Physical and Chemical Properties of Australian ... - UNSWorks

Physical and Chemical Properties of Australian Continental Aerosols by

Majed Radhi

A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE FACULTY OF SCIENCE THE UNIVERSITY OF NEW SOUTH WEALS SYDNEY- AUSTRALIA August 2010

ii

Acknowledgments First of all I owe my deepest gratitude to both my supervisor Associate Professor Michael Box and my co-supervisor Dr. Gail Box of the School of Physics, University of New South Wales, who provide me with many helpful suggestions, important advice and constant encouragement during this work and especially their patient proof reading of my thesis. I also wish to express my appreciation to both of them for sending me to various conferences and workshop (in Australia and overseas). I wish to express my cordial appreciation to Dr. Ross Mitchell of the Commonwealth Scientific and Industrial Research Organisation (CSIRO), for his valuable assistance in the Birdsville and Muloorina field campaigns, as well as providing valuable suggestions that improved the quality of this study. Without his help, these field experiments would not have been done. Sincere thanks are extended to Dr. David Cohen and Ed Stelcer of the Australian Nuclear Science and Technology Organisation (ANSTO), for their supervision of Ion Beam Analysis, in particular Ed who’s help was always offered cheerfully without hesitation. I would like to express my heartiest thanks to Dr. David French of the CSIRO for his numerous ideas and useful suggestions for QEMSCAN analysis. Without his help this analysis would not have been completed. My keen appreciation goes to Dr. Melita Keywood of the CSIRO for her supervision of the Ion Chromatography analysis, and providing numerous ideas and useful discussions.

iii

I would like to thank Professor Mike Gal for his constant encouragement; the computing unit, mainly David and Kristien for their help. Special thanks are due to Stephen Lo for his constant encouragement to finish this work. I would like to acknowledge Australian Research Council as my research was supported by Grant DP0451400, and AINSE which supported the Ion Beam Analysis under the Grants AINGRA08006 and 09057. I would like to show my gratitude to all my friends inside and outside the School of Physics for their support and engagements. Finally, my deepest gratitude goes to my family members (brothers and sisters) for their unflagging love, support and encouraging me to pursue this degree. I have no suitable words that can fully describe my mother, a never ending support. Special thanks go to my wife Rabab for helping me to concentrate to finish this work.

iv

Abstract This thesis has examined the characteristics of Australia’s two major continental aerosol types – mineral dust and biomass burning – using both remote sensing data, and the laboratory analysis of field samples. Australia is the dominant mineral dust source in the southern hemisphere, yet the physical, chemical and optical properties of Australian dust aerosol are presently poorly understood. In this study three field campaigns have been conducted within the Lake Eyre Basin in order to collect size resolved aerosol particles using a 12 stage impactor. The physical properties (size distribution) of these aerosols have been investigated by studying the mass of aerosol on each filter, which showed that the fine fraction made approximately 50% of the TSP during non dust storm conditions. The chemical properties of these aerosols have been investigated by using a range of different analysis methods. Ion Beam Analysis was used to determine the elemental composition of all filter samples and as expected, Si is a strongly abundant element in all size fractions during all sampling events at all sites. Ion Chromatography provides good information about secondary aerosol components, and about the soluble and insoluble components of elements such as sodium and chlorine. A new technique, QEMSCAN, has been used for the first time to determine the mineralogy of dust aerosols in the atmosphere. This showed that Australian dust particles are an internal mixture, and that much of the Fe is not in the form of iron oxide minerals. The scatter plot slopes of Fe against Al were found to be in the range 0.77 – 0.94 within LEB, which higher than the values have been reported for the North

v

Hemisphere sites (0.4-0.7) and higher than the global crustal average, confirming that Australian dusts are comparatively rich in Fe. The fraction of salt in transported dust was found to be in the range 0.2-1% during non dust storm days, and somewhat less during dust storms. Evidence of the contribution of marine biogenic aerosol to the total atmospheric aerosol over the LEB is provided by the presence of MSA, and also nssSO42Ground-based remote sensing measurements of aerosol optical depth have been used to investigate the optical and physical properties of aerosol and their variation with time for three regions across the Australian continent. The results show that the tropical north sites show seasonal cycles in AOD and α with fine fraction aerosol dominating during spring month due to biomass burning activities, while for the desert site the coarse fraction dominates during dust storm activity. A careful analysis of AOD data during a recent bushfire season showed clear evidence of smoke aging.

vi

Publications Journal Publications Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, D. D. Cohen, E. Stelcer, M. D. Keywood, “Optical, Physical and Chemical Characteristics of Australian Continental Aerosols: Result from a Field Experiment” Atmospheric Chemistry and Physics, 10(13): 5925-5942. Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, D. D. Cohen, E. Stelcer, M. D. Keywood, “Size-resolved Mass and Chemical Properties of Dust Aerosols from Australia’s Lake Eyre Basin” Atmospheric Environment, 44(29): 3519-3528. Radhi M, Box M. A., Box G. P., Gupta P, and Christopher S. A. “Evolution of the optical properties of biomass-burning aerosol during the 2003 southeast Australian bushfires”. Appl. Opt., 48, 1764-1773, 2009.

Refereed Conference Proceedings Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, D. D. Cohen, E. Stelcer, M. D. Keywood, D. French, (2007) “Physical Chemical and Optical Properties of Australian desert dust aerosols”, 14th IUAPPA world Congress, BrisbaneAustralia. 9–13 September 2007. Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, D. D. Cohen, E. Stelcer, M. D. Keywood, D., (2009) “Size-resolved Mass and Chemistry of Australian Desert Aerosol”, 16th AINSE Conference on Nuclear and Complementary Techniques of Analysis, Sydney –November 2009. T. Hallal, G. P. Box, M. Radhi, M.A. Box, D.D. Cohen and E. Stelcer. “Size – Resolved properties of Atmospheric Aerosol in Sydney and Regional NSW.” 14th IUAPPA world Congress, Brisbane-Australia. 9–13 September 2007.

vii

Conference and workshop presentations Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, D. D. Cohen, E. Stelcer, M. D. Keywood, D., (2009) “Size-resolved Mass and Chemistry of Australian Desert Aerosol”, 16th AINSE Conference on Nuclear and Complementary Techniques of Analysis, Sydney –November 2009 (poster). Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, D. D. Cohen, E. Stelcer, M. D. Keywood, D. French,. “Size-Resolved Mass, Chemistry and Mineralogy of Australian Desert Aerosol” 5th Australian New Zealand Aerosol Seminar, Auckland, New Zealand 22-24 July 2009 (oral). Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, M. D. Keywood, D. French, “Physical and chemical characterization of Australian desert aerosol.” 9th International

Conference

on

Southern

Hemisphere

Meteorology

and

Oceanography, Melbourne, Australia 9 -13 February 2009. (poster) Gail Box, Majed Radhi, Tomas Alarcon, Michael Box. “Size-resolved properties of Sydney urban aerosols.” 9th International Conference on Southern Hemisphere Meteorology and Oceanography Melbourne, Australia 9-13 February 2009. (oral) Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, M. D. Keywood, D. French, “Physical, chemical and optical properties of Australian desert dust a multiply analysis from several sites.” The third International workshop on mineral dust, Leipzig, Germany. 15-17 September 2008. (poster) Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, “Size-Resolved chemistry of Australian desert dust samples.” IGAC- The International Global Atmospheric Chemistry, Annecy- France, 7-12 September 2008. (poster)

viii

Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, D. D. Cohen, E. Stelcer, M. D. Keywood, D. French,. “Physical and Chemical properties of Australian desert dust.” The 4th Australian New Zealand Aerosol Seminar, ANSTO SydneyAustralia. 16-18 July, 2008. (oral) Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, B. W. Forgan “Climatology of Australian

continental

aerosol.”

15th

Australian

Meteorological

Oceanographic Society (AMOS). Geelong, Victoria, Australia.

and

29 January-

February 1 2008. (oral) Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, D. D. Cohen, E. Stelcer, M. D. Keywood, D. French, “Physical, Chemical and Optical Properties of Australian Desert Dust Aerosols.” 14th IUAPPA world Congress, Brisbane-Australia. 9–13 September 2007. (oral) T. Hallal, G. P. Box, M. Radhi, M.A. Box, D.D. Cohen and E. Stelcer. “Size – Resolved properties of Atmospheric Aerosol in Sydney and Regional NSW.” 14th IUAPPA world Congress, Brisbane-Australia. 9–13 September 2007. (oral) Radhi M, M. A. Box, G. P. Box, R. M. Mitchell, D. D. Cohen, M. D. Keywood, D. French, S. Campbell “Physical, Chemical and Optical Properties of Australian Desert Dust Aerosols. (ADDA).” Australian Aerosol Workshop 2007. CSIRO Aspendale, Victoria, Australia. 4-6 July 2007. (oral) Radhi M, M. A. Box, G. P. Box, “Characteristics of biomass burning aerosol in Wagga Wagga during the Southeast Australia bushfires of 2003.” 14th AMOS National Conference: Climate, weather and marine forecasting – Glenelg Sailing Club, Adelaide, SA, Australia, 5 to 8 February 2007. (oral) Radhi M, M. A. Box, G. P. Box, “Optical and physical properties of biomass burning aerosol over Wagga Wagga during Southeast Australia bushfires in summer

ix

2003.” International Laboratory for Air quality and Health. Queensland University of technology- Brisbane- Australia. 5-7 July 2006. Radhi M, M. A. Box, G. P. Box, B. Forgan. “Physical Properties of Aerosols at Wagga Wagga and Tennant Creek, Australia.” 13th AMOS National Conference: Climate, water and sustainability – Newcastle City Hall, Newcastle, NSW, Australia, 6 - 8 February 2006. (oral)

x

Table of Contents Acknowledgments

i

Abstract

iii

Publications

v

Introduction

1

1.1

Motivation

1

1.2

Atmospheric Aerosols

2

1.2.1

Aerosol Properties

2

1.2.2

Australian Aerosols

4

1.3

Aims and Objectives

7

1.4

Thesis outline

8

Measurements, Instruments, Methods and Data

9

2.1

Introduction

9

2.2

Aerosol Sampling

10

2.2.1

Instrumentation

10

2.2.2

Gravimetric Mass Distributions

13

2.2.3

Chemical Analysis

14

2.3

Aerosol Remote Sensing

16

2.3.1

Instruments

16

2.3.2

Langley Analysis

18

2.3.3 Inversion

21

2.4

Mineralogical Analysis

24

2.5

Computing optical properties

28

xi

Characterization of Aerosol from the Lake Eyre Basin 3.1

3.2

3.3

3.4

30

Introduction

30

3.1.2

30

Mineral Dust Aerosol

3.1.1 Lake Eyre Basin

32

Birdsville

35

3.2.1

Site Location

35

3.2.2

Aerosol Optical Properties

35

3.2.3

Aerosol Samples and Gravimetric Mass Distributions

45

3.2.4

Elemental Composition and Source Apportionment

49

3.2.5

Water-Soluble Ions

56

Muloorina

65

3.3.1

Site Location and Aerosol Samples

65

3.3.2


67

3.3.3


70

3.3.4

Water-Soluble Ions

77

Fowlers Gap Station

84

3.4.1

Site Location, Aerosol Samples and Weather Conditions

84

3.4.2


85

3.4.3


87

3.5

Mineralogy

95

3.6

Discussion

100

3.6.1

Physical Properties

100

3.6.2 Chemistry: Metals

101

3.6.3

Salt Entrainment

104

3.6.4 Secondary Components

106

xii

Aerosols from Savanna and Woodland Areas

107

4.1

Introduction

107

4.2

Northern Sites

110

4.2.1

110

4.3

4.4

Site Locations

4.2.2 Aerosol Optical Depth (AOD)

112

4.2.3

Angstrom Exponent (α)

113

4.2.4

Seasonal Statistics

119

4.2.5

Size Distributions

126

4.2.6 Chemical Properties for Darwin and Jabiru

135

4.2.7

139

Summary of the Northern Sites

Southeast Site (Wagga Wagga)

140

4.3.1

140


4.3.2 Aerosol Optical Properties during the 2003 Bushfires

144

4.3.3 Satellite Observation

147

4.3.4

Ångström Exponents

150

4.3.5

Study of Two Intense Smoke Periods

152

4.3.6

Summary of Wagga Wagga

162

Discussion and Conclusion

163

Conclusion

165

5.1

Summary

165

5.2

Future work

171

Glossary

173

References

175

xiii

List of Figures

Figure1.1: Satellite image of Australian dust storm

6

Figure 1.2: Australian Fire season

6

Figure 2.1: 12 stages Micro Orifice Uniform Deposition Impactor (MOUDI).

12

Figure 2.2: Nominal collection efficiency curves for MOUDI stages

12

Figure 2.3: Cimel Sun photometer.

20

Figure 2.4: A Middleton SPO2 sun photometer

20

Figure 2.5: QEMSCAN mineral image

26

Figure 3.1: Lake Eyre Basin map

34

Figure 3.2: Daily and monthly of AOD over Birdsville

40

Figure 3.3: Daily and monthly of α over Birdsville

40

Figure 3.4: Scatter plot of daily means of α vs. AOD over Birdsville

41

Figure 3.5: Seasonal frequency distributions of daily mean AOD, α and Seasonal scatter plots of daily means of α vs. AOD over Birdsville

42

Figure 3.6: Volume size distributions over Birdsville

44

Figure 3.7: Size resolved mass concentrations, Birdsville

48

Figure 3.8: Size-resolved mass ratios of some elements to Si for all Birdsville samples.

51

Figure 3.9: Scatter plots of some elements vs. Si (Birdsville).

52

Figure 3.10: Back-trajectories during sampling days at Birdsville.

54

Figure 3.11: Size-resolved concentrations of major water-soluble ionic species during sampling periods C and E (Birdsville).

58

Figure 3.12: Back-trajectories during periods A, E, DSN and DSS at Muloorina. 66

xiv

Figure 3.13: Size-resolved mass concentrations for all samples days (Muloorina). 69 Figure 3.14: Size-resolved mass ratios of some elements to Si for all samples.

71

Figure 3.15: Scatter plots of some elements vs. Si (Muloorina).

72

Figure 3.16: Size-resolved concentrations of major water-soluble ionic species

78

during sampling periods A and E (Muloorina). Figure 3.17: Size resolved mass concentrations sampled at FGS.

86

Figure 3.18: Size-Resolved mass ratios of elements to Si (FGS).

88

Figure 3.19: Scatter plots of some elements vs. Si (FGS).

89

Figure 3.20: Back-trajectories during the sampling days at FGS.

90

Figure 3.21: Images of minerals particles analysed by QEMSCAN technique.

98

Figure 4.1: Australian map showing locations of radiometer sites and

111

mineral dust sites. Figure 4.2: Daily mean values of AOD at 500 nm, over northern sites.

114

Figure 4.3: Monthly mean values of AOD at 500 nm, over northern sites.

115

Figure 4.4: Daily mean values of α, over northern sites.

116

Figure 4.5: Monthly mean values of α, over northern sites.

117

Figure 4.6: Frequency distribution of daily means of AOD over a) Lake Argyle. b) Jabiru. c) Darwin. d) Tennant Creek.

123

Figure 4.7: Frequency distribution of daily means of α over a) Lake Argyle. b) Jabiru. c) Darwin. d) Tennant Creek.

124

Figure 4.8: Scatter plot of daily mean of α against AOD over a) Lake Argyle. b) Jabiru. c) Darwin. d) Tennant Creek.

125

Figure 4.9: Monthly mean of volume size distribution over a) Lake Argyle. b) Jabiru. c) Darwin. d) Tennant Creek. Figure 4.10: Daily mean of volume size distribution over Lake Argyle

128-129

xv

and Jabiru for selected days.

132-133

Figure 4.11: Air mass back trajectory analysis over Lake Argyle and Jabiru for selected days

134

Figure 4.12: Daily mean values of AOD at 500nm, from March 2001 to December 2004, over Wagga Wagga.

142

Figure 4.13: Seasonal frequency distribution of daily means of AOD over Wagga Wagga.

142

Figure 4.14: Daily mean values of α412-778, from March 2001 to December 2004, over Wagga Wagga. Figure 4.15: Seasonal frequency distribution of daily means of α412-778.

143 143

Figure 4.16: Plume of smoke travelling to both the southeast and northwest on 25 January 2003. (Image from Aqua at 03:35 UTS)

146

Figure 4.17: Hourly mean values of AOD over Wagga Wagga during summer 2003 southeastern Australia bushfire.

146

Figure 4.18: MODIS-Terra image on January 24, 2003, showing smoke plumes over Wagga Wagga with Aerosol optical depth retrieval for this smoke plumes.

148

Figure 4.19: Correlation between ground-based and satellite measurements of AOD over Wagag Wagga during bushfire events Figure 4.20: Scatter plot of hourly means of AOD vs. both αS and αL.

149 151

Figure 4.21: Hourly mean values of wind direction from 9-13th of January 2003 over Wagga Wagga. Figure 4.22: Back trajectory analysis for 24 hr from 9 to 13th January 2003.

155 155

xvi

Figure 4.23: Scatter plot of AODH vs. both αS and αL , for 9th, 10th, 11th and 12th January 2003 over Wagga Wagga.

156

Figure 4.24: Particle volume-weighted size distributions from 7th-13th January 2003 over Wagga Wagga.

156

Figure 4.25: Scatter plots of τH vs. both αS and αL on January 25 over Wagga Wagga: a) whole day; b) morning period; c) afternoon period.

156

Figure 4.26: Wind direction during 25th January 2003, over Wagga Wagga.

160

Figure 4.27: Hourly particle volume-weighted distribution during 26 January 2003 over Wagga Wagga: a) morning and b) afternoon.

161

xvii

List of Tables

Table 1.1: Annual source strength for present day emissions of aerosol precursors (Tg N, S or C /year).

3

Table 3.1: Times of the sample collections at Birdsville.

46

Table 3.2: Mass fractions at Birdsville.

46

Table 3.3: Ratio of elements to Si at Birdsville.

50

Table 3.4: Ion Chromatographic analysis statistic at Birdsville site.

57

Table 3.5: Timings of sample collection and TSP mass concentrations at Muloorina.

65

Table 3.6: Mass fractions at Muloorina.

68

Table 3.7: Ratios of elements to Si at Muloorina.

70

Table 3.8: Ion Chromatographic analysis statistic at Muloorina site.

77

Table 3.9: Time of sample collection at FGS.

85

Table 3.10: Mass fractions at Muloorina.

86

Table 3.11: Ratios of elements to Si at FGS site.

87

Table 3.12: Filters sent for QEMSCAN analysis.

95

Table 3.13: The mass percentages of minerals.

97

Table 3.14: The slope and R2 values from the scatter plots of Al, Fe, Ca, Mn, Ti and K against Si, at the three sites.

101

Table 3.15: Ratio of Fe/Al for different locations around the world.

102

Table 4.1: Monthly statistics of aerosol optical depth for Northern sites

118

Table 4.2: Monthly statistics of Angstrom exponent for Northern sites

118

xviii

Table 4.3: Seasonal statistics of aerosol optical depth for Northern sites.

122

Table 4.4: Seasonal statistics of Angstrom exponent for Northern sites.

122

Table 4.5: Elemental concentration for Darwin and Jabiru.

138

Table 4.6: Summary of statistics of aerosol optical properties over Wagga Wagga

145

Table 4.7: Hourly mean values of τ500 and αs during 25th January 2003 over Wagga Wagga

160

2161Chapter 1 Introduction

1

Chapter 1 Introduction

1.1

Motivation

Aerosol particles are known to play an important role in the atmosphere-land-ocean system, as they may alter the Earth’s radiation budget (Charlson et al., 1992; Tegen et al., 1996; Sokolik et al., 2001), change the physical and radiative properties of clouds (Ramanathan et al., 2001), affect atmospheric chemistry (Li and Shao, 2009), modify ocean biogeochemistry (Prospero et al., 1989; Duce et al., 1991), and have an impact on human health (Brook et al., 2002; Englert, 2004). Hence aerosol characterization, such as particle size, shape, and chemistry, is important to understand these processes. The radiative forcing by most aerosol types is reasonably well estimated, but the uncertainties remain large, especially for mineral dust (IPCC, 2007). Thus continuous aerosol monitoring is important to improve our understanding of such impacts. The Australian continental atmosphere is dominated by both biomass burning and mineral dust aerosols, with contributions from marine aerosol when the winds are onshore, and pollution aerosol due to human/industrial activities. Thus ongoing monitoring and characterization of the physical and optical properties of this aerosol mix is important to improve our understanding of its role in Australian climate (e.g. Rotstayn et al., 2009). The work of this thesis is thus aimed at increasing our understanding of these characteristics. More comprehensive overview and literature background, with emphasis on current gaps in our knowledge, will be given in Chapters 3 (for mineral dust) and 4 (for biomass burning).


1.2

Atmospheric Aerosols

1.2.1

Aerosol Properties

2

Aerosol particles can be injected directly into the atmosphere (primary aerosols) from natural and anthropogenic sources, such as dust particles from desert areas, volcanic eruptions, sea salt ejected from the ocean, soot from bushfires, and industrial processes. With the exception of soot, they are mostly one to several µm in size. Secondary aerosols are aerosols created in the atmosphere after chemical and physical processes whereby low volatility compounds condense, either to form new particles, or on the surface of existing particles. Examples include methanesulfonate and non sea salt sulfate, as the end products of the oxidation of dimethylsulfide emitted from the ocean (Hatakeyama and Akimoto, 1983; Bates et al., 1992), and gas-to-particle conversion of industrial waste gases such as SO2 and NO2. Different aerosols emitted to the atmosphere have different chemical properties (see Table 1.1 from IPCC, 2001). Processing in the atmosphere, such as condensation of volatile gases on existing particles, and coagulation of particles either directly or within cloud droplets which subsequently evaporate, can further complicate these properties. The size of aerosol particles is generally specified as the radius (assuming a spherical shape) or ‘equivalent radius’. Aerosols are typically dispersed into three size ranges: nuclei or ultrafine mode (particles smaller than ~0.1 µm radius); fine or accumulation mode (0.1-1.0 µm radius); and coarse mode (particles > 1.0 µm radius). Because coarse and fine mode aerosols often come from different sources (e.g. primary versus secondary sources/mechanisms), their chemistry is likely to vary.


3

Table 1.1 Annual source strength for present day emissions of aerosol precursors (Tg N, S or C /year). Northern Hemisphere

Southern Hemisphere

Globala

NOx (as TgN/yr)

32

9

41

(see also Chapter 4).

Fossil fuel (1985)

20

1.1

21

Benkovitz et al. (1996)

Aircraft (1992)

0.54

0.04

0.58

0.4-0.9

Penner et al. (1999b); Daggett et al. (1999)

Biomass burning (ca. 1990)

3.3

3.1

6.4

2-12

Liousse et al. (1996); Atherton (1996)

Soils (ca. 1990)

3.5

2.0

5.5

3-12

Yienger and Levy (1995)

Agricultural soils

2.2

0-4


Natural soils

3.2

3-8


Range

Source

Lightning

4.4

2.6

7.0

2-12

Price et al. (1997); Lawrence et al. (1995)

NH3 (as TgN/yr)

41

13

54

40-70

Bouwman et al. (1997)

Domestic animals (1990)

18

4.1

21.6

10-30


Agriculture (1990)

12

1.1

12.6

6-18


Human (1990)

2.3

0.3

2.6

1.3-3.9


Biomass burning (1990)

3.5

2.2

5.7

3-8


Fossil fuel and industry (1990)

0.29

0.01

0.3

0.1-0.5


Natural soils (1990)

1.4

1.1

2.4

1-10


Wild animals (1990)

0.10

0.02

0.1

0-1


Oceans

3.6

4.5

8.2

3-16


SO2 (as TgS/yr)

76

12

88

67-130

Fossil fuel and industry (1985)

68

8

76

60-100

Benkovitz et al. (1996)

Aircraft (1992)

0.06

0.004

0.06

0.031.0

Penner et al. (1998a); Penner et al. (1999b); Fahey et al. (1999)

Biomass burning (ca. 1990)

1.2

1.0

2.2

1-6

Spiro et al. (1992)

Volcanoes

6.3

3.0

9.3

6-20

Andres and Kasgnoc (1998) (incl. H2S)

DMS or H2S (as TgS/yr)

11.6

13.4

25.0

12-42

Oceans

11

13

24

13-36

Kettle and Andreae (2000)

Land biota and soils

0.6

0.4

1.0

0.4-5.6

Bates et al. (1992); Andreae and Jaeschke (1992)

Volatile organic emissions (as TgC/yr)

171

65

236

100560

Anthropogenic (1985)

104

5

109

60-160

Piccot et al. (1992)

Terpenes (1990)

67

60

127

40-400

Guenther et al. (1995)


1.2.2

4

Australian Aerosols

1.2.2.1 Mineral dust aerosol A large part of Australia is desert or semi-arid area, and is the greatest contributor of mineral aerosol in the Southern Hemisphere with 5% contribution to the total annual global dust emission (Tanaka and Chiba, 2006). Mineral dust aerosol is a primary, coarse mode aerosol which is essentially natural, although human activities which enhance desertification may increase emissions. Australian desert aerosol is highly absorbing in the blue wavelength region (Qin and Mitchell, 2009), which is reflected in the fact that Australian’s deserts are reddish, in contrast to the Sahara, for example, which is more yellow. This difference is a reflection of mineralogy, particularly in relation to iron oxides such as hematite. These authors also assert that “a systematic characterization of the composition and optical properties of Australian dust aerosol is currently lacking”. Australian dust aerosol is the main terrestrial source to the open and remote ocean (Hesse and McTainsh, 2003) (Figure 1.1: image of the last dust storm over Sydney). More overview and literature background on Australian mineral dust aerosol, including the gaps in our knowledge, will be given in Chapter 3.

1.2.2.2 Biomass burning aerosol The second major aerosol type over the Australian continent is biomass burning. This aerosol type, which may be either natural or anthropogenic, consists of both fine mode soot particles, and biogenic gases which may condense to form secondary aerosol. Australia is second in burning area after Sub-Saharan Africa with 34 x 106 ha yr-1, contributing approximately 8% of global carbon emissions due to grassland,


5

woodland and forest burning (Ito and Penner, 2004). Figure 1.2 shows the seasonal distribution of the fires that occur in Australia. The savanna biomass burning in the tropical north of Australia is largest source of biomass burning aerosol in Australia, and occurs during the dry season of winter (June–August) and spring (September– November). In the southeast of the continent fires are more common during summer and early autumn months. More overview and literature background on Australian biomass burning aerosol will be given in Chapter 4.

1.2.2.3 Marine aerosols Sea salt (a primary aerosol) is the largest natural aerosol produced by the evaporation of sea spray droplets over the ocean. The marine atmosphere is also influenced by biogenic emission of dimethylsulfide (DMS), the oxidation of which produces methanesulfonate (MSA) and non sea salt sulfate aerosol (Bates et al., 1987; Leck and Rodhe, 1991), both of which are natural, secondary aerosols. These marine aerosols become land aerosol when the wind is onshore.

1.2.2.4 Pollution aerosols Anthropogenic aerosols from sources such as industry, transport, power stations and agriculture are the main source of outdoor pollution in Australia. These kinds of emissions produce fine mode secondary particles which may affect human health (Pope, 2000; Sunyer, 2001). Keywood et al. (1999) found that on average the fraction of particles less than 2.5μm in diameter (PM2.5) made up 60% of aerosol (by mass) collected in six Australian cities, and the PM1 fraction made up 72% of this, which indicates that fine fraction particles are dominant in Australian urban areas.


6

Mt. Isa in western Queensland is the site of a very large copper, lead and zinc mine, and is one of the largest point sources of SO2 in the world.

a)

b)

Figure 1.1: a) Dust storm over eastern Australia; Satellite: Terra – Pixel size: 2km, on 23/9/2009 at 00.05UTC. b) Australian dust northwest of New Zealand; Satellite Terra – Pixel size: 2km, on 23/9/2009 at 23.10 UTC.

Figure 1.2: Australian fire season.


1.3

7

Aims and Objectives

The aim of this thesis is to improve our understanding of the role of Australian continental aerosols – mineral dust and biomass burning – in the Earth System. More specifically, the aim is to determine their physical, chemical and optical properties, in order to better understand their interaction with electromagnetic radiation, over a range of wavelengths. In order to achieve this, we need to know the chemical composition, including the mixing state, of these aerosol types, as functions of particle size. It is also desirable to obtain this information, along with the size distribution, and its variability, under both ‘background’ and elevated (e.g. dust storm; bushfire) conditions. Shape information may or may not be important. If the particles are simple enough – either chemically ‘pure’ or uniformly mixed, and of close to spherical shape – then the computation of optical properties (extinction and absorption coefficients, phase function) is straightforward, using Mie theory. If this is not the case – e.g. agglomerated particles of highly irregular shape – then far more sophisticated algorithms (such as spheroidal, discrete dipole, or T-matrix) must be used. This would add an entirely new dimension to the task. The work of this thesis has two thrusts: mineral dust; and biomass burning aerosols. The first has involved the investigation of size-resolved field samples of Australian mineral dust aerosol, supported by an investigation of the optical and physical properties of dust aerosol at Birdsville using three years of remote sensing data. The second investigated the optical and physical properties of biomass burning aerosol, based primarily on remote sensing data, in both the tropical north, and the southeast of Australia.


8

Three different field campaigns were carried out in the Lake Eyre Basin (LEB), the largest dust source region on the Australian continent, to collect size-resolved dust samples. Elemental composition of these samples was determined using acceleratorbased Ion Beam Analysis (IBA). The IBA results have been used to obtain the ratios of elements to the most abundant element in the Australian desert atmosphere to build a signature for Australian dust aerosol. After IBA determination, some of the sample sets have been chosen for Ion Chromatography (IC) analysis to determine the concentration of soluble ions. (Information about the instrumentation and methods of analysis are provided in Chapter 2.) A few of the filters were also selected for a new form of mineralogical analysis, QEMSCAN. This technique, which only became available for my samples towards the end of my work, provides unprecedented new information on the size, shape, and internal mineralogy (including agglomeration state), of thousands of individual particles. Analysis of these results will continue into the future. The optical and physical properties of the tropical north biomass burning aerosol have been investigated at four sites – Lake Argyle, Jabiru, Darwin and Tennant Creek – by using aerosol optical depth (AOD) data from ground–based instruments. South-eastern Australia is known to be strongly affected by the El Niño/La Niña cycle, which has led to several major droughts in the past century. These conditions have been responsible for a number of disastrous bushfire seasons involving considerable property loss as well as the loss of many lives. February 2009 and January–February 2003 were the most recent examples. The optical and physical properties of the January–February 2003 bushfire aerosol have been investigated using AOD data obtained from a sun photometer at Wagga Wagga.


1.4

9

Thesis outline

This thesis is constructed in the following order. Chapter 2 provides a description of instruments used in this work and the methods of analysis used to determine optical properties, aerosol mass concentrations, chemical composition and mineralogy. Where appropriate, an indication of uncertainty levels is also included. Chapter 3 focuses on characterizing Australian mineral dust aerosol through three field campaigns to different sites within the Lake Eyre Basin, in the heart of the Australian arid and semi-arid region. Three years of remote sensing data have also been analysed. Chapter 4 focuses on the properties of biomass burning aerosol from savanna and woodland areas in Australia, using aerosol optical depth data from five stations to study the optical and physical properties of aerosol over northern and southeast areas. Chapter 5 summarizes the main findings and conclusions of this work, including what still remains to be done, and provides directions for future research.

Chapter 2 Measurements, Instruments, Methods and Data

10


2.1

Introduction

Because of the complexities of aerosol particles, as reflected in the great variations in their physical and chemical properties, no one technique can be said to be capable of supplying all the information which might be needed. Direct sampling – the collection and subsequent analysis of airborne particulates – is the only approach capable of providing detailed information on aerosol properties at a given place and time. However, it is expensive and time-consuming, and cannot be used except in localized or campaign modes. Remote sensing, either from ground or space, can provide information on the gross (averaged) properties of an aerosol population, which may or may not hide vital information. In this thesis, direct sampling has been used to characterize mineral dust aerosol, as its chemistry is sufficiently variable that such approaches are essential. Remote sensing observations have been used to supplement this information, and also to study the longer term variability of aerosols from several locations.


2.2

11

Aerosol Sampling

Collecting aerosol samples on filters provides a good opportunity to study the mass concentrations, physical properties (shape and size), chemical composition and mineralogy of the particulates suspended in the atmosphere. This section will describe the instrument and different analysis methods used to investigate the physical and chemical properties of Australian aerosol collected during the course of the thesis work.

2.2.1

Instrumentation

A 12 stage Micro Orifice Uniform Deposition Impactor (MOUDI) (Marple et al., 1991) was used to collect size-segregated aerosol particles (Figure 2.1). At each stage jets of particle-laden air are directed at a flat impaction plate. Large particles having significant inertia are collected on the impaction plate whilst smaller particles below a threshold follow the airflow out of the impaction area. The cut-off between the size of aerosol particles collected and those that follow the airflow is very sharp and is determined by their aerodynamic diameter. Particles in particular size ranges are collected by passing the aerosol through multiple stages of impaction plates with each following stage collecting particles smaller than the previous one. The inlet to the MOUDI consisted of two bowls with a 2 cm gap between them, allowing air to be sampled equally from all directions, while protecting against rain. The stage cuts are at 18.0, 10.0, 5.6, 3.2, 1.8, 1.0, 0.56, 0.32, 0.18, 0.1 and 0.056 μm aerodynamic diameters, plus an after filter (< 0.056 μm). The cut-points and collection efficiencies (Figure 2.2) of each stage have been specifically calibrated for this instrument by the manufacturer, MSP Inc., with flow rate 30 l/min. (The


12

relationship between the aerodynamic diameter, density, and the effective diameter is: ( Dae= d(ρ/ρae)1/2 ≡ d ρ1/2). The density of dust particles depends on the chemical composition, and may be quite variable. Thus there was no correction made for the aerodynamics diameter of the dust particles collected by MOUDI for the density. In the case of mineral dust particles with their relatively high density, this is likely to be an overestimate of the true or effective diameter by about 30%.) The multiple stages allow for later identification of any compositional differences between particles of different size as the aerosols in many locations are known to be chemically variable, with significant differences between fine and coarse modes. For convenience we define coarse mode aerosols as particles with (aerodynamic) diameters greater than 1.0 μm, and fine mode aerosols as those with smaller diameters. The collection substrates used on the first 11 stages were polycarbonate Isopore filters 47mm in diameter with 0.8µm pore size. The final stage substrate was a Teflon-backed Fluoropore filter 47mm in diameter with 1µm pore size.


Figure 2.1: 12 stages Micro Orifice Uniform Deposition Impactor (MOUDI).

Figure 2.2: Nominal collection efficiency curves for MOUDI stages.

13


2.2.2

14


The sum of the masses on all MOUDI filters can be considered a good measurement of Total Suspended Particulates in the atmosphere (TSP) for one sampling period because all particles for different sizes are collected at the one time between the inlet and the after filter stages. The substrates were weighed before and after sampling at the Institute for Environmental Research (IER) of the Australian Nuclear Science and Technology Organization (ANSTO) using a Mettler Toledo MX5 balance with repeatability o

0.0008 mg at gross load. Temperature during the weighing process was 22 ± 1.5 C, and the humidity was 50 ± 5%. The mass and standard deviation on each substrate was calculated according to the following procedure. The substrates were weighed three times before sampling and the average and standard deviation calculated to get the mass before, Mb, with standard deviation, SDb. After the substrates were exposed this process was repeated to obtain the mass after sampling, Ma, with standard deviation, SDa. Then the final aerosol mass on each substrate, Mf, is

M f = Ma − Mb

(2.1)

and the final error is obtained by adding errors in quadrature SD f = ( SD a + SDb ) 2

(2.2)

The TSP for one sampling period is TSP = ∑ M f

(2.3)

The error in TSP is obtained adding the SDf in quadrature SDTSP =

∑ (SD )

2

f

(2.4)


2.2.3

15

Chemical Analysis

2.2.3.1 Ion Beam Analysis The filter samples were analysed at ANSTO by Ion Beam Analysis techniques (IBA) using the 2.6MeV van de Graaff accelerator (Cohen, 1993, 1998; Cohen et al., 1996). Typical beam currents of 10 - 15 nA over a diameter of 8 mm were used, the currents kept below 0.2 nA/mm2 to avoid damage to the filter and reduce elemental loss. Run times under these conditions were around 5 minutes per filter for full analysis. IBA analysis is a reliable, fast and non-destructive technique for analysing particulate elemental composition with sufficient sensitivity: no sample preparation is required and it is ideal when large numbers of samples are involved. The techniques used simultaneously throughout this study were: Proton Induced XRay Emission (PIXE), used to determine Al, Si, P, S, Cl, K, Ca, Ti, V, Cr, Mn, Fe, Co, Cu, Ni, Zn, Br, and Pb, from a few ngm-3 upwards; and Proton Induced Gamma ray Emission (PIGE), used for the analysis of light elements, such as Li, B, F, Na, Mg, Al, and Si. Since PIGE originates from a nuclear process rather than an atomic reaction, it is less sensitive for aerosol filter analysis than PIXE, and can only detect concentrations above 100 ngm-3. Long term precision for PIXE is ±1.6% for major elements, while accuracy ranges from 5% to 10% depending on elemental composition. Accuracy for PIGE is 5% to 15% and precision is 5%. Cohen et al. (2002) discuss in more detail the minimum detection limits, measurement accuracy, and precision. Statistical counting errors depend on elemental concentrations. Errors are then added in quadrature. For major elements like Si, Fe etc the final result is likely to be 12%, while for minor elements like Ni this could approach 35%.


16

As the spatial distribution of sample on each filter (stage) varied, and the ion beam only sampled the central 10% of the filter, the IBA results have been used primarily to obtain the ratio of elements to the most abundant element in the samples. Elemental carbon (EC) was also measured for Darwin and Jabiru using the Laser Integrating Plate Method. Light from a 4 mW He-Ne laser is diffused and collimated to give a uniform beam that passes through the loaded filter. The light scattered by particles is angularly integrated by a Lambert scattering opal glass plate, so that the change in light transmission is attributed only to the particle absorption. Aerosol particles are on the surface away from the light source. The transmitted signal intensity is measured using a photodiode detector. Each filter is measured before and after sample collection because optical properties of the filters vary noticeably. The absorption is described by the Lambert Beer law and EC concentration is given by:

EC =

A ⎛ I0 ⎞ ln⎜ ⎟ εV ⎝ I ⎠

(2.5)

where Io is the unexposed intensity, I is the exposed transmission intensity, ε is the absorption coefficient, and is 7 m2g-1 for the ANSTO system, A is the exposed area of the filter, and V is the volume of air sampled. More details of the method used are given in Taha et al. (2007).

2.2.3.2 Soluble Ion Analysis After IBA, some samples were analysed for the concentration of soluble ions at CSIRO by suppressed ion chromatography (IC). The filters were extracted in 5 ml of MQ-grade water (18 mΩ de-ionized water). The sample is then preserved using 1% chloroform. Anion and cation concentrations are determined by using a Dionex DX500 gradient ion chromatograph. Anions are determined using an AS11 column


17

and an ASRS ultra-suppressor and a gradient eluent of sodium hydroxide. Cations are determined using a CS12 column and a CSRS ultra-suppressor and a methanesulfonate acid eluent. The uncertainty of the soluble ion measurements is 6.6% at the 95% confidence level.

2.3

Aerosol Remote Sensing

Aerosol Optical Depth (AOD) describes the attenuation of incoming sunlight by a column of aerosol. Hence AOD is a measure of aerosol loading in the atmosphere. AOD typically decreases with increasing wavelength and is much smaller for longwave radiation than for shortwave. Values of AOD vary widely depending on the number and size of the aerosols. There are many methods of measuring the AOD: satellite-based instruments, aircraft and ground–based instruments. In this study ground–based measurements, and some satellite data, are used.

2.3.1

Instruments

2.3.1.1 Cimel Sun photometers Lake Argyle, Jabiru and Birdsville AOD data were measured by Cimel Sun photometers (Figure 2.3) which form part of CSIRO’s Aerosol Ground Station Network (AGSNet) that is affiliated with NASA’s Aerosol Robotic Network (AERONET). Instrument calibration and the generation of AOD and aerosol microphysical data (phase function, size distribution, refractive index and derivative quantities such as single scattering albedo and symmetry factor) were performed as part of the standard AERONET processing stream. These instruments and data


18

products are described in detail by Holben et al. (1998) and Dubovik and King (2000). Nevertheless, a short description will be given here. The instrument makes direct sun measurements and sky-scans automatically with a 1.2o full field of view at least every 15 minutes at the nominal wavelengths of 340, 380, 440, 500, 675, 870, 940 and 1020 nm. A full measurement set, including three solar extinction measurements at each wavelength, takes 1 minute. These measurements are used to compute aerosol optical depth at each wavelength apart from the 940 nm channel, which is used to retrieve water vapour. Eck et al. (1999) found that the uncertainties in computed the AOD are in the range 0.01-0.02 for field instruments. The AOD data from this instrument are cloud screened following the methodology of Smirnov et al. (2000). The temporal variability of AOD measurements is the principal filter used for cloud screening. It was found that the uncertainty in daily average AOD at 500 nm varied from month to month, with approximately 90% of the daily averages having standard deviations less than 0.05 for all three sites, but this reduced to around 80% during the biomass burning and dust activity seasons (spring months).

2.3.1.2 SPO2 sun photometer The second instrument was a Middleton SPO2 sun photometer (Figure 2.4) mounted on an automatic solar tracker with active sun tracking operated by the Australian Bureau of Meteorology (BoM) which was used to obtain the Darwin, Tennant Creek and Wagga Wagga AOD measurements. The tracking sensor ensures sun alignment within 0.02° when the sun is seen with four 25 mm diameter interference filters. The filters have 10 nm full-width-half-maximum transmissions, and 4 Hamamatsu silicon photodiodes centred at the nominal wavelengths of 412, 500, 610 and 778 nm.


19

The SPO2 sun photometer measures direct solar radiation. AOD is calculated based on the Beer-Lambert-Bouguer law. The error analysis for this instrument shows that the uncertainty in AOD for all 4 wavelengths at the 95% confidence level was less than 0.010 (Mitchell and Forgan, 2003). Intercomparison of BoM and CSIRO sun photometers at sites in the Australian outback shows that each is capable of measuring total optical depth to 0.01 at the 95% uncertainty level (Mitchell and Forgan, 2003). The minute by minute AOD data is received from BoM and the uncertainty (standard deviation) in daily average values has been found to be less than 0.05 at Wagga Wagga, on approximately 95% of days during all months except January and February (84% and 90% respectively) due to the influence of the 2003 bushfires. Darwin and Tennant Creek data show that 90% of days had uncertainties less than 0.05, although this percentage was lower during the biomass burning (late winter and spring months) and dust activity months.

2.3.2 Langley Analysis As a beam of sunlight travels through the atmosphere it is attenuated due to scattering or absorption by aerosols, molecules and gases in the atmosphere. The Beer-Lambert-Bouguer law gives the new intensity: I λ = I 0,λ R −2 exp(−τ λ m)

(2.6)

where λ is the wavelength, I0,λ is the intensity at the top of the atmosphere, Iλ is the intensity reaching earth’s surface, R is the Earth-sun distance in astronomical units, τλ is the atmospheric optical thickness and m is the relative air mass factor, where m = sec(θ), θ being the solar zenith angle.


20

The total optical thickness of the atmosphere is the sum of optical thicknesses of the atmospheric components. It can be expressed as:

τ = τ R + τ gas + τ aer

(2.7)

τR is the molecular (Rayleigh) scattering optical thickness, which depends on pressure and wavelength (both are well known); τgas is the gaseous absorption optical thickness due to O3 and H2O; and τaer is the aerosol optical thickness. Substituting equation (2.7) in equation (2.6) gives I λ = I 0,λ R −2 exp[− m(τ R + τ gas + τ aer )]

(2.8)

Equation (2.7) can be rewritten in logarithmic form as ln I λ = ln I 0,λ R −2 exp[ −m(τ R + τ gas + τ aer )]

(2.9)

Plotting ln Iλ against m for a series of measurements (typically for 2 < m < 6) gives a Langley plot. This should be a straight line if atmospheric conditions remain constant. The slope is the total optical thickness of the atmosphere, and the intercept is the logarithm of the measured intensity at the top of the atmosphere, which can be used for self calibration of the system. The relative air mass m is dependent on the solar zenith angle, and for its determination it is necessary to know the time and location of an observation. Subtracting τR , and τgas from total τ gives τaer .


Figure 2.2: Cimel Sun photometer.

Figure 2.3: A Middleton SPO2 sun photometer

21


2.3.3

22

Inversion

Remote sensing data, whether extinction as a function of wavelength, scattering as a function of angle, or a combination, contains information relating to the size distribution of aerosol particles in the column of atmosphere above the instrument. There are many techniques available to make use of this information.

2.3.3.1 Ångström Exponent, α The spectral dependence of aerosol optical thickness can often be reasonably well represented by the Angstrom exponent, which can be obtained in several ways. The spectral dependence of aerosol optical thickness may be described by the Ångström law:

τ aer = β λ−α

(2.10)

Here τaer is the aerosol optical thickness at wavelength λ, β is the turbidity parameter and α is the Ångström exponent. We may obtain α by taking the natural logarithm of both sides of Equation (2.10). A plot of ln τaer vs ln λ should give a straight line with an intercept ln β and α is the negative of the slope; alternatively,

α = −d ln τ aer / d ln λ

(2.11)

Ångström exponent values are used to give a qualitative idea of aerosol size distribution, with smaller values corresponding to large particle sizes (in general), and vice versa (Kambezidis and Kaskaoutis, 2008). The accuracy in the Ångström exponent data is dependent on the choice of channels (especially the wavelength range) and the type of aerosols (Eck et al., 1999). In this study the Ångström exponent computed from the 440-870 nm wavelengths has been chosen from the


23

AERONET data base as these are highly accurate channels (Eck et al., 1999; Holben et al., 1998). For BoM data Eq. (2.11) have been used to compute the Ångström exponent from the 412, 500, 610 and 778 nm wavelengths.

2.3.3.2 Size Distribution Retrieval Multispectral extinction (optical thickness) measurements relate the size distribution of the aerosol particles in the atmosphere to the optical thickness through the following integral equation ∞

τ aer (λ ) = ∫ πr 2 Q(r , λ , m)n(r )dr

(2.12)

0

where τ aer (λ ) is the optical depth at wavelength λ, Q is the Mie extinction efficiency factor, m is the complex refractive index, r is the particle radius and n(r) is the number of particles per unit area in a vertical column of atmosphere, with radii in the size range r to r + dr. A number of different inversion techniques have been employed to extract the particle size distribution from such measurements. In this study the volume-weighted size distributions for BoM stations (Darwin, Tennant Creek and Wagga Wagga) were obtained by inverting the AOD data using the constrained linear inversion technique developed by King et al. (1978) which uses optical thickness data only. Note that no error levels have been included. The major contributor to retrieval errors would be the limited spectral range of the AOD measurements. There is a direct connection between the wavelength range of the measurements, and the radius range for which we can expect reliable retrievals (e.g King et al., 1978; Viera and Box, 1987). For the BoM instruments this range is approximately 0.1 to 2.0 µm – the accumulation (fine) mode. The inversion code may be able to suggest the presence of a coarse mode, but can say nothing about its


24

structure. An information content analysis (Box et al., 1996) suggested that this data contains one to two pieces of information, which will be taken here to be the peak radius (assuming this lies in the accumulation mode), and possibly its width. The focus here will be on the peak radius. Over Lake Argyle, Jabiru and Birdsville the size distribution has been obtained from the AERONET web site from retrieval version 2. The research and development of this inversion are described in the papers by (Dubovik and King 2000; Dubovik et al., 2000; Dubovik et al., 2002a; Dubovik et al., 2002b; Dubovik et al., 2006a; Sinyuk et al., 2006). By combining both multispectral extinction measurements with scattered light measurements (almucantar scan), these instruments are capable of providing information on course mode particles as well as accumulation mode. Thus we expect these retrievals to be more realistic than our own inversions. Spherical and non-spherical aerosol particles are assumed in the retrievals. Dubovik et al. (2000) estimated the error in the value of the volume size distribution (i.e. dv/d(lnr)) which is not less than one tenth of the maximum.


2.4

25

Mineralogical Analysis

Selected filters from our sample collection trips were sent for mineralogical analysis using QEMSCAN, a newly developed technique which has never before been used in aerosol studies. As such, all of the results presented in Chapter 3 should be treated as preliminary, and will require further assessment. QEMSCAN (an automated electron beam image analysis system) was initially developed for applications in the mineral processing industry, and as such has features which include the provision of morphological information in association with textural characteristics and phase associations. This is possible as QEMSCAN collects compositional information on a pixel-by-pixel basis rather than on a particle basis as is done in conventional CCSEM (computer controlled scanning electron microscopy). (This description from Dr. David French, private communication.) Originally developed within the Commonwealth Scientific and Industrial Research Organisation (CSIRO) of Australia as QEM*SEM (or Quantitative Evaluation of Minerals by Scanning Electron Microscopy), the system uses a Windows-based PC operating system to control an automated SEM, fitted with four energy dispersive Xray spectrometers (EDS) which have enhanced light element capability (carbon and oxygen) via the use of thin polymer windows. This enables improved mineralogical identification and classification with the ability to discriminate phases such as iron oxides (hematite and limonite) and siderite in mineral matter. The ability to detect the characteristic radiation of carbon and oxygen further provides the potential to partially analyse some organics, which is of value in the analysis of aerosol particulates.


26

The measurement modes employed in data acquisition for QEMSCAN are fundamentally different from those used in most CCSEM systems in that an X-ray spectrum is acquired at each point of a rectangular grid superimposed on the particle, thus building up a phase composition map of the particle. QEMSCAN uses both backscattered electron and energy dispersive X-ray signals to create digital images in which each pixel corresponds to a mineral species, or phase, in a region under the electron beam. The backscattered electron signal is used to discriminate between the mounting medium and the particles of interest. After calibration of the BSE signal, the sample is moved under computer control to a particular field of view (referred to as a frame) which is rapidly scanned by the electron beam to identify the mineral particles from the mounting medium based upon preset threshold values of the BSE intensity. Further discrimination can be made to identify particles that are on the frame boundary, particles that are either too large or too small for consideration, or particles having a particular shape as defined by the operator. After particles have been identified and selected for measurement, an X-ray spectrum is collected from each point of a user-defined grid. The spacing of analytical points can be varied to give three basic scan modes, depending on the type of information and statistical accuracy required to satisfactorily address the particular issue. QEMSCAN has several forms of analysis. Particle Mineralogical Analysis (PMA) is the most detailed, and is based on the area scan of closely spaced points in the X and Y direction. This mode of analysis provides the most detailed information on particle composition, shape, size, and association. However, fewer particles are measured than in the MMA or BMA modes, so that more information is obtained at the expense of statistical accuracy. Depending on the point spacing selected, the analysis


27

Mineral Carbon Quartz Carbon + Fe Carbon + FeS Carbon + SiFe Carbon + AlFe Carbon + AlSiFe Carbon + CaMgFe Carbon + S Carbon + SiS Carbon + AlS Carbon + AlSiS Carbon + Al Carbon + Si Carbon + AlSi Other

Figure 2.5: QEMSCAN mineral maps of two air particulate samples: top an urban air particle sample and bottom, one collected in a major road tunnel showing distinct differences in the nature and morphology of the particles. Particles are sorted by decreasing area.


28

also takes significantly longer. This was the measurement mode used to analyse the particles collected on the filters. A 1,000-count X-ray spectrum (typically collected within 5 milliseconds) is obtained from each point, and the elements present and their X-ray intensities are used to identify the mineral species present at that point. This is done by comparison with a look-up table (the Species Identification Program), from which a species number is assigned to the pixel. The X-ray spectrum is matched to a mineral whilst the next spectrum is being acquired, thus allowing approximately 100,000 spectra to be processed and identified in one hour of measurement. The beam is then moved to another point and the process repeated until all selected particles in the frame have been measured. Another frame is then selected and another set of points measured, the process being repeated until a preset number of particles has been analysed. The SIP table is one of the most important elements of the QEMSCAN system as it is the primary phase classification table. Potentially it may have several hundred entries that contain information on the BSE response, the elements present and their relative intensities. It may also contain supplementary information in the form of logical operators as to which elements must be present and those which may be present. The SIP is created from the analysis of standards from which reference Xray spectra are obtained. These reference spectra are then used to generate simulated 1,000-count spectra against which the SIP is tested for correct assignment of each spectrum to a particular mineral phase. A list of the mineral species likely to be present in a particular sample group is built up by combining entries in the SIP to produce what is known as a Primary Species list. Secondary and Tertiary mineral lists may then be constructed, each of which contain a reduced number of mineral species that can be used in specific


29

applications. This capability is illustrated in the mineral maps of two air particulate samples, one from an urban area and the other from a major road tunnel, presented in Figure 2.5.

2.5

Computing optical properties

To calculating the dust optical properties the fundamental microphysical properties (such as particle size distribution, chemical and mineralogy composition, reflective index, mixture state and particle shape) is required. If these characteristics do not vary (assuming the particles are spherical and homogeneous), then standard Mie codes can be used to compute the optical properties of this aerosol.

∞

β (λ i ) = ∫ πr 2Qext (r,λ i ,mi )n(r )dr

(2.13)

0

where β(λ) is the optical depth at wavelength λi, Qext is the Mie extinction efficiency factor, m is the complex refractive index, r is the particle radius and n(r) is the number of particles per unit area in a vertical column of atmosphere. The Mie extinction efficiency factor Qext depends on the Mie Scattering Qsct and Mie Absorption Qabs efficiencies,

Q ext =Q sca +Q abs

(2.14)


30

The complex reflective index depends on the chemical composition and the mixture state of the particles. If there are different types of aerosol dominated in the atmosphere then the equation 2.13 will be the sum of the optical properties of each type of aerosol related to different size and reflective indexes. If the particles are external mixture that means these particles will be more scattering while internal mixture means these particles are more absorbing. If the dust particles are found to be not spherical, heterogeneous and the chemistry varies with size distribution then the spheroidal models (Mishchenko, et al., 2000 and Dubovik, et al., 2006) and T-matrix method (Trautmann, et al., 2009) should be use extensively until a combination of efficient computational techniques to be find, which returns adequate agreement with the direct measurements (in situ). See section 5.2.

Chapter 3 Characterization of Aerosol from the Lake Eyre Basin

31


3.1

Introduction

3.1.1

Mineral Dust Aerosol

Global annual dust emission is estimated to be ~ 1877 Tg yr-1, of which Australia contributes around 5% of the total, and is the greatest contributor of mineral aerosol in the Southern Hemisphere (Tanaka and Chiba, 2006; Mackie et al., 2008). The lifetime of dust in the atmosphere varies from a few hours to 10 days or more, depending on particle size (Tegen and Fung, 1994; Tegen and Lacis, 1996; Mahowald et al., 1999; Ginoux et al., 2001). This aerosol will influence the climate system by, firstly, affecting the Earth’s radiation budget (Charlson et al., 1992; Tegen et al., 1996; Sokolik et al., 2001) since it both scatters and absorbs solar radiation, while also absorbing and emitting some infrared radiation. IPCC (2007) reported that the radiative forcing of mineral dust is in the range of -0.1±0.2 Wm-2, indicating that even the sign is uncertain. A more recent study by Balkanski et al. (2007) estimated the radiative forcing at the top of atmosphere to be -0.47 to -0.24 Wm-2. The variability of optical properties of dust particles is dependent on the dust source region (Dubovik et al., 2002; Kubilay et al., 2003; Lafon et al., 2006). Therefore, physical

(size

and

shape)

and

chemical

(composition

and

mineralogy)


32

characterization of mineral aerosol for different source regions is important to understand the influence of dust aerosol on the climate system. Secondly, dust aerosol may change the physical and radiative properties of clouds (Ramanathan et al., 2001). Thirdly, dust particles may have an impact on human health, as the finer particles can travel deeply into the lungs (Englert, 2004). Fourthly, many studies have shown that dust particles provide surfaces for heterogeneous reactions leading to the formation of sulfate and/or nitrate layers (Li and Shao, 2009; Galindo et al., 2008; Bauer et al., 2007; Bauer and Koch, 2005; Jordan et al., 2003; Zhang and Iwasaka, 1999; Zhou et al., 1996; Iwasaka et al., 1988). By using electron microscopy, Iwasaka et al. (1988) found that dust particles were coated by sulfate. Li and Shao (2009) found that mineral particles in the Beijing atmosphere were coated with nitrate. Bauer et al. (2007) found that due to the coating process between sulfate and mineral dust particles, the radiative forcing of anthropogenic sulfate was reduced from -0.25 to -0.16 W/m2. Finally, dust particles are the main terrestrial source to the open and remote ocean (Prospero et al., 1989; Duce et al., 1991). Iron aerosol carried with aeolian dust was found to influence marine ecosystems after being deposited and dissolved in the ocean (Duce and Tindale, 1991; Martin et al., 1989), affecting the growth of phytoplankton in some oceanic regions (Karl et al., 2002), and hence biological productivity and the carbon cycle in the ocean (Schroth et al., 2009; Uematsu et al., 1983).


33

3.1.2 Australian Dust Aerosol: Current Knowledge and Gaps While much important work has been done on a number of aspects of Australian mineral dust (summarized by Greene et al., 2009), there is insufficient information on the size-resolved physics, composition, and optical properties of Australian mineral dust aerosol. For example, Kiefert et al. (1996) report that Australian dust mainly shows a particle size mode under 10 µm, but they did not identify the size distribution of this dust, or its chemical composition as a function of the size of the particles, essential information for the computation of optical properties. Australian dust aerosol has been found to be the main source to the Indian Ocean, Tasman Sea and Pacific Ocean (Bowler 1976; Petit et al., 1983; 1999; Hesse, 1994; Kiefert and McTainsh, 1996; Greene et al., 2001; Pillans and Bourman, 2001; Hesse and McTainsh, 2003). Knight et al. (1995) have estimated that on average around 3.8-6.8 Mt/y of Australian dust transported to the southeast is deposited into the South Pacific Ocean. Studies by Shaw et al. (2008) showed an increase in the phytoplankton standing stock in Queensland coastal waters immediately after a dust storm. Iron aerosol carried with aeolian dust was found to influence marine ecosystems after being deposited and dissolved in the ocean, therefore identifying the fraction, size and mixture statue of iron in Australian dust aerosol is very important to understand this influence. Using remote sensing data, Qin and Mitchell (2009) were able to classify Australian aerosol into four classes: aged smoke; fresh smoke; coarse mineral dust; and a superabsorptive class of unknown nature. They found that their dust aerosol class is highly absorbing in the blue wavelength region, which suggests that the Australian arid zone is rich in hematite (and internally mixed), by contrast with Northern Hemisphere dust regions. This is consistent with the fact that Australia’s deserts are reddish (Bullard


34

and White, 2002), in contrast to the Sahara, for example, which is more yellow, a reflection of the mineralogy, particularly in relation to iron oxides (Sokolik and Toon, 1996). Moreover, that study found that the energy deposited in the aerosol layer for this class is 63 Wm-2τ-1 by contrast with 31 Wm-2τ-1 for the Saharan region, and 33 Wm-2τ-1 for the Arabian Peninsula (Balkanski et al., 2007), a significant difference. Qin and Mitchell (2009) further assert that “a systematic characterization of the composition and optical properties of Australian dust aerosol is currently lacking”. In particular, this requires that these properties are characterised as a function of particle size (and mixing state if important). The aim of the work in this Chapter is to take the first key steps in filling this significant gap.

3.1.2

Lake Eyre Basin

Three field campaigns have been undertaken within the Lake Eyre Basin (LEB) (Birdsville in 2006; Muloorina Station in 2007; and Fowlers Gap Station in 2009: see Figure 3.1) to investigate the optical properties, size-resolved mass and chemical properties, and mineralogy of Australian mineral dust aerosol. These campaigns each involved the collection of size-resolved aerosol samples for subsequent analysis: ion beam analysis, ion chromatography, and mineralogy. In addition, we have three years of ground-based remote sensing data from Birdsville. The Lake Eyre Basin (LEB) is the largest closed drainage system in the leastdeveloped arid and semi-arid zone in heart of Australia: Figure 3.1. Lake Eyre is the lowest point in the basin, and all the rivers in the basin flow towards it (although much of the time no water actually reaches the Lake). This region covers around 1.2 million km2 with different sedimentary environments: further details of the


35

geomorphology and hydrology of this area can be found in Tyler et al., (1990). The deserts that have formed in the LEB constitute the largest airborne dust source region on the Australian continent and southern hemisphere (Middleton et al., 1986; Prospero et al., 2002; Washington et al., 2003; McTainsh and Lynch, 1996; Bullard et al., 2008; Baddock et al., 2009). Studies by Washington et al. (2003) and Bullard et al. (2008) have found that there is significant contribution to the airborne dust in this region from floodplains, ephemeral lakes and dry lake beds. Mitchell et al. (2010) found that dust activates over the Lake Eyre Basin has increased from 20032007, by contrast with the period 1997-2002. In addition to mineral dust, there are a number of other significant aerosol sources, either within, or close to, the Lake Eyre Basin. Dry salt lakes, of which Lake Eyre is the largest, are a source of NaCl (and other salts), potentially exacerbating salinity problems in inland Australia. At the northern end of the LEB, biomass burning aerosol from the winter/spring fires in northern Australia can be a significant part of the aerosol mix. Marine advection can bring sea salt, as well as biogenic compounds to the region. Mt Isa, on the northern edge of the LEB, is one of the world’s largest point sources of SO2. Finally, although vegetation is sparse, biogenic compounds may also contribute to the total aerosol mix within the LEB.


Birdsville

Muloorina

FGS

Figure 3.1: Lake Eyre Basin

36


3.2

37

Birdsville

3.2.1 Site Location Birdsville in south-west Queensland (25.9 S, 139.34 E, 48 m elevation) is a small outback town situated on the banks of the Diamantina River between the sands of the Simpson Desert and the gibbers of Sturt’s Stony Desert (Figure 3.1). Rainfall averages 167 mm per year, occurring mostly in summer, with September being the driest month. Daytime temperatures reach 38 °C in summer (November - February). Annual average relative humidity is 48%, with highest values of around 67% in June. Dust devils are common. The Birdsville permanent population currently stands at approximately 120, but is augmented at times (especially winter/early spring) by tourism, peaking at around 6000 for the “Birdsville Races” in early September.

3.2.2


A Cimel sun photometer at Birdsville has been monitoring aerosol optical depth (AOD) over a range of wavelengths since the spring of 2005. This instrument forms part of the CSIRO Aerosol Ground Station Network (Mitchell and Forgan, 2003), which is affiliated with NASA’a global Aerosol Robotic Network, AERONET. Instrument calibration and the generation of AOD and aerosol microphysical data (phase function, size distribution, refractive index and derived quantities such as single scattering albedo and asymmetry factor) are performed as part of the standard AERONET processing stream (Holben et al., 1998; Dubovik and King, 2000). In this study we will use AOD at 500 nm as our measure of aerosol loading. Approximately 95% of the daily average found to have a standard deviation less that 0.05 during


38

autumn and winter months , but this value have been reduce to around 85 % during some spring and summer months due to dust activates. Another parameter which can be extracted from this data is the Angstrom exponent,

α, which is obtained from a straight line fit of ln(AOD) vs. ln(λ), from 440 to 870 nm.

3.2.2.1 Daily and monthly data From plots of the daily and monthly means of AOD for the period August 2005 to June 2008, shown in Figure 3.2, a weak seasonal variability in aerosol optical depth can be seen. Daily averages of AOD (Figure 3.2a) display less variability during winter, reflecting lower wind speed that dominates during this season, resulting in a lack of significant local dust sources. During spring, summer and autumn, the variability in the daily average AOD is increased due to the influence of dust activity. Figure 3.2b shows the monthly means and standard deviations. The annual mean of AOD computed from the monthly means is 0.06 ± 0.03. The daily and monthly means of α are presented in Figure 3.3, which shows a very clear seasonal cycle, along with large variation in daily measurements, including negative values on some days – and especially during summer and early autumn – which is likely due to dust storm activity. (Inadequate cloud screening is unlikely to be a factor.) Higher values in AOD, along with a wide range of α values during spring months, is suggestive of a contribution from a different aerosol type such as regional biomass burning, and/or long range (intercontinental) transport (Rosen, et al., 2000; Gloudemans et al., 2006), and/or the possible influence of marine biogenic emission when the air mass is advected from the ocean. The situation is more


39

complex during winter months, when low AOD values are associated with high α values. A scatter plot of daily average Angstrom exponent, α, against daily average AOD at 500 nm, is presented in Figure 3.4a. This figure shows a wide range of α associated with AOD less than 0.03. To isolate the low AOD measurements, the time series of all AOD < 0.03 has been plotted in Figure 3.4b, which shows that nearly all of these measurements were during late autumn and winter. As a check, the time series of

α was replotted, excluding all data where AOD was less than 0.02: the seasonal cycle was still obvious. Winter is the dry season, characterized by clear skies, along with lowest temperatures and wind speeds. It is also the burning season in the ‘top end’ of northern of Australia, so that a biomass burning signature from that region at our Birdsville instrument is likely, whenever the winds are from an appropriate direction. (In addition, winter is a tourism season, when many people are camping under the stars and using firewood for cooking and heating, which produces biomass burning aerosol, although only in very small amounts). However, it should also be noted that the uncertainty in α increases significantly whenever it is computed from AOD data less than 0.02. For this reason, AOD data below 0.02 was not used when computing α because of the high probability of unacceptable errors.


40

3.2.2.2 Seasonal statistics The seasonal frequency distributions of daily means of AOD and α are presented in Figure 3.5. The left panel shows that the AOD distribution was very narrow during autumn and winter, with ~ 60% of the data observed in the 0.04 bin in both seasons. During summer the distribution was broader and the peak was in the 0.06 bin which accounted for ~ 40% of the data observed, with around 20% of the data observed in each of the two adjoining bins. The spring distribution was very broad, again peaking in the 0.06 bin, but with a long tail, with around 40% of the data above 0.08. Overall around 80% of daily AOD means are below 0.1. Around 80% are less than 0.06 during autumn and winter months, but only 50% fall in that range during spring and summer. The frequency distributions of daily means of α – central panel – were broad in all seasons, indicative of a range of particle sizes entering the atmosphere, or a range of atmospheric processing. This Figure shows a skewed distribution during summer with approximately half the data observed in the lowest two bins, which indicates that coarse mode particles predominate. During autumn the peak was shifted to the 1.0 and 1.3 bins, and accounted for ~ 40% of the total daily means, while around 30% of the daily means were in the 0.4 and 0.7 bins. The winter pattern showed a normal distribution with around 65% of the data in the 0.7-1.6 range. This indicates that fine mode particles predominate during this season, a result of local wood burning and regional biomass burning, plus (presumably) fine dust particles. The spring distribution was approximately normal, with around 70% of the data in the 0.7-1.6 range, which again implies both fine mode biomass burning aerosol (plus,


41

presumably, fine dust) and a contribution from a different aerosol type, such as coarse dust. The annual mean of α computed from the monthly means is 0.9 ± 0.3. Scatter plots of daily average α vs. AOD for each season in Figure 3.5 (right panel) show that during all seasons a wide range of α is associated with AOD less than 0.05. On the other hand, the α values generally decreased (including some negative values) as the value of AOD increased beyond 0.1 in all seasons, which implies that dust is the main contributor to the higher optical depth values during those seasons. However, the spring pattern shows a second aerosol signature with larger α values contributing to the larger optical depth measurements, indicating the influence of fine mode particles, most likely biomass burning aerosol.


0.8

42

a)

0.7 0.6 AOD

0.5 0.4 0.3 0.2 0.1 0 Aug-08

May-08

Feb-08

Nov-07

Aug-07

May-07

Feb-07

Nov-06

Aug-06

May-06

Feb-06

Nov-05

Aug-05

May-05 0.25

b)

AOD

0.20 0.15 0.10 0.05 0.00 Dec

Nov

Oct

Sep

Aug

Jul

Jun

May

Apr

Mar

Feb

Jan

Figure 3.2: a) Daily means of AOD from August 2005–Jun 2008. b) Monthly means and standard deviations of AOD for same period.

3.50

a)

3.00 2.50

α

2.00 1.50 1.00 0.50 0.00 -0.50 Dec-08

Jun-08

Nov-07

Apr-07

Oct-06

Mar-06

Sep-05

Feb-05 2.50

b)

2.00

1.50

α

1.00

0.50

0.00

Figure 3.3: a) Daily means of α from August 2005–Jun 2008. b) Monthly means and standard deviations of α for same period.

Dec

Nov

Oct

Sep

Aug

Jul

Jun

May

Apr

Mar

Feb

Jan

-0.50


3.5

43

a)

3.0 2.5

α

2.0 1.5 1.0 0.5 0.0 -0.5 0.72

0.68 0.64

0.60

0.56

0.52 0.48

0.44

0.40 0.36

0.32

0.28

0.24 0.20

0.16

0.12

0.08 0.04

0.00

AOD

0.035

b)

0.030

AOD

0.025 0.020 0.015 0.010 0.005 0.000

b) Daily means of AOD less than 0.03.

Jul-08

Jun-08

Apr-08

Feb-08

Dec-07

Oct-07

Aug-07

Jun-07

Apr-07

Feb-07

Dec-06

Oct-06

Aug-06

Jun-06

Apr-06

Feb-06

Dec-05

Oct-05

Aug-05

Figure 3.4: a) Scatter plot of daily means of α vs. daily means of AOD.


50%

Summer

2.0

30% 20%

1.5

20% α

Frequency %

40% Frequency %

2.5

Summer

30%

Summer

44

1.0

10%

0.5

0%

-0.5

10%

0.0

0% 80%

Autumn

30%

Autumn

Autumn

2.5

60% 40% 20%

20%

1.5 α

Frequency %

Frequency %

2.0

1.0

10%

0.5

0%

-0.5

0.0 0% 80%

Winter

3.0

Winter

30%

Winter

2.5

40% 20%

2.0

20% α

Frequency %

Frequency %

60%

1.5 1.0

10%

0.5 0%

0.0

0%

30%

Spring

3.0

Spring

30%

Spring

20%

10%

2.0

20% α

Frequency %

Frequency %

2.5

10%

1.5 1.0 0.5

0%

0.0

Right panel: Seasonal scatter plots of daily mean α vs. daily mean AOD.

0.8

Central panel: Seasonal frequency distributions of daily mean α.

0.7

Figure 3.5: Left panel: Seasonal frequency distributions of daily mean AOD.

0.6

AOD

0.5

0.4

0.3

0.2

0.1

α

0

>2.5 2.5 2.2 1.9 1.6 1.3 1.0 0.7 0.4 0.1

>0.2 0.22 0.2 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 AOD

0%


45

3.2.2.3 Aerosol size distribution The daily average (columnar) volume size distributions from the AERONET web site have been used to calculate seasonal average volume size distributions: these are presented in Figure 3.6a. This graph shows a clear bimodal pattern in all seasons, as well as some significant differences. Summer and spring have the largest aerosol loadings, with a coarse mode peak at 3.5 μm radius, however this peak varied from day to day depending on wind speed and direction. Figure 3.6b shows the volume size distributions for selected days during spring 2007, and it can be seen that the coarse mode peak shifts from day to day. The spring distribution shows the strongest fine mode, with a peak between 0.10 and 0.15 μm, which is further evidence of a second aerosol contribution (most likely biomass burning) during this season. Figure 3.6b shows that on 8 October 2007 the fine mode peak was dominant, and the air mass back trajectory analysis for this day shows advection from the north, where hot spot satellite imagery confirmed that biomass burning was occurring (as expected for this time of year). The autumn and winter distributions show lower total aerosol content, with the coarse mode again dominating. Qin and Mitchell (2009) performed a cluster analysis of AERONET retrievals from a number of Australian sites (3 in northern Australia subject to biomass burning, 2 desert sites, plus others), including Birdsville, yielding 4 classes which they identify as aged smoke, fresh smoke, coarse dust, and an undetermined super-absorptive class. While all 4 classes have bimodal size distributions, it is only the coarse dust class which has a dominant course mode (and a very small fine mode). The aerosol size distribution over Birdsville is clearly bimodal: a fine mode which is believed to be primarily biomass burning, and a coarse mode which is believed to be mineral dust. The contributions of each vary seasonally in predictable ways.


α 440-870 AOD 440nm

dv/d (lnr) [μm3/μm 2]

0.020 0.018

summer 0.55

0.05

0.016

Autumn 1.0

0.03

Winter

1.4

0.03

Spring

1.02

0.09

0.014 0.012

46

0.010 0.008 0.006 0.004 0.002 0.000 0.01

0.1

1

10

100

Radius (μm)

Figure 3.6a: Seasonal average volume size distributions.

AOD440 0.23 21/09/07 0.18 24/09/07 0.15 0.39 8/10/07 11/10/07 0.53

0.08

16/09/07 0.07

dv/d (lnr) [μm3/μm2]

0.06 0.05 0.04 0.03 0.02 0.01 0 0.01

0.1

1

10

100

Radius (µm)

Figure 3.6b: Daily average volume size distribution for selected days


47

The fine mode occurs mainly in spring, the peak of the biomass burning season in northern Australia. The coarse mode strength is primarily dictated by wind speeds, which are stronger in spring and summer.

3.2.3

Aerosol Samples and Gravimetric Mass Distributions

In November 2006 a 12 stage Micro Orifice Uniform Deposition Impactor (MOUDI) was deployed to collect size-segregated aerosol particles. During this field campaign 5 set of samples – A, B, C, D and E – were collected. Sample A was used for equipment testing only, and has not been included in the analysis. Table 3.1 gives the start and finish times of sampling periods. Note that collection times varied, depending on dust activity conditions. The size-resolved mass concentrations for different data sets are presented in Figure 3.7, which shows that the size spectrum was quite variable during this field campaign. Total suspended particulates (TSP) during the sampling period are in order C > B > E > D: see Table 3.2. The mass concentration was significantly higher during periods B and C due to the influence of moderate to high wind speeds and local dust storm activity on those days. During sampling period B there was significant raised dust, although not associated with a dust storm. (A dust devil may have contributed to this.) The mass size distribution during this period was multi-modal, but without a definable ‘structure’. During period C, a weak local dust storm occurred (but for few a moments only). The resulting size distribution was tri-modal with peaks at 0.18, and 0.56 μm, and a broad coarse mode peak. During the other sampling periods, the distributions were relatively flat.


48

Table 3.1: Times of the sample collections at Birdsville. Start

End

date

time

date

time

Total hours

B

1st Nov. 2006

9.10AM

2nd Nov. 2006

9.10AM

24

C

2nd Nov. 2006

4.00PM

3rd Nov. 2006

4.00PM

24

D

3rd Nov. 2006

5.00PM

5th Nov. 2006

5.00PM

48

E

5th Nov. 2006

6.00PM

7th Nov. 2006

6.00AM

36

Collection

Table 3.2: Mass fractions at Birdsville. B

C

D

E

3

57.2

107.5

17.5

28.8

PM10 / TSP

73%

78%

80%

72%

76% ± 4

PM2.5 / TSP

42%

44%

53%

47%

46% ± 5

PM1 / TSP

32%

27%

32%

34%

31% ± 3

PM2.5 / PM10

58%

57%

66%

65%

61% ± 5

PM1 / PM10

44%

35%

40%

47%

41% ± 5

PM1 / PM2.5

76%

62%

60%

73%

68% ± 8

TSP μg/m

Average

A smoothing/inversion procedure was applied to these 4 data sets (Keywood et al., 1999), and the results are presented in Figure 3.7b. From these retrievals the masses in key size ranges can be extracting (see below). AERONET retrievals were available on two of these days, and particularly sampling period C, which gave the most clearly defined modal distribution. In Figure 3.7c the inverted mass distribution for this period was replotted, along with the AERONET retrieval. Several factors may explain the differences, particularly for sizes above 10 µm. Firstly, the MOUDI


49

inverted results are for the surface only while AERONET is column retrieval (hence the different scales): the larger particles are likely to be found only in the lowest atmospheric layers. Secondly, the MOUDI mass results are based on a 24 hour sampling period. Finally, a number of authors have questioned the reliability of AERONET retrievals in this size range, given that the longest measurement wavelength is only 1.02 µm. In particular, the comparison shown in Figure 3.7c suggests that the AERONET inversion leads to an artificially steep decline on the large particle side of the coarse mode peak, an issue discussed by Qin and Mitchell (2009). Given these caveats, we consider the agreement to be as good as could be expected. Table 3.2 shows the ratios between PM10, PM2.5, PM1 and TSP; PM2.5, PM1 and PM10; and PM2.5 and PM1. (‘PMn’ is “particulate matter with diameters less than n

μm”.) On average the PM10, PM2.5 and PM1 components make up 76%, 46% and 31% of TSP respectively. PM2.5 and PM1 comprise 61% and 41% of PM10 respectively, and PM1 makes up 68% of PM2.5. From the above analysis the concluded are that the PM2.5 component (sometimes taken as the fine mode for air quality purposes) accounts for approximately 50% of all particulate mass in the Birdsville atmosphere at this time of year. The coarse particles are, of course, more rapidly removed by gravitation.


B

25.0

C

D

50

E

Mass cons. µg/m3

20.0 15.0 10.0 5.0 0.0 18

10

5.6

3.2

1.8

1

0.56

0.32

0.18

0.1

0.056

< 0.056

Aerodynamic diameter, μm (MOUDI Stage)

Figure 3.7a: Size resolved mass concentrations for all 4 samples. 80

B

C

D

E

Mass dM/dlogdp (µg m-3 )

70 60 50 40 30 20 10 0 100.0

10.0

1.0

0.1

0.0

Diameter (µm)

Figure 3.7b: Smoothed size distributions.

80

0.030 0.025

dv /d (lnd) [µ m3 /µ m 2 ]

M as s dM /dlogdp (µ g m -3 )

70 60 50 40 30 20

0.020 0.015 0.010 0.005

10 0.000

Diameter (µm)

Figure 3.7c: Comparison of MOUDI size distribution with AERONET retrieval for sample C.

100.0

10.0

1.0

100.0

10.0

1.0

0.1

0.0

Diameter (µm)

0.1

0


3.2.4

51


Ion beam analysis showed that, as expected, Si is an abundant element in all size fractions, and is used in this study as a dust ‘indicator’ for Australian desert aerosol. The elemental concentrations determined through the IBA are used to calculate the mass ratio of elements to Si: these ratios are summarised in Table 3.3, along with the mass ratios of these elements in the Earth’s crust from CRC Handbook (Lide, 1997). Size-resolved mass ratios for selected elements are presented in Figure 3.8. Scatter plots for some of these elements, based on all 48 data values, and are presented in Figure 3.9. (it is worth to not that, “The ratio presented in Table 3.3 is the sum of the mass of the element occurring on all filters over the sum of the mass of silicon occurring on all filters collected during one sampling period, while the mass ratio presented in Figure 3.8 is the ratio of element to silicon for individual filters”. The elements are classified into two groups: crustal elements (Al, Fe, Ca, K, Mn and Ti); and Na and Cl.

3.2.4.1 Crustal elements The TSP Fe/Si mass ratios are in the range 0.215 - 0.232, which is moderately higher than the values in the Earth’s crust: this most probably reflects the high amount of Fe which occurs naturally in Australian desert soil. The Fe/Si mass ratios for all samples are shown in Figure 3.8, which shows a complex structure, reasonably consistent across all four samples. The scatter plot of Fe versus Si in Figure 3.9a shows an excellent linear relationship with R2 = 0.98, indicating that soil is the source of Fe in the samples collected. This ratio of Fe to Si may be used to build a signature for Australian soil.


52

Table 3.3: Ratio of elements to Si at Birdsville. Ratio/Si

Element Na

B

C

D

E

Crust

8.63E-02

7.14E-02

3.80E-01

1.77E-01

8.37E-02

Al

2.71E-01

2.92E-01

2.73E-01

2.71E-01

2.92E-01

P

1.18E-02

4.36E-03

2.03E-02

3.56E-02

3.72E-03

Cl

5.54E-02

2.69E-02

3.70E-01

1.48E-01

5.14E-04

K

6.18E-02

5.49E-02

5.24E-02

4.75E-02

7.41E-02

Ca

3.06E-02

3.20E-02

5.91E-02

4.39E-02

1.47E-01

Ti

2.23E-02

2.55E-02

1.81E-02

1.87E-02

2.00E-02

Cr

7.36E-03

3.37E-03

1.25E-02

1.27E-02

3.62E-04

Mn

3.77E-03

3.68E-03

4.77E-03

4.61E-03

3.37E-03

Fe

2.15E-01

2.25E-01

2.32E-01

2.28E-01

2.00E-01

Co

1.42E-03

1.30E-03

1.22E-03

1.54E-03

8.87E-05

Ni

9.71E-03

4.17E-03

1.75E-02

1.98E-02

2.98E-04

Cu

1.92E-03

9.89E-04

1.66E-03

1.54E-03

2.13E-04

Zn

1.30E-03

8.96E-04

1.55E-03

1.82E-03

2.48E-04

Br

3.09E-03

1.45E-03

6.88E-03

4.89E-03

8.51E-06

Pb

3.22E-03

1.02E-03

1.89E-03

2.23E-03

5.32E-08

The TSP Al/Si mass ratios are in the range 0.271 - 0.292, close to the value in the Earth’s crust, and the scatter plot of Al against Si in Figure 3.9a again shows an excellent linear relationship with R2 = 0.99. The TSP Ti/Si mass ratios were between 0.018 and 0.026, which is in good agreement with Earth’s crustal value. The scatter plot of Ti against Si again shows an excellent linear relationship with R2 = 0.97. Both of these ratios can be incorporated in our soil signature. The Ca/Si TSP mass ratios are lower than the value in the Earth’s crust by ~ 60-80% for all samples. The scatter plot of Ca against Si appears to show two populations, indicative of two sources. The primary population displays an excellent linear relationship with R2 = 0.97, so that this population can be assumed to be associated with Si in Australian soil. The second population comprised particles between 1.8 to 10.0 μm, and from samples D and E only. This population also shows a reasonable


B

C

D

E

1

0.1

B

1 F e/Si M as s R atio

N a / Si M as s R atio

10

0.01

D

E

1 K/Si M as s R atio

C l / Si M as s R atio


1 0.1 0.01 0.001

B


E

0.1 0.01

M n/Si M as s R atio

C a / Si M as s R atio

D

1

B

C

D

E

0.1 0.01 0.001

18 10 5.6 3.2 1.8 1 0.56 0.32 0.18 0.1 0.056