Studying fire weather with ACCESS - BFCRC Legacy · Studying fire weather with ACCESS Robert Fawcett CAWCR High Impact Weather Research 14 May 2013 . The Centre for Australian Weather

The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

Studying fire weather with ACCESS

Robert Fawcett

CAWCR High Impact Weather Research

14 May 2013

www.cawcr.gov.au

The Centre for Australian Weather and Climate Research

A partnership between CSIRO and the Bureau of Meteorology

Acknowledgements

• Co-authors

• Will Thurston, Jeff Kepert and Kevin Tory

High Impact Weather Research

Weather and Environmental Prediction


Docklands, Victoria

• The Earth System Modelling group in CAWCR



What are we trying to accomplish?

• to model the meteorology of days of significant fires (e.g., the Eyre Peninsula on 11 January 2005)

• using the Australian Community Climate and Earth-System Simulator (ACCESS), including the UK Met Office atmospheric model

• at high resolution (0.012° latitude/longitude spacing = 1.2 km approx.)

• and very high resolution (0.004° latitude/longitude spacing = 400 metres approx.)

• EP is the second of 10 cases simulated so far

• to see which aspects of the modelled meteorology might be important for fire behaviour and to see if aspects of the observed fire behaviour can be “explained” by the meteorology

• to see which aspects of the very-high-resolution meteorology are also seen in the lower-resolution simulations

• lower resolution simulations are cheaper and faster to run, and therefore closer to operational implementation

• outputs

• meteorological grids have been provided to project partners (FIRE-DST)

• accompanying documentation (technical reports; conference presentations, posters and papers; journal papers)



Cases simulated so far …

• [4] Melbourne Dust Storm (Victoria) 8 February 1983

• [1] Blue Mountains (NSW) 25 December 2001

• Canberra (ACT) 18 January 2003

• [1] Eyre Peninsula (SA) 11 January 2005

• Boorabbin (WA) 30 December 2007

• [1,3] Black Saturday (Victoria) 7 February 2009

• [2] Margaret River (WA) 23 November 2011

• Dunalley/Tasman Peninsula (Tasmania) 4 January 2013

• Coonabarabran (NSW) 13 January 2013

• Dereel (Victoria) 28 March 2013

Some of these simulations have already been looked at in detail, others represent future work.

[1] “Official” case studies providing grids to project partners [2] Project deliverables from this case study [3] Results from this case study presented at Hobart RAF last year [4] Ash Wednesday (Victoria) 16 February 1983 being simulated by Uni. Melb. team



How do we validate the simulations?

• By comparing the simulations against

• surface observations

• one-minute and thirty-minute automatic weather station (AWS) observations as

available in the computational domains

• upper-air observations

• radiosonde and wind-profile balloon flight observations as available in the

computational domains

• radar data

• where available (minimal for EP, partial for BS)

• satellite observations

• visual, infra-red imagery

• None of these observational data are assimilated into the model

simulations, so the simulation can be validated against completely

independent data.



What aspects of the modelling can we verify?

• timing of the main wind change on 11 January 2005

• forecast maximum temperature

• forecast maximum wind speeds

• upper-level temperature and moisture

• features observed in the radar and satellite imagery if available

Source: Coroner’s report Source: NASA Earth Observatory



Primary wind change on 11 January 2005

134 135 136 137 138

-36.0

-35.0

-34.0

-33.0

Longitude

Latitu

de

Wind change isochrones every half-

hour as analysed (BoM 2005) and

simulated using at 0.012°-resolution

from 11:00 to 14:00 EDT on 11 January

2005.

Bureau of Meteorology

2005: Meteorological

report on the Wangary

and Black Tuesday Fires:

Lower Eyre Peninsula,

10-11 January 2005



Animations

Temperature (°C)

10 15 20 25 30 35 40 45

1600 UTC 10 Jan to 0900 UTC 11 Jan

0230 EDT 10 Jan to 1930 EDT 11 Jan

Wind direction

N E S W N

0.004°-resolution simulation

EP EP



What have we learnt studying these wind changes?

• They come in many different flavours – e.g., rippled over water and

simpler over land – and can interact with each other

Wind direction

N E S W N

Wind direction

N E S W N

EP EP (1230 EDT) (0900 EDT)



• In fact, they can change from one type into another


132 134 136 138 140

-38

-36

-34

-32

-30

Longitude (°)

La

titu

de

(°)

Cross-section moving

along the orange

arrow. Sea-level to

6,000 metres. Colours

show potential

temperature in Kelvin.

Black (grey) contours

show positive

(negative) vertical

velocity in 1.5 m/s

intervals. From the

0.012°-resolution

simulation.

(K)

EP

Longitude (°)

Ele

va

tio

n (

me

tre

s)

133°E 138°E



• In fact, they can change from one type into another


132 134 136 138 140

-38

-36

-34

-32

-30

Longitude (°)

La

titu

de

(°)

Cross-section moving

along the orange

arrow. Sea-level to

6,000 metres. Colours

show potential

temperature in Kelvin.

Black (grey) contours

show positive

(negative) vertical

velocity in 1.5 m/s

intervals. From the

0.012°-resolution

simulation.

(K)

EP

Longitude (°)

Ele

va

tio

n (

me

tre

s)

133°E 138°E




• The wind change is not always a straight line and can have embedded

substructures and secondary wind changes

Wind direction

N E S W N

Forest Fire Danger Index

0 20 40 60 80 100 120 140

BS BS 20090207 0800 UTC (1900 EDT) 20090207 0840 UTC (1940 EDT)

NSW

Vic NSW

Vic

Echuca

Echuca

embedded

vortices




• The wind change can stall and generate an undular bore, with significant

consequences for fires underneath it (including nocturnal intensification of the

fire)

Wind direction

N E S W N

BS

Yarrawonga radar

Radar reflectivity

and modelled

wind direction

around the

Yarrawonga (Vic)

radar. Animation

starts at 1000

UTC (2100

EDT).

Beechworth-

Mudgegonga

fire

NSW

Vic

Echuca




• Their interactions with topography and coastlines can be complicated

Wind direction

N E S W N

DU

142 144 146 148 150 152

-46

-44

-42

-40

Longitude

La

titu

de



Why is this important?

• Wind changes

• can send fires off in new directions and into previously

unburnt areas

• or blow the fire back onto a previously burnt area …

• change narrow-fronted fires into broad-fronted fires

• broad-fronted fires typically propagate faster than narrow-

fronted fires

• can be hazardous to fire-fighters and the general public

• a very large fraction of the fire deaths occur within one

hour of the wind change

• are predictable on the broad synoptic scale

• on the fine scales shown here are beyond the Bureau’s

current operational capabilities

• their prediction in research mode is possible, but the

calculation takes too long to be run in real time with

current resources

The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

Robert Fawcett

High Impact Weather Research

Phone: 03 9669 4296

Email: [email protected]

Web: www.cawcr.gov.au

Thank you www.cawcr.gov.au



Nesting details

• The modelling begins with a

global model run. The second level

is at 0.11° resolution covering all of

Australia and surrounding waters.

Further levels of nesting

increasingly focus on the area of

interest.

• Boundaries of the third (red),

fourth (green) and fifth (blue)

nesting levels for the Black

Saturday model runs.

• Model boundaries are chosen, as

far as is possible, to avoid areas of

elevated topography.

• Comparison test runs also done

at 0.0075° (fifth level).

132 134 136 138 140

-38

-36

-34

-32

-30

Longitude (°)

La

titu

de

(°)

0.036°

0.012°

0.004°

0.0075°

approximate fire location



-20 0 20 40 601

00

05

00

20

01

00

Temperature (°C), Wind speed (m/s)

Pre

ssu

re (

hP

a)

-140 °C -120 °C -100 °C -80 °C -60 °C -40 °C -20 °C

-20 0 20 40 60

10

00

50

02

00

10

0

Temperature (°C), Wind speed (m/s)

Pre

ssu

re (

hP

a)

-140 °C -120 °C -100 °C -80 °C -60 °C -40 °C -20 °C

Above-surface data

Vertical temperature traces for Adelaide Airport at 09:30 EDT (left) and 21:30 EDT

(right) UTC (right) on 11 January 2005, from the 3600-metre resolution model runs.

Temperatures are skewed.

air

temperature

dewpoint

temperature incre

asin

g h

eig

ht

model

obs. wind

speed

09:30 EDT 21:30 EDT



0.0 0.5 1.0 1.5 2.0 2.5 3.0

010

20

30

40

50

018012 (1200m) CEDUNAAMO (-32.13,133.698)

Days since midnight on 20050110 (LCT)

Tem

pera

ture

(°C

), W

indspeed (

m/s

)

Tmax bias: 1.75

Dewpoint temperature

Wind speed

Air temperature

Wind direction

00:00 12:00 00:00 12:00 00:00 12:00 00:00 CDT

10 Jan 10 Jan 11 Jan 11 Jan 12 Jan 12 Jan 13 Jan

Surface data – Ceduna Airport T

em

pe

ratu

re (

°C),

Win

d s

pe

ed

(m

/s)

0 1

0 2

0

3

0

4

0

5

0

model

obs

obs

model

0.012°-resolution

simulation

Documents

Studying fire weather with ACCESS - BFCRC Legacy · Studying fire weather with ACCESS Robert Fawcett CAWCR High Impact Weather Research 14 May 2013 . The Centre for Australian Weather