Atmospheric Science
Free Access

Violent pyro‐convective storm devastates Australia's capital and pollutes the stratosphere

Michael Fromm

E-mail address:mike.fromm@nrl.navy.mil

U.S. Naval Research Laboratory, Washington, D.C., USA

Search for more papers by this author
Andrew Tupper

Northern Territory Regional Office, Commonwealth Bureau of Meteorology, Casuarina, Northern Territory, Australia

Search for more papers by this author
Daniel Rosenfeld

Institute of Earth Sciences, Hebrew University of Jerusalem, Jerusalem, Israel

Search for more papers by this author
René Servranckx

Canadian Meteorological Center, Environment Canada, Dorval, Quebec, Canada

Search for more papers by this author
Rick McRae

ACT Emergency Services Authority, Curtin, A.C.T., Australia

Search for more papers by this author
First published: 15 March 2006
Cited by: 38

Abstract

[1] Headline‐making firestorms in southeast Australia in 2003, responsible for at least 500 destroyed buildings and four lost lives, culminated with pyro‐cumulonimbus (pyroCb) “eruptions” that ravaged Canberra on 18 January. Here we reveal that in their 3‐hour lifetime, the Canberra pyroCbs also produced a stratospheric smoke injection that perturbed the hemispheric background analogous to the theorized “nuclear winter.” We use an unprecedented array of data to analyze the Canberra pyroCbs' distinctive stratospheric impact, microphysics, energetics, and surface manifestations—including suppressed precipitation, an F2 tornado, and black hail.

1. Introduction

[2] Forest fire plumes in the stratosphere, hypothesized in the “nuclear winter” theory [Turco et al., 1983], have been discovered in recent years [Fromm et al., 2000]. Initial speculation that the stratospheric smoke was from severe convection was confirmed in a later case study [Fromm and Servranckx, 2003] of a pyro‐convective blowup in Alberta in 2001. This extreme combination of heat‐energy release and convection has since been identified as the “smoking gun” for observations of biomass‐burning particles and carbon monoxide [Jost et al., 2004], acetonitrile [Livesey et al., 2004], and increased ozone [Fromm et al., 2005] deep in the stratospheric “overworld” [Holton et al., 1995]. Until now the internal dynamics of the pyroCb have not been well observed—they have occurred far from dense observational networks. Moreover, it is still unclear how widespread the pyroCb phenomenon is and what its climate implications are. Herein we report on a pyroCb that advances our understanding significantly on both fronts.

2. The Australian Fires and PyroCb of January 2003

2.1. Bushfire Setup Conditions

[3] Devastating fires over the Canberra and Snowy Mountains regions of southeastern Australia (see Figures 1, 2, and 4 for maps) started from lightning strikes on 8 January [Webb et al., 2004]. Prior to the fire season, the 2002/03 El Niño event had brought abnormally low rainfall and high temperatures to large parts of eastern Australia [Webb et al., 2004]. Between 8 and 17 January, the fires persisted as drought levels remained elevated in mild, anti‐cyclonic weather conditions. On 18 January the passage of a low pressure trough brought strong, dry westerly winds across the area. Day‐time heating caused mixing downward of the stronger upper level winds. Wind speeds of 13 m/s with gusts to 22 m/s were recorded shortly after 0420 UTC (1520 LT). High temperatures (maximum 37.4°C at Canberra airport, roughly 12 km from the fires) and low relative humidity (minimum 8%), combined with the very dry vegetation and the strong wind, produced extreme fire danger conditions [Webb et al., 2004]. The upper air sounding from Wagga Wagga, 163 km to the west, suggests the regional atmosphere was potentially unstable despite dry air near the surface (see auxiliary material). In fact, thunderstorms did form on the ranges and the coast to the north of Canberra on the afternoon of 18 January. In this environment the Canberra pyroCb grew explosively (Figures 2a–2d).

image
Locations and observed phenomena discussed in the text. The green outline represents the border of the ACT; yellow represents urban area; light grey represents final burnt area; dark grey represents fire extent before 18 January 2003; and red hatched represents most intense convection at about 0400 UTC (1500 LT) 19 January (1500 18 January LT). Red line represents mapped tornado damage trail. (a) Photograph from Wanniassa, at 0408 UTC (1508 LT), 18 January, looking WNW. Blue arrow points to apparent tornado. Photographed by Jim Venn. (b) Detail from photogrammetric reconstruction of tornado path. Hatched area is the damage zone; the black lines are sightlines from the photo. GIS‐measured width of the damage zone (blue line) gives a 440m basal diameter.
image
(a) Aqua/MODIS true‐color image, 0330 UTC, 18 January 2003, of smoke plumes, pyroconvection, and “regular” convection. Outline of ACT and NSW coast in green. (b) Plan Position Indicator (PPI) view from Captain's Flat radar at 0340 UTC. (c) Range Height Indicator (RHI) view through pyroCb along WNW‐ESE axis indicated in Figure 2b. (d) RHI view through SW‐NE axis showing smoke (0–60 km) and developing Cb near Wollongong (160 km). (e) Particle effective radius (reff) and cloud top temperature relations, using MODIS 0330 UTC data, for pyroCb area (red) and storms near Wollongong (blue). Shown are the 15% (dashed line), 50% (solid line) and 85% (dashed line) percentiles of the reff for each 1°C interval. The vertical green line at 14 μm represents the precipitation threshold [Rosenfeld and Gutman, 1994].

2.2. The 18 January PyroCb

[4] During the peak flaming and fire spread in the afternoon of 18 January we use infrared linescans (see auxiliary material) to estimate that the instantaneous flaming zone of the combined firefronts close to Canberra occupied 5000 ha, generating an energy release of 4.3 × 1013 kJ. In the peak 10 minutes of flaming at about 0410 UTC (1510 LT), 3.5 × 1012 kJ were released. This is equivalent to 22 kT of TNT, more than the Hiroshima atomic bomb (15 kT), and exceeds typical energy release in thunderstorm cells by 1–2 orders of magnitude [Doswell, 1996]. The Australian Bureau of Meteorology (BOM) weather radar east of Canberra at Captains Flat captured the entire sequence of cloud development as well as the flying ash in the dry smoke plumes from the fires. The radar sequence (see auxiliary material) shows four main complexes of pyroCb over the Australian Capital Territory (ACT) with smaller complexes developing over the Snowy Mountains further south. The life of the Canberra pyroCb complex was ∼3 hours, with individual cells of diameter ∼10 km developing near the fire fronts and collapsing as they were advected to the east in westerly winds. The echoes from within the pyroCb were surprisingly and persistently weak when contrasted with the vigor of the clouds. We find no obvious supercell characteristics in the radar sequence. The maximum height of the echoes observed on radar was ∼15 km at 0450 UTC (1550 LT), a conservative estimate of the true cloud/plume height considering the radar particle‐size resolution.

[5] Effective particle radii have been calculated for the Canberra pyroCb cluster and a cell to the northeast (Figures 2a and 2e), using established 3.7μm‐based reflectivity techniques [Rosenfeld and Gutman, 1994; Rosenfeld and Lensky, 1998]. Figure 2e shows that the cloud particle effective radii in the pyroCb over Canberra were grossly reduced. The cloud above the −40°C isotherm was composed of ice particles with extremely small effective radius (consistent with the weak radar echoes), suggesting homogeneous freezing of most condensates while still in the form of small cloud droplets, which implies strongly inhibited precipitation‐forming processes. The cloud to the northeast had cloud droplets exceeding the precipitation threshold of 14 μm at and above −28°C, indicating greater precipitation potential. No precipitation was observed in or near Canberra apart from some “black hail” 30 km to the east, a manifestation of the strong updrafts and suppressed precipitation within the smoky pyroCb. The few precipitation particles that managed to form had a large supply of highly supercooled water to feed on, and only the large sooty hailstones could fall through the strong updrafts. The blackness of the hailstones shows that they were formed in the smoke plume that was the core of the pyroCb.

[6] Eyewitnesses reported multiple vortices in the pyroCb. Figure 3 shows quasi‐simultaneous observations near the developing pyroCb, close to the origin of the tornado damage path in Figure 1, near Mount Coree. Views include a photograph of a possible tornadic vortex (and antecedent tree damage), IR linescan hot spot, a photograph of active and mature pyroconvection, and a radar echo‐top map. Damage consistent with an F2 [Fujita, 1971] tornado occurred near Weston Creek (Figure 1). Three aspects of the damage path indicate a tornado rather than fire‐induced whirls or roll vortices: 1. the path dimensions–20 km long, up to 450 m across; 2. breaks in the damage path (consistent with the vortex temporarily lifting); and 3. damage extent well beyond the burn zone. Moreover, the vortex observations occurred close in time and space to the most rapid cell growth and pyroCb maturity. For example, the photograph in Figure 1, aligned with the damage path, near pyroCb maturity, indicates a possible funnel cloud. The observed cells and vortices are most consistent with relatively intense forms of non‐supercell tornadoes [Wakimoto and Wilson, 1989], fire‐induced vortices stretched by updrafts from rapidly growing pyroCb cells, forming F2‐strength tornadoes.

image
(a) Damage in pine trees east of Mount Coree on NW border of ACT, showing trunks snapped ∼3 m above ground and fine branches stripped off limbs (image taken 0200UTC, 31 May 2005). (b) Aircraft photo looking eastward, of pyroCb growing above location of Figure 1a at 0404UTC, 18 January 2003 (square 1), and mature pyroCb tops in distance (square 2). (c) 3D view of 1 dBz radar contour at 0410UTC, looking eastward to show the two pyroCb complexes identified in Figure 3b. (d) Fire linescan above location of Figure 3a, showing spot‐fire in clearing at ∼0412UTC. (e) Apparent tornadic vortex (square 3) to right of spot fire in Figure 3d at 0408UTC. The damage shown in Figure 3a occurred in the lower right area of this photo, presumably after the photo was taken.

[7] Taken together, the observations of weak echoes, small effective radii, black hail, and tornadic vortices are all consistent with fire‐enhanced deep convection having suppressed particle aggregation and precipitation processes [Andreae et al., 2004; Rosenfeld, 2000], which inhibits the formation of strong downdrafts and hence the efficient removal of the smoke from the atmosphere. As we shall see next, the resultant violent updrafts penetrated the tropopause and injected smoke into the stratosphere.

2.3. Stratospheric Impact

[8] The first indication of the Canberra pyroCb's potential climate impact came from satellites. Aerosol Index (AI) data from NASA's Total Ozone Mapping Spectrometer (TOMS), which has been successfully employed to identify large‐scale stratospheric smoke transport [Fromm et al., 2005; Jost et al., 2004; Livesey et al., 2004], detected several extreme smoke outbreaks from southeastern Australia between 18 January and 1 February. The 18 January pyroCb ejecta were detected on 19 January (Figure 4a), when several distinctly brown clouds over the Tasman Sea were captured in Moderate Resolution Imaging Spectroradiometer (MODIS) true‐color imagery. The 19 January AI map (Figure 4b), within one hour of MODIS, reveals a super‐intense node of smoke aerosols at 32°S, 162°E. AI is a strong function of plume altitude [Fromm et al., 2005, and references therein] ‐values of this extreme have only been observed on one other occasion in the 25‐year TOMS era, after the Chisholm pyroCb [Fromm and Servranckx, 2003]. We constrain the effective altitude of the plume with isentropic back trajectories, calculated on a dense vertical grid, all beginning at the TOMS measurement time and ending at 0330 UTC 18 January, when the Canberra pyroCb was active. The best temporal‐spatial match was found at the 385K potential temperature surface, in the stratospheric overworld.

image
(a) Terra/MODIS true color image of southeast Australia, 0000 UTC, 19 January 2003. (b) TOMS AI image for the same time. Backward isentropic trajectories, using analyses from the Met Office, from the location of extreme AI (denoted by an asterisk on both images) for potential temperatures 300K to 440K at 5K intervals. The trajectory start time is the TOMS observation time, ∼0100 UTC 19 January. The 385K back‐trajectory is a close (and the best) match for smoke from the Canberra fires, 20 hours previously. Other selected trajectories, from above and below the optimal match, are in red.

[9] Within one week of 18 January, stratospheric aerosol‐layer detections were recorded by NASA's Stratospheric Aerosol and Gas Experiment (SAGE) III instrument. The enhancements in SAGE III aerosol extinction profiles increased in frequency after 18 January and became a hemispheric feature within one month. Daily mean and median aerosol extinction at 385 K are shown in Figure 5, along with median extinction at 435 K (above the injection altitude). Figure 5 shows three periods. The pre‐blowup quiescence—relatively low levels of aerosol extinction—is evident in early January. The post‐blowup regional stratospheric impact (25 January–14 February) is manifested by the early increases in zonal mean and latter rise in the median. The full hemispheric pollution and decay period (14 February on) is characterized by the meeting of the mean and median, and the gradual decrease through March. At its peak hemispheric impact, the zonal median 385K aerosol loading was nearly 2.5 times the pre‐injection level.

image
Daily mean and median SAGE III 1020 nm extinction at 2 isentropic levels. SAGE III made 14 profile measurements equally spaced longitudinally each day, at latitudes between 34° and 46°S, between January and March 2003.

3. Summary

[10] This is the first major pyroCb event detected and analyzed for the Southern Hemisphere. Three aspects of this remarkable phenomenon require increased understanding. First, the tornadic and fire‐induced destruction underneath the Canberra pyroCb indicates the necessity of developing appropriate forecasting/warning techniques. Second, the pyroCb appears to be a unique storm, combining elements of “normal” (i.e., instability‐related) convection with explosive heat‐energy release at the ground. The heat‐aided convection produces a plume akin to certain volcanic eruptions. Indeed, the grossly suppressed cloud‐top particle sizes in the convection are quite similar to the 1991 Mt. Pinatubo eruption clouds [Tupper et al., 2005]. Lastly, it appears that the brief (∼3 hrs) and microscale (∼10 km diameter) pyroCb has now been shown to have the explosive capacity to pollute the austral stratosphere with absorbing aerosols that intercept solar radiation–with significant implications for weather and climate [Fromm et al., 2004].

Acknowledgments

[11] We thank Rob Webb, Clem Davis, and John McBride of the Australian BOM for help and discussions, the Australian BOM for radar data, and the NSW Rural Fire Service for photographs and eyewitness accounts.

    Number of times cited: 38

    • , Pyroconvection Risk in Australia: Climatological Changes in Atmospheric Stability and Surface Fire Weather Conditions, Geophysical Research Letters, 45, 4, (2005-2013), (2018).
    • , Stratospheric Smoke With Unprecedentedly High Backscatter Observed by Lidars Above Southern France, Geophysical Research Letters, 45, 3, (1639-1646), (2018).
    • , Pyrocumulonimbus lightning and fire ignition on Black Saturday in southeast Australia, Journal of Geophysical Research: Atmospheres, 122, 14, (7342-7354), (2017).
    • , Sub-micrometer refractory carbonaceous particles in the polar stratosphere, Atmospheric Chemistry and Physics, 17, 20, (12475), (2017).
    • , A Conceptual Model for Development of Intense Pyrocumulonimbus in Western North America, Monthly Weather Review, 10.1175/MWR-D-16-0232.1, 145, 6, (2235-2255), (2017).
    • , Detection and Inventory of Intense Pyroconvection in Western North America using GOES-15 Daytime Infrared Data , Journal of Applied Meteorology and Climatology, 10.1175/JAMC-D-16-0226.1, 56, 2, (471-493), (2017).
    • , Sensitivity of atypical lateral fire spread to wind and slope, Geophysical Research Letters, 43, 4, (1744-1751), (2016).
    • , Simulating the Black Saturday 2009 smoke plume with an interactive composition‐climate model: Sensitivity to emissions amount, timing, and injection height, Journal of Geophysical Research: Atmospheres, 121, 8, (4296-4316), (2016).
    • , A transdisciplinary approach to understanding the health effects of wildfire and prescribed fire smoke regimes, Environmental Research Letters, 10.1088/1748-9326/11/12/125009, 11, 12, (125009), (2016).
    • , Impact of the New South Wales fires during October 2013 on regional air quality in eastern Australia, Atmospheric Environment, 131, (150), (2016).
    • , Bibliography, Infrared Observation of Earth's Atmosphere, (201-210), (2015).
    • , Wildland fire emissions, carbon, and climate: Science overview and knowledge needs, Forest Ecology and Management, 317, (1), (2014).
    • , The worldwide “wildfire” problem, Ecological Applications, 23, 2, (438-454), (2013).
    • , An Australian pyro-tornadogenesis event, Natural Hazards, 65, 3, (1801), (2013).
    • , An assessment of satellite-based high resolution precipitation datasets for atmospheric composition studies in the maritime continent, Atmospheric Research, 122, (579), (2013).
    • , Defining adequate means of residents to prepare property for protection from wildfire, International Journal of Disaster Risk Reduction, 6, (67), (2013).
    • , Numerical and laboratory prediction of smoke lofting in the atmosphere over large area fires, Applied Mathematical Modelling, 37, 3, (876), (2013).
    • , Fine-dispersed particles in the natural and anthropogenic geosystems, Izvestiya, Physics of the Solid Earth, 48, 3, (256), (2012).
    • , Anatomy of a catastrophic wildfire: The Black Saturday Kilmore East fire in Victoria, Australia, Forest Ecology and Management, 284, (269), (2012).
    • , Wind - terrain effects on the propagation of wildfires in rugged terrain: fire channelling, International Journal of Wildland Fire, 21, 3, (282), (2012).
    • , Transport and evolution of the 2009 Australian Black Saturday bushfire smoke in the lower stratosphere observed by OSIRIS on Odin, Journal of Geophysical Research: Atmospheres, 116, D6, (2011).
    • , Coupled atmospheric and land surface dynamics over southeast Australia: a review, analysis and identification of future research priorities, International Journal of Climatology, 31, 12, (1758-1772), (2010).
    • , Severe convective storms initiated by intense wildfires: Numerical simulations of pyro‐convection and pyro‐tornadogenesis, Geophysical Research Letters, 36, 12, (2009).
    • , A new look at the role of fire-released moisture on the dynamics of atmospheric pyro-convection, International Journal of Wildland Fire, 18, 5, (554), (2009).
    • , An overview of mountain meteorological effects relevant to fire behaviour and bushfire risk, International Journal of Wildland Fire, 18, 7, (737), (2009).
    • , Tall clouds from small eruptions: the sensitivity of eruption height and fine ash content to tropospheric instability, Natural Hazards, 51, 2, (375), (2009).
    • , Atmospheric composition change – global and regional air quality, Atmospheric Environment, 10.1016/j.atmosenv.2009.08.021, 43, 33, (5268-5350), (2009).
    • , An aerosol boomerang: Rapid around‐the‐world transport of smoke from the December 2006 Australian forest fires observed from space, Journal of Geophysical Research: Atmospheres, 114, D21, (2009).
    • , Injection Heights of Biomass Burning Debris Estimated From WSR-88D Radar Observations, IEEE Transactions on Geoscience and Remote Sensing, 47, 8, (2599), (2009).
    • , Measurement of methanol emissions from Australian wildfires by ground‐based solar Fourier transform spectroscopy, Geophysical Research Letters, 35, 8, (2008).
    • , Stratospheric impact of the Chisholm pyrocumulonimbus eruption: 1. Earth‐viewing satellite perspective, Journal of Geophysical Research: Atmospheres, 113, D8, (2008).
    • , Stratospheric impact of the Chisholm pyrocumulonimbus eruption: 2. Vertical profile perspective, Journal of Geophysical Research: Atmospheres, 113, D8, (2008).
    • , A review of global stratospheric aerosol: Measurements, importance, life cycle, and local stratospheric aerosol, Atmospheric Research, 90, 2-4, (223), (2008).
    • , On the Meteorological Mechanisms Driving Postfire Flash Floods: A Case Study, Monthly Weather Review, 136, 5, (1778), (2008).
    • , Nuclear winter revisited with a modern climate model and current nuclear arsenals: Still catastrophic consequences, Journal of Geophysical Research: Atmospheres, 112, D13, (2007).
    • , Facing the Challenges of the International Airways Volcano Watch: The 2004/05 Eruptions of Manam, Papua New Guinea, Weather and Forecasting, 22, 1, (175), (2007).
    • , The origin of the savanna biome, Global Change Biology, 12, 11, (2023), (2006).
    • , 30-year lidar observations of the stratospheric aerosol layer state over Tomsk (Western Siberia, Russia), Atmospheric Chemistry and Physics, 10.5194/acp-17-3067-2017, 17, 4, (3067-3081), (2017).