the plume) overlying a colder tropospheric cloud deck. He also commented on a lack of evidence for ash at lower levels and wondered what role sulfate may have played.

Figure with caption Figure 6. Detailed view of the spreading eruption cloud from Shishaldin on 19 April, taken from a series of images taken by the GOES 10 satellite (channel 1). Unimak Island is outlined. The top frame was imaged at 1200; the following frames at subsequent half-hour increments. The cloud labeled "A" was at higher altitude and moved N; cloud "B" was at lower altitude and moved S. Courtesy of NOAA/NESDIS.

TOMS observations. The TOMS instrument rides aboard NASA's Earth Probe satellite and collects information about airborne gases and particles, including ozone, SO2, and volcanic ash. TOMS passed over Shishaldin at 2142 GMT on 19 April, two hours after the eruption began as a small white plume in the GOES images. Thus, TOMS captured an early stage of the event while the eruption column was actively growing. This early post-eruption data reflected very high concentrations of SO2 and ash in a pixel over the volcano and smaller amounts in two adjacent pixels (unshaded boxes, figure 7). The TOMS images can now retrieve ash as well as SO2 concentrations; the dense 19 April plume, however, was not conducive to realistic SO2 measurement.

Figure with caption Figure 7. TOMS measurements of SO2 near Alaska on 19 April (unshaded boxes, no scale) and 20 April 1999 (shaded boxes, see scale at right) with respect to local coastal margins (lines) and Shishaldin volcano (triangle). For 20 April, shaded boxes (pixels) indicate SO2 gas concentrations of up to 40 milliAtm · cm. Each pixel represents a footprint of TOMS (about 40 x 40 km) containing SO2. The total SO2 depicted in the 20 April image, obtained by summing the pixels, was 20 ktons. (TOMS Orbits 15142 and 15168.) Courtesy of Arlin Krueger and Steve Schaefer, NASA/GSFC.

On 20 April the Shishaldin cloud was still found close to the volcano as an arc-shaped plume of SO2 (figure 7) to the N of and disconnected from the volcano. However, no detectable ash remained in the plume. This dispersed cloud was used to determine that the mass of SO2 in the eruption was 20 ktons. Traces of this SO2 cloud still remained on 21 April after drifting slightly to the N, but were gone on 22 April.

POAM III and SAGE II satellite observations. As discussed on their web site (NRL, 1999) the POAM instrument was developed by the U.S. Naval Research Laboratory (NRL) to measure the vertical distribution of atmospheric ozone, water vapor, nitrogen dioxide, aerosol extinction, and temperature. Solar extinction by the atmosphere is measured using the solar occultation technique; the sun is observed through the Earth's atmosphere as it rises and sets as viewed from the satellite. POAM data on stratospheric aerosols provide information on how the aerosol burden varied with altitude, latitude, season, and annually in a record going back over 3 years. The data have good vertical resolution (1 km), wide geographic coverage, and dense sampling in the polar regions over the latitude range 55°N-71°N. The following discusses data collected by the instrument's latest version (POAM III). SAGE II, another very similar satellite-based, limb-profiling technique has also contributed data.

As shown on figures 8 and 9, trajectory modeling and observational data from POAM III and SAGE II indicated that air parcels moving away from the eruption column at different altitudes took very different paths during the days following the eruption. The forward trajectory model (figure 8) shows strong correspondance with those run independantly by Barbara Stunder at the same altitudes. The modeling indicated that the part of the plume at ~12 km altitude first moved slightly SW, then E, then NE, and finally ESE. Modeling also indicated that the part of the plume at ~18-19 km altitude moved N and varying amounts to the E. In accord with this, high altitude volcanic aerosol material was detected N of 70°N latitude on 23 April by SAGE II. Finally, the modeling indicated that the part of the plume at ~14-16 km altitude branched away from the higher altitude material and began heading E. On 23 April the plume was observed on a POAM III profile (figures 8 and 9).

Figure with caption Figure 8. Simulated air parcel trajectories that originated at Shishaldin on 19 April (squares), and profiles actually observed by POAM III (open circles). The figure illustrates stratospheric circulation and observations during 9 days beginning at 2000 GMT on 19 April and ending at 2000 GMT on 28 April. The trajectory paths were estimated from the motion of their respective associated air parcels at the indicated altitudes; each successive square indicates the results of 24 hours movement. The three aerosol layers sensed by POAM III limb profiling are depicted as circles that have these dates, center coordinates, and altitudes: 23 April at 62.7°N, 197.1°W, 15 km altitude; 25 April at 62.4°N, 207.9°W, 14 km altitude; and 27 April at 62.0°N, 218.7°W, 13 km altitude. Courtesy of Mike Fromm, NRL.
Figure with caption Figure 9. POAM III aerosol extinction ratios on 23, 25, and 27 April. The peaks are due to volcanic aerosols from Shishaldin. The measurement locations are given in the caption for figure 12. Normal background ratios are 1-2. Courtesy of Mike Fromm, NRL.

Figure 9 illustrates aerosol extinction ratios for the aerosol layers seen on 23, 25, and 27 April (circles, figure 8). The peak values shown in figure 9 lie 3-4 standard deviations above the normal background. Anomalously high extinction ratios in the lower stratosphere such as these continued well into May. The plot indicates the plume's height progressively decreased during the course of the three observations, descending from altitudes of ~15 to ~13 km, implying that the volcanic particulate settled out at roughly 0.5 km per day.

Figure 10 illustrates the POAM III results for several weeks following the 19 May Shishaldin eruption. It maps the location of all available POAM profiles (+ symbols) and 14 profiles with varying loads of enhanced stratospheric aerosols (circles). Larger circles indicate larger aerosol loads; more specifically, the circle sizes vary in proportion to the peak aerosol enhancement, determined in relation to the standard deviation of the aerosol extinction ratio in relevant background conditions. The altitudes of peak extinctions varied from 12-15 km.

Figure with caption Figure 10. A polar orthographic projection showing POAM III profiles taken 23 April-11 May 1999. The locations of all possible profiles during the period appear as crosses. The locations of 14 profiles with enhanced stratospheric extinction appear as open circles (labeled with their observation dates). Circle sizes vary in proportion to the peak aerosol enhancement (see text). In back-trajectory models, the profiles with starred dates can be traced back to Shishaldin on 19 April (see text). The map's line-spacings are as follows: latitude, 10°; longitude, 30°. Courtesy of Mike Fromm, NRL.

Looked at on the scale of weeks after the eruption, the atmospheric circulation carried 19 April eruptive products towards the W. For the starred profiles on figure 10, isentropic (constant entropy, which assumes conservation of potential temperature) modeling of back trajectories strongly suggested Shishaldin as the source. POAM III continued to detect enhancements of aerosols in the lowest stratosphere at least until 23 May. The latitudes of the profile's center points moved gradually from about 62°N in late April to 57°N in late May.

Attempts to link additional POAM III observations (those that lack stars) with Shishaldin through isentropic trajectory analysis is in progress, but thus far some of them have failed to lead either back to Shishaldin or to another clear source. Around 5-6 May, for example, two stratospheric aerosol layers resided over or near Hudson Bay, Canada and were also not traceable to Shishaldin in trajectory models. As for another layer at that time, S of Iceland, the models indicate a likely source at the eruption.

Ground-based lidar observations. Lidars (light radars), which measure the amount of backscattered laser energy due to plume and atmospheric conditions (Jørgensen and others, 1997), detected aerosol layers over Germany, Virginia, and Greece. Beginning on the evening of 6 May, Horst Jäger detected a stratospheric layer while profiling with a 532-nm wavelength lidar operated in Garmisch-Partenkirchen, Germany (47.5°N, 11.0°E). His 6 May data showed a small but pronounced peak in the scattering ratio (figure 11). The source of the anomaly was between 15.1 and 15.4 km altitude, well above the estimated maximum altitude of the local troposphere (11.4 km, as determined by a midnight radiosonde from Munich). A maximum scatter ratio of 1.35 occurred at 15.2 km.

Figure with caption Figure 11. Lidar backscatter ratios as a function of height as measured from the 532-nm lidar at Garmisch-Partenkirchen, Germany on 16 May 1999. Courtesy of Horst Jäger.

On 9 May the atmosphere lacked detectible layers in the expected altitude region. On 16 May the lidar achieved maximum scatter ratios of 1.1-1.2 at 14.3, 15.6, 16.3, and 17.3 km. Thus, over Germany, the layers did not form a major perturbation to the stratosphere; these faint backscatters became prominent only because of the low aerosol background during the times of measurement.

The altitude and timing of the peak in German lidar suggested a link to the 19 April Shishaldin eruption plume. The last eruption to produce similar results at the Garmisch-Partenkirchen site was the October 1994 eruption of Kliuchevskoi (Bulletin v. 19, no. 10). That plume reached heights of 25 km.

At Hampton, Virginia, ground-based 694-nm lidar also showed high-altitude peaks (table 17). Measurements there on 11 May detected a diffuse layer (with a peak ratio of 1.17) that was narrow (~1 km thick) and located at 16.9 km altitude, well above the tropopause height. Measurements on 21 May also disclosed two narrow layers. One had a peak ratio of 1.10 at 17.5 km; the other, a peak ratio of 1.19 at 14.5 km. The presence of particles at this height are generally considered to be associated with an eruption; the timing of these observations suggested the layers were due to the 19 April Shishaldin eruption. This may imply that the erupted aerosols had reached mid-latitudes during the month following the eruption.

Table 17. Lidar data from Virginia, USA, for February-May 1999 showing altitudes of aerosol layers. Backscattering ratios are for the ruby wavelength of 0.69 µm. The integrated values show total backscatter, expressed in steradians-1, integrated over 300-m intervals from the tropopause to 30 km. Courtesy of Mary Osborne.

DATE LAYER ALTITUDE (km) (peak) BACKSCATTERING RATIO BACKSCATTERING INTEGRATED
Hampton, Virginia (37.1°N, 76.3°W)
11 Feb 1999 11-27 (23.5) 1.10 5.03 x 10-5
23 Feb 1999 10-27 (24.1) 1.09 5.93 x 10-5
05 Mar 1999 09-25 (10.7) 1.11 6.61 x 10-5
14 Apr 1999 15-27 (22.1) 1.09 2.49 x 10-5
11 May 1999 12-26 (16.9) 1.17 4.72 x 10-5
21 May 1999 13-27 (14.5) 1.19 4.48 x 10-5

Commenting on research conducted on the Mediterranean island of Crete (35°30'N, 23°43'E), Christos S. Zerefos reported that the portable VELIS lidar instrument of Gian P. Gobbi also detected an aerosol layer during 10-13 May. Profiles disclosed increased aerosols at 15-16 km altitudes. Aerosols were seen again on 14 May, but they were not detected on 15 May. The optical depth at 532 nm was at most 0.02.

Conclusions. The 19 April Shishaldin eruption provided a modest injection to ~17-19 km altitude and a TOMS estimate the next day found ~20 kt of SO2 . In trajectory models, components of the plume at various altitudes moved away from the source in 3 branches; POAM III profiles on the ENE-directed path showed the plumes there decreased in altitude with time. Trajectory models have yet to confirm that several POAM III profiles came from the Shishaldin eruption and at this point their source remains ambiguous. The exact trajectories that presumably carried the Shishaldin aerosols over the German, Crete, and Virginia lidar systems have yet to be either consistently traced or modeled.

References. Hans, E., Jørgensen, H.E., Mikkelsen, T., Streicher, J., Herrmann, H., Werner, C., and Lyck, E., 1997, Lidar calibration experiments, Applied Physics B, Lasers and optics, v. 64, no. 3, Springer-Verlag, p. 355-61.

F. Congeduti, F. Marenco, E. Vincenti, P. Baldetti, and G.P. Gobbi, 1998, The new transportable lidar facilities at IFA: 9-eyes and VELIS, in Proceedings of the Workshop onSynergy of Active Instruments in the Earth Radiation Mission,M. Quante and others (eds.), http://aragorn.gkss. de/deutsch/Radar/workshop_papers.html

Naval Research Lab, 1999, Remote Sensing Division, Remote Sensing Physics Branch, Middle Atmospheric Physics Section, POAM Home page, http://wvms.nrl. navy.mil/POAM/poam.html.

Lidar Researchers Directory (including a bibliography produced by NASA) URL: http://arbs8.larc.nasa.gov/lidar/directory.html.

Sparks, R.S.J., Bursik, M.I., Carey, S.N., Gilbert, J.S., Glaze, L.S., Sigurdsson, H., and Woods, A.W., 1997, Volcanic plumes: John Wiley and Sons, Ltd., ISBN-0-471-93901-3, 574 p.

Information Contacts: Horst Jäger, Fraunhofer - Institut für Atmosphärische Umweltforshung (IFU), Kreuzeckbahnstrasse 19, D-82467 Garmisch-Partenkirchen, Germany; Mike Fromm, Computational Physics, Inc., 2750 Prosperity Avenue, Fairfax, Virginia, 22031 USA; Naval Research Laboratory, 4555 Overlook Ave. SW, Washington, DC 20375 (URL: http://www.nrl. navy.mil); Barbara Stunder, U.S. National Oceanic and Atmospheric Administration (NOAA), Air Resources Laboratory, SSMC3, Rm. 3151 (R/E/AR), 1315 East-West Highway, Silver Spring, MD 20910, USA; Mary Osborn, NASA Langley Research Center (LaRC), Hampton, VA 23681 USA; Christos S. Zerefos, Aristotle University of Thessaloniki, Physics Department, Laboratory of Atmospheric Physics, Campus Box 149, 540 06 Thessaloniki, Greece; Arlin J. Krueger and Steve Schaefer; TOMS Instrument Scientists, Code 916, Building 33, Room E413, Goddard Space Flight Center, Greenbelt, MD 20771, USA, Dave Schneider, Alaska Volcano Observatory (see Shishaldin).