3.1b: Supporting Information

Indicator 3.1b: Area of forest burnt by planned and unplanned fires

This indicator reports on the area of forest burnt by planned or unplanned fires in the five years of the period 2011–12 to 2015–16. Monthly fire data are collated and reported annually, by financial year. The data are then reported both as the cumulative area of forest fire in the five-year period, and as the total area of forest burnt during the period.

The cumulative area of fire in forest in the five-year period is the sum of the annual forest fire area totals. Some areas of forest burnt in multiple years of the reporting period, and thus over a five-year period the cumulative area of fire in forest substantially exceeds the total area of forest that experienced fire. Over many years, the cumulative area of fire in forest would exceed the total area of forest in a region.

For SOFR 2018, this indicator therefore also reports the total area of forest burnt one or more times during the period, a metric that counts an area of burnt forest only once no matter how many times it burns in a reporting period. The total area of forest burnt can never exceed the total area of forest in a region.

Where and when do bushfires occur?

The Australian climate is generally hot, dry and prone to drought. At any time of the year, some parts of Australia are prone to bushfire, with the widely varied fire seasons reflecting the continent’s different weather patterns (Figure 3.12). For most of south-eastern and south-western Australia, the fire danger period is summer and autumn. For areas in northern New South Wales and southern Queensland, peak risk usually occurs in spring and early summer. Most fires in the Northern Territory, Queensland and northern Western Australia are in the monsoonal dry season, which coincides with the southern winter and spring.

Bushfires in eucalypt forests tend to occur when fuel loads have dried out, usually following periods of low rainfall and high temperatures. In grasslands, however, and in woodlands with a grassy understorey, fires frequently occur after good periods of rainfall which result in abundant growth that dries out in subsequent hot weather.

The potential for extreme fire weather varies greatly throughout Australia, both in frequency and severity. The greatest extent of fire is in the Northern Territory and northern areas of Western Australia and Queensland, where there are large, sparsely settled areas with few roads, and where dry-season fires started by lightning or other causes burn large areas. Most loss of life and economic damage occurs in the areas around cities and regional towns in south-eastern and south-western Australia, where homes are commonly in close proximity to flammable vegetation.

Figure 3.12: Distribution of bushfire seasonality across Australia

Map source: Bureau of Meteorology (www.bom.gov.au/weather-services/bushfire/about-bushfire-weather.shtml); see also Luke and MacArthur (1978)

Source: Adapted text from Geoscience Australia (www.ga.gov.au/scientific-topics/hazards/bushfire)

Policy and coordination of fire management in Australia

The National Bushfire Management Policy Statement for Forests and Rangelands (FFMG 2014) outlines Australian, state and territory government objectives and policies for the management of landscape-level fire in Australia’s forests and rangelands. The statement was developed by the Forest Fire Management Group, a national body within the Australian Government ministerial council structure, which has the role of providing information to governments on major forest fire-related issues, policies and practices affecting land management.

The Australasian Fire and Emergencies Authorities Council is the national peak organisation that provides advice on a range of policies and standards. Research on bushfires is performed by a number of organisations, including the Bushfire and Natural Hazards Cooperative Research Centre162, which brings together experts from universities, fire and emergency management agencies, CSIRO, and other Australian, state and territory government organisations for long-term programs of collaborative research.

Determining the extent of fire in Australia’s forests: data sources and analysis

Australia has no nationally coordinated approach to the systematic mapping and reporting of fire areas. For reporting in SOFR 2018, annual spatial coverages of fires for the period 2011–12 to 2015–16 were therefore sourced from each state and territory separately, either by direct provision by the state or territory or from the North Australia and Rangelands Fire Information (NAFI) website164. Most jurisdictions create their fire area dataset from multiple sources, including satellite imagery, aerial photography, aerial reconnaissance, and operational and on ground knowledge and measurement.

Data sources for jurisdictions

Meaningful datasets of fires or burnt areas in woodland forests, such as the savannas of northern Australia, can be derived from satellite-based platforms carrying Advanced Very High Resolution Radiometer (AVHRR165), Moderate-resolution Imaging Spectroradiometer (MODIS166) and Landsat ETM167 sensors. The different satellites detect areas affected by fire in different ways (for example, through hot-spots, smoke plumes or vegetation changes), and combining the fire area data from different sensors gives a fire area statement that is larger than that from each satellite individually.

The extent and distribution of fire or burnt areas in open or closed forests, such as in the forests of southern Australia, is determined by combining satellite data with ground-based measurements and high-resolution aerial photography. Spatial data collated in this way were provided by the Australian Capital Territory, New South Wales, South Australia, Tasmania, Victoria and Western Australia. For these jurisdictions, each fire was allocated as either unplanned or planned by the jurisdiction that provided the data, as the seasonal distribution of planned and unplanned fires differs between jurisdictions.

Fire area provided by the Northern Territory and Queensland were derived solely from remote sensing. Fire area data provided by the Northern Territory were derived from the NAFI website, which combines MODIS satellite data with data from satellites carrying an AVHRR sensor, and also incorporates Landsat satellite data. Queensland agencies provided data from NAFI as well as Landsat data used in the Queensland Statewide Landcover and Trees Study program, with these two datasets being combined by ABARES.

Planned and unplanned fires in Australia

The main factors required for propagation of fire are the presence of fuel, oxygen and an ignition source. Fires can originate from human activity and from natural causes, with lightning nearly always the natural source of fire. Fire intensity and the speed at which a fire spreads depend on fuel load and arrangement, fuel moisture, prevailing temperature, wind speed and slope angle. The most intense fires occur when temperatures are high, humidity is low, winds are strong, and the arrangement of fuel allows rapid propagation. Box 3.1 summarises the occurrence of bushfires in Australia. Detailed geographic descriptions of Australia’s fire regimes have also been published (Murphy et al. 2013).

For the Northern Territory and Queensland, the allocation of fires as unplanned or planned was based on their month of occurrence. Northern Territory fires occurring from January to July were allocated as planned fires, while those occurring from August to December were allocated as unplanned fires. Queensland fires occurring between January and June were allocated as planned fires, while those occurring between July and December were allocated as unplanned fires.

The extent and intensity of unplanned fires, or bushfires, vary with latitude and seasonal rainfall (see Box 3.1), and the drivers of fire are substantially different across the continent:

Unplanned fires

The incidence of fire in northern Australia is essentially limited by fuel loads, and low-intensity fires burn over large areas in each dry season. The incidence of fire in southern Australia is essentially limited by fuel dryness, and some areas of south-eastern and south-western Australia are prone to severe bushfires: hot, dry and windy summer conditions, especially following periods of drought, lead to fires in eucalypt forest that are often very intense and difficult to control. Such bushfires can result in the loss of human life, and destroy assets such as buildings, fences, bridges and powerlines as well as standing stocks of wood (native forest and plantations). They can also have a significant impact on ecological values, and affect water supplies. Bushfires are rare in the tropical rainforests of northern Australia, and are occasional in the subtropical, temperate and cool-temperate rainforests of southern Australia. However, during prolonged droughts even these forests can be damaged by fire entering from adjacent grasslands or eucalypt forests.

Climate change and weather pattern variability are among the key factors that are predicted to affect the future occurrence and severity of bushfires. Projected increases in summer temperatures and declines in rainfall are predicted to exacerbate the risk of fire and increase the challenges associated with fire management.

More frequent and intense bushfires could also increase the incidence and severity of certain pests, diseases and weeds. For example, populations of bark beetles (Ips spp.) may increase in response to a higher availability of fire-damaged (dead, dying or stressed) trees that can be colonised. Furthermore, forests affected by pests, diseases and weeds may become more vulnerable to bushfires as a result of increases in fuel loads due to tree mortality (Singh et al. 2010). Indicator 3.1a provides more information on pests, diseases and weeds affecting forest health.

Planned burning

Planned or prescribed burning is the deliberate use of fire to achieve particular management objectives, and is an important management tool on both public and private land. Case study 3.3 on the National Burning Project: Prescribed Burning Guidelines and Frameworks describes how the principles underpinning planned burns were articulated and put into practice. Management objectives for planned burns can include reducing the levels of flammable fuels (fuel reduction burning), protection and enhancement of biodiversity in fire-adapted ecosystems, and promoting regeneration after wood harvesting; not all of these objectives are necessarily served equally by the same burn frequency or intensity.

Planned burning does not prevent unplanned fires. However, in some ecological communities previous planned burning can reduce the intensity of unplanned fires, aid control efforts by widening the range of weather and other conditions under which an unplanned fire may be controlled, and potentially allow firefighters to break the run of large fires (McCaw 2013). This can lead to a reduced area of unplanned fire, and a lower impact.

In the tropical savannas of northern Australia, woodland forests with a grassy understorey are part of a patchy landscape mosaic. Rapid growth occurs in the wet season, and this is converted to fuel during the dry season, with an increased risk of high-intensity fires late in the dry season. Up to 50% of some northern Australian landscapes may be burnt in a single year, and most areas burn at least once every three years. Land managers in northern savannas are increasingly employing traditional, early dry-season burning techniques, where burning occurs at low intensity and in a patchy mosaic, so as to reduce the risk of extensive, high-intensity late-season fire and consequential carbon dioxide emissions. Complete suppression of fire, on the other hand, can lead to increased tree and shrub invasion, which may adversely affect biodiversity and habitat values and reduce pastoral productivity, and can also lead to cumulative increases in fuel loads and an associated increased fire risk. Case study 5.3 in Indicator 5.1a on the Western Arnhem Land Fire Abatement project describes one such planned burning program in northern Australia.

The EcoFire project in the Kimberley, Western Australia (Legge et al. 2011), a partnership between landholders, private conservation organisations and government agencies, is also working to use planned early dry-season fires to minimise the area of extensive, intense, uncontrolled and unplanned mid-to-late dry season fires and thereby improve habitat quality and the proportion of long-unburnt vegetation across the landscape.

Prescribed burning for fuel reduction needs to bring together contrasting expectations: the public expectation that fuel hazards will be managed to protect life and property, and concerns that inappropriate burning will affect biodiversity and other values (McCaw 2013). Prescribed burning regimes will have undue impacts if burning is more frequent, intense or uniform than the natural fire regime for a particular ecological community, or if it occurs at times of the year when natural processes are adversely affected. Area targets for fuel-reduction burning can be designed to balance community safety and asset protection with protection of ecological values and maintenance of ecological processes. Whether or not area targets are achieved depends on weather and fuel conditions: unseasonably warm, dry or windy weather can make prescribed burning too risky, and unseasonably cold or wet weather can make prescribed burning ineffective.

 

Table 3.6: Area of forest fire, 2011–12 to 2015–16, by year and jurisdiction, separately for planned and unplanned fire (’000 hectares)

Jurisdiction

Forest fire area

Cumulative area of fire in forest, 2011–12 to 2015–16a

Proportion of cumulative area of fire in forest, 2011–12 to 2015–16

2011–12

2012–13

2013–14

2014–15

2015–16

Planned fire

ACT

0

9

0

3

5

17

0.1%

NSW

75

221

125

122

248

791

2.4%

NT

3,810

1,969

2,475

3,059

1,853

13,166

40%

Qld

2,700

2,371

3,102

3,056

1,930

13,159

40%

SA

7

11

2

2

3

24

0.1%

Tas.

10

14

10

23

4

60

0.2%

Vic.

99

118

46

146

118

526

1.6%

WA

1,536

1,484

828

651

686

5,184

16%

southern WAb

96

35

80

139

155

504

1.5%

northern WAc

1,441

1,450

748

512

531

4,681

14%

Australia

8,236

6,197

6,587

7,061

4,847

32,927

100%

Unplanned fire

ACT

0

0

0

0

0

0

0.0%

NSW

11

318

485

119

41

975

1.3%

NT

6,299

8,107

4,310

7,532

6,152

32,399

44%

Qld

11,940

12,360

3,088

6,082

3,273

36,743

50%

SA

6

28

221

8

16

279

0.4%

Tas.

3

51

7

6

80

147

0.2%

Vic.

2

136

318

28

14

498

0.7%

WA

363

153

467

398

457

1,837

2.5%

southern WAb

104

70

26

155

313

668

0.9%

northern WAc

259

83

441

243

143

1,169

1.6%

Australia

18,623

21,154

8,896

14,174

10,032

72,880

100%

All fire

ACT

0

9

0

3

5

17

0.0%

NSW

85

540

610

241

289

1,766

1.7%

NT

10,109

10,076

6,784

10,591

8,004

45,565

43%

Qld

14,640

14,731

6,190

9,138

5,203

49,902

47%

SA

13

39

222

10

19

302

0.3%

Tas.

13

65

18

29

84

208

0.2%

Vic.

100

255

363

174

132

1,025

1.0%

WA

1,899

1,637

1,294

1,049

1,142

7,022

6.6%

southern WAb

199

104

106

294

469

1,172

1.1%

northern WAc

1,700

1,533

1,189

755

674

5,849

5.5%

Australia

26,860

27,351

15,483

21,235

14,879

105,807

100%

a Cumulative area of fire in forest is the sum of the five annual area totals, and therefore counts multiple times any forest areas that were burnt in two or more years of the five-year period. This metric can therefore exceed the total forest area.

b Data for forest south of the Tropic of Capricorn, calculated using an interim area mask.

c Data for forest north of the Tropic of Capricorn, calculated using an interim area mask.

Table 3.8: Cumulative area of fire in forest, 2011–12 to 2015–16, by tenure and jurisdiction, separately for planned and unplanned fire

Jurisdiction

Forest fire area ('000 hectares)

Leasehold forest

Multiple-use public forest

Nature conservation reserve

Other crown land

Private forest

Unresolved tenure

Total

Planned fire

ACT

0

1

16

0

0

0

17

NSW

5

83

622

20

61

1

791

NT

4,132

0

5

1,294

7,728

8

13,166

Qld

6,527

412

1,114

536

4,524

46

13,159

SA

1

1

21

0

1

0

24

Tas.

0

23

22

12

4

0

60

Vic.

0

279

223

18

6

0

526

WA

210

240

3,045

1,241

449

0

5,184

southern WAa

3

240

228

22

11

0

504

northern WAb

207

0

2,817

1,218

438

0

4,681

Australia

10,874

1,039

5,067

3,121

12,773

54

32,927

Unplanned fire

ACT

0

0

0

0

0

0

0

NSW

18

131

479

28

320

0

975

NT

9,722

0

18

747

21,899

14

32,399

Qld

21,562

1,409

2,250

855

10,511

156

36,743

SA

47

10

181

2

38

0

279

Tas.

0

48

34

28

37

0

147

Vic.

0

159

303

3

33

0

498

WA

140

107

910

510

170

0

1,837

southern WAa

29

107

349

127

56

0

668

northern WAb

110

0

561

384

114

0

1,169

Australia

31,488

1,864

4,175

2,174

33,008

170

72,880

All fire

ACT

0

1

16

0

0

0

17

NSW

22

214

1,100

48

380

1

1,766

NT

13,853

0

22

2,041

29,627

22

45,565

Qld

28,090

1,820

3,364

1,391

15,035

202

49,902

SA

48

11

202

2

39

0

302

Tas.

0

71

56

41

40

0

208

Vic.

0

438

526

21

39

0

1,025

WA

349

347

3,955

1,751

619

0

7,022

southern WAa

32

347

576

149

67

0

1,172

northern WAb

317

0

3,379

1,602

552

0

5,850

Australia

42,362

2,903

9,242

5,295

46,229

225

105,807

a Data for forest south of the Tropic of Capricorn, calculated using an interim area mask.

b Data for forest north of the Tropic of Capricorn, calculated using an interim area mask.

Area of forest burnt one or more times

The data indicated that some areas had burnt more than once in any one financial year. In such situations, only the first fire in that year (whether planned or unplanned) was retained in the data. This approach had only a small (<2%) effect on total area figures.

The fire datasets for each jurisdiction were then intersected with the forest cover dataset (Indicator 1.1a) to produce forest fire statistics.

Large areas of northern Australia were reported as having burnt in multiple years of the five-year reporting period, (2011–12 to 2015–16). This indicator therefore reports separately the cumulative area of fire in forest (the sum of the five individual-year forest fire areas; this counts every time an area of forest was burnt in the five-year period) and the total area of forest burnt (in which a burnt area of forest is counted only once in the five-year period, even if it was burnt more than once in that period). SOFR 2008 and SOFR 2013 reported only the cumulative area of fire in forest in the five-year reporting period (the sum of the five individual-year areas).

Both the data sources and the methods used to derive area of forest fire for SOFR 2018 are different to those used for SOFR 2008 and SOFR 2013, and therefore the results cannot be directly compared between these reports. This is particularly so for Northern Territory, Queensland and northern Western Australia, where only the MODIS dataset was used for SOFR 2008 and SOFR 2013, compared to the wider range of datasets used for SOFR 2018. Furthermore, different algorithms and data resolutions were used in analysis of the fire datasets reported in SOFR 2018. Lastly, a larger and more accurate forest coverage was used for SOFR 2018, particularly in the Northern Territory (see Indicator 1.1a).

When fire areas burnt in multiple years are allowed for, the total area of forest that was burnt once or more in the period 2011–12 to 2015–16 was determined as 55 million hectares, which is 41% of Australia’s forest area (Table 3.7). This is the total area impacted by fire once or more during the reporting period, and is represented as the various colour areas labelled 1–5 on Figure 3.14.

Table 3.7: Area of forest burnt by number of times burnt, by jurisdiction, 2011–12 to 2015–16 ('000 hectares)

Jurisdiction

Total forest area

Forest area not burnt, 2011–12 to 2015–16

Forest area burnt, 2011–12 to 2015–16

Number of times forest was burnt by area

Area burnt one or more timesa

Proportion of total area by jurisdiction

Area burnt two or more timesb

Proportion of total area by jurisdiction

One

Two

Three

Four

Five

ACT

142

125

17

0

0

0

0

17

12%

0

0.0%

NSW

20,368

18,637

1,696

35

0

0

0

1,731

8%

35

0.2%

NT

23,735

3,689

5,434

6,836

5,106

2,212

459

20,046

84%

14,612

62%

Qld

51,830

23,539

15,397

6,913

3,708

1,809

463

28,291

55%

12,894

25%

SA

5,060

4,759

300

1

0

0

0

301

6%

1

0.0%

Tas.

3,699

3,495

200

4

0

0

0

204

6%

4

0.1%

Vic.

8,222

7,228

964

30

0

0

0

995

12%

30

0.4%

WA

20,981

17,737

1,447

589

660

324

224

3,244

15%

1,797

9%

southern WAc

17,357

16,220

1,107

25

4

1

0

1,136

7%

30

0.2%

northern WAd

3,624

1,516

341

564

656

323

224

2,108

58%

1,768

49%

Australia

134,037

79,208

25,456

14,408

9,474

4,345

1,146

54,829

41%

29,374

22%

Proportion of total national forest area

100%

59%

19%

11%

7%

3.2%

0.9%

41%

 

22%

 

 

a Sum of the areas burnt one times, two times, three times, four times and five times in the period 2011–12 to 2015–16, giving the total area of forest burnt one or more times. The forest areas burnt on multiple occasions are counted only once in these totals.

b Sum of the areas burnt two times, three times, four times and five times in the period 2011–12 to 2015–16, giving the total area of forest burnt multiple times.

c Data for forest south of the Tropic of Capricorn, calculated using an interim area mask.

d Data for forest north of the Tropic of Capricorn, calculated using an interim area mask. Note: Multiplying each area of forest burnt by the number of years in which it was burnt in the period 2011-12 to 2015-16, and summing the results, gives the cumulative area of forest fire over the reporting period (Table 3.6).

 

The largest areas of forest burnt in the period 2011–12 to 2015–16 were in Queensland (28 million hectares, 55% of Queensland’s total forest area, and 52% of the total national area of forest burnt) and the Northern Territory (20 million hectares, 84% of the Northern Territory’s total forest area, and 37% of the total national area of forest burnt).

Of the total area of forest burnt in the period 2011–12 to 2015–16, 29 million hectares (22% of Australia’s forest area) was burnt multiple times. These areas of forest were almost completely confined to northern Australia (Figure 3.14), including substantial areas of forest in the Northern Territory (15 million hectares, 62% of the Northern Territory’s total forest area), Queensland (13 million hectares, 25% of Queensland’s total forest area) and northern Western Australia (1.8 million hectares, 49% of the total forest area of northern Western Australia). A small area of forest was burnt in all five years of the reporting period (Table 3.7). Only very small areas of forest in southern Australia were burnt more than once between 2011–12 and 2015–16. Figure 3.15 (see page 270) shows the distribution of burnt forest in Australia coloured according to whether the area was burnt by planned fire, by unplanned fire, or by both planned and unplanned fires, in the period 2011–12 to 2015–16.

Nationally, 79 million hectares of forest (59% of Australia’s forest area) were not burnt at all in the period 2011–12 to 2015–16. The jurisdictions with the highest proportions of forest area not burnt during this period were Tasmania (94%) and South Australia (94%) (Table 3.7).

The high fire frequency in northern Australia is driven by the characteristics of vegetation and climate. In open tropical forests with a grassy understorey, periods of prolific annual growth in the wet season are followed by rapid drying in the dry season, and lightning associated with storm events leads to frequent and extensive unplanned fires, especially late in the dry season. Case study 5.3 in Indicator 5.1a describes how planned burns early in the dry season, implemented by traditional owners and land managers, are being used to reduce the extent and impact of unplanned fires late in the dry season.