ArticlePDF Available

Multi-decadal variability of forest fire risk - Eastern Australia

Authors:

Abstract and Figures

This study investigates the influence that the El Nio/Southern Oscillation (ENSO) and the Inter-decadal Pacific Oscillation (IPO) have on long term daily weather conditions pertinent to high forest fire danger in New South Wales, Australia. Using historical meteorological data for 22 weather stations to compute the daily value of McArthur's Forest Fire Danger Index (FFDI), it is shown that a strong relationship exists between climate variability, on a range of time scales, and forest fire risk. An investigation into the influence of ENSO on fire risk demonstrates that the proportion of days with a high, or greater than high, fire danger rating is markedly increased during El Nino episodes. More importantly, this study also shows that the already significantly enhanced fire danger associated with El Nino events was even further increased during El Nino events that occurred when the IPO was negative. The potential to use simple indices of climate variability to predict forest fire risk is therefore demonstrated to be significant.
Content may be subject to copyright.
CSIRO PUBLISHING
www.publish.csiro.au/journals/ijwf International Journal of Wildland Fire, 2004, 13, 165–171
Multi-decadal variability of forest fire risk—easternAustralia
Danielle C. VerdonA,B,Anthony S. KiemAand Stewart W. FranksA
ASchool of Engineering, University of Newcastle, NSW 2308,Australia.
BCorresponding author. Telephone: +61 2 4921 6058; fax: +61 2 4921 6991;
email: danielle.verdon@studentmail.newcastle.edu.au
Abstract. This study investigates the influence that the El Niño/Southern Oscillation (ENSO) and the Inter-decadal
Pacific Oscillation (IPO) have on long term daily weather conditions pertinent to high forest fire danger in New
South Wales, Australia. Using historical meteorological data for 22 weather stations to compute the daily value of
McArthur’s Forest Fire Danger Index (FFDI), it is shown that a strong relationship exists between climate variability,
on a range of time scales, and forest fire risk. An investigation into the influence of ENSO on fire risk demonstrates
that the proportion of days with a high, or greater than high, fire danger rating is markedly increased during El Niño
episodes. More importantly, this study also shows that the already significantly enhanced fire danger associated
with El Niño events was even further increased during El Niño events that occurred when the IPO was negative.
The potential to use simple indices of climate variability to predict forest fire risk is therefore demonstrated to be
significant.
Additional keywords: El Niño/Southern Oscillation (ENSO); Inter-decadal Pacific Oscillation (IPO); Pacific
Decadal Oscillation (PDO); bushfire; climate variability.
Introduction
The incidence of forest fires in Australia poses a significant
threat to lives and property. Devastating forest fires have been
found to occur during specific meteorological conditions.
‘Fire weather’ is a convenient term to use when reference
is made to the effect of climate and weather conditions on
the chances of a fire starting, its behaviour and the diffi-
culty of suppression. The weather variables that generally
increase the risk of forest fires are low precipitation and rel-
ative humidity combined with high temperature and wind
speed (Luke and McArthur 1986). The high variability of
rainfall and temperature in northern and eastern Australia
has been strongly associated with the regional influence of
the El Niño/Southern Oscillation or ENSO (e.g. Ropelewski
and Halpert 1987; Allan 1988; Stone and Auliciems 1992;
Kiem and Franks 2001). These studies indicate that large-
scale climate variability on an annual/interannual time-scale
influences at least two of the four weather variables that
contribute to forest fire risk in Australia.
In addition to the annual/interannual effects of ENSO, a
number of studies have also examined decadal and longer
scale climate variability. Observational records suggest that
the Inter-decadal Pacific Oscillation (IPO) modulates the
strength and nature of the ENSO cycle (Power et al. 1998,
1999; Folland et al. 1999; Allan 2000). The IPO is the
coherent pattern of sea surface temperature variability occur-
ring on inter-decadal time scales over the Pacific Ocean
and is similar to the Pacific Decadal Oscillation or PDO
(Mantua et al. 1997; Franks 2002a). Importantly, Power et al.
(1999) showed that individual ENSO events (i.e. El Niño and
La Niña) had stronger impact across Australia during the
negative phase of the IPO. Furthermore it has recently been
shown that, in addition to influencing the magnitude of ENSO
impacts, the IPO also appeared to modulate the frequency
of extreme ENSO events, leading to multi-decadal periods
of elevated flood or drought risk depending on the phase
of the IPO (Franks 2002b; Franks and Kuczera 2002; Kiem
and Franks 2004; Kiem et al. 2003). However, whilst the
relationship between ENSO and fire risk has been studied
briefly (Love and Downey 1986; Williams and Karoly 1999),
the influence that decadal/multi-decadal climate variability
has on forest fire risk has not been investigated prior to this
study.
In this study the Forest Fire Danger Index (FFDI), devel-
oped by McArthur (1967), is used to assess the daily risk of
forest fire in New South Wales (NSW), Australia. The influ-
ence of ENSO on NSW forest fire risk is then evaluated by
comparing the proportion of days with a high (or greater) fire
danger rating based on the FFDI, during El Niño and non-
El Niño years. In addition, multi-decadal IPO enhancement
© IAWF 2004 10.1071/WF03034 1049-8001/04/020165
166 D. C. Verdon et al.
Bourke
Coonabarabran
Dubbo
Glen Innes
Tamworth
Lismore
Ya m b a
Williamtown
Cessnock
Nobbys
Bathurst
Sydney
Nowra
Coffs Harbour
Port Macquarie
Hay
Wagga
Wagga
Burrinjuck Dam
Canberra
Moruya Heads
Merimbula
Cooma
Fig. 1. Location of the 22 study sites within New South Wales,
Australia.
of El Niño events, and hence elevated forest fire risk, is also
assessed. This is achieved by comparing the fraction of days
with a high (or greater) fire danger rating during IPO negative
El Niño years with the fraction of high (or greater) fire risk
days in all other El Niño years.
Data
Daily data records for precipitation, maximum temperature,
dewpoint temperature and average wind speed were obtained
from the Australian Bureau of Meteorology for 22 weather
stations located across NSW (Fig. 1). Daily relative humidity
is then calculated from the daily maximum temperature and
the corresponding dew-point temperature. The selection of
the study sites is primarily based on the length of meteoro-
logical data available. It is also desirable to obtain a number
of stations that are positioned so as to observe spatial trends
across NSW. As can be seen from Fig. 1, very few stations
are available in the middle and western thirds of NSW.
Definition of fire danger
The fire danger rating employed in this study is based on
the FFDI, which is designed for general fire danger fore-
casting purposes and is in common use throughout eastern
Australia. The FFDI provides an assessment of the chances of
a fire starting, its rate of spread, fire intensity and difficulty
of suppression (Noble et al. 1980). It is important to note
that the FFDI does not provide a measure of actual forest
fire events, but rather the climatological conditions associ-
ated with severe fire danger. The fire danger rating for any
given day is determined based on the daily value of the FFDI,
as shown in Table 1. Throughout the remainder of this paper
the term ‘high’ fire danger will be taken to mean high, very
high or extreme fire danger (i.e. FFDI >12).
Table 1. Fire danger ratings based on the Forest
Fire Danger Index (FFDI)
Forest fire danger index Fire danger rating
0–5 Low
5–12 Moderate
12–24 High
24–50 Very high
50–100 Extreme
The daily FFDI value is calculated using the equations
provided by Noble et al. (1980) and shown below:
FFDI =2.0×exp(0.450 +0.987 ×ln(D) 0.0345
×H+0.0338 ×T+0.0234 ×V), (1)
where His the minimum relative humidity in percent,Tis the
maximum air temperature in 24 h in C,Vis the average wind
velocity (km h1) in the open at a height of 10 m, and Dis the
drought factor for that day. The drought factor (D) is a discon-
tinuous variable obtained using the Keetch-Byram Drought
Index (KBDI, Keetch and Byram 1968). The equation used
to derive the drought factor is shown below:
D=0.191 ×(I +104)×(N +1)1.5
(3.52 ×(N +1)1.5+P1),(2)
where Iis the daily KBDI, Pis the daily precipitation (in
mm), and Nis the number of days since rain.
Methods
To determine the probability of daily ‘high’ fire risk in any
year, regardless of ENSO or IPO, the proportion of days
between September and February, inclusive, with a fire dan-
ger rating of ‘high’ (FFDI>12) is calculated for each station
over the period that data are available. Only the spring and
summer months (September to February) are investigated,
as this is generally the period of peak forest fire activity for
NSW (Luke and McArthur 1986).
The impact of ENSO on forest fire risk is assessed by clas-
sifying all years, and the corresponding proportions of days
with a ‘high’ fire danger rating, as either El Niño, La Niña
or Neutral based on the 6-month October to March average
Multivariate ENSO Index (MEI) value. The MEI, developed
by Wolter and Timlin (1993, 1998) is derived from multiple
climate parameters and has been shown to reflect the nature
of the coupled ocean–atmosphere system better than either
the Southern Oscillation Index or sea surface temperature
based indices (Kiem and Franks 2001). The probability of
daily ‘high’fire risk during El Niño years is then determined
by calculating the proportion of days with FFDI>12 for all
El Niño years during the September to February forest fire
season at each study site. A similar process is performed
to determine the probability of daily ‘high’ fire risk during
non-El Niño years and the results are compared.
Multi-decadal variability of forest fire risk 167
2
1.5
1
0.5
0
0.5
1
1.5
2
IPO
1920 1940 1960 1980 2000
Year
Fig. 2. The Inter-decadal Pacific Oscillation (IPO) from 1920 to 1999.
The probability of daily ‘high’ fire danger (FFDI >12) in
El Niño years occurring in the negative IPO periods, when
ENSO impacts are enhanced, is then compared with the prob-
ability of ‘high’ fire danger in all other El Niño years to
determine whether there is a significant difference. To achieve
this the El Niño years, and the corresponding proportions
of days with a ‘high’ fire danger rating, are further strati-
fied based on the IPO and the method used by Power et al.
(1999). In classifying the different IPO phases, Power et al.
(1999) used the thresholds of ±0.5 to distinguish positive,
neutral and negative IPO phases. Figure 2 shows the time
series of the IPO from 1920 to 1999. During this period there
have been three major phases of the IPO:Two positive phases
(IPO >0.5) between 1924–1943 and 1979–1997 and a neg-
ative phase (IPO<–0.5) from 1946 to 1976. These phases
exclude the 10 years from 1958 to 1967 when the absolute
value of the IPO index was less than 0.5. Only 14 of the
22 study sites had sufficiently long records of ‘fire weather’
variables for the IPO analysis to be performed.
To further investigate the magnitude of fire risk associated
with IPO-negative El Niño years, due to the IPO induced
amplification of El Niño impacts (Power et al. 1999), the
probability of daily ‘high’fire danger in IPO negative El Niño
years and all non-El Niño years is also compared.
Results
Figure 3 shows the probability of having a day associated with
a ‘high’ fire danger rating (FFDI>12) during the September
to February forest fire season at each of the 22 study sites.
Figure 3 demonstrates that some degree of ‘high’ fire risk
exists at all stations; however, the magnitude of risk varies
considerably. The results show that the occurrence of days
with a ‘high’ fire risk rating is generally lower along the
coast than it is at the inland stations. This is because rain-
fall totals and relative humidity tend to be higher at coastal
locations while temperatures are generally lower than they are
inland. The low probability of ‘high’fire risk obser ved at Glen
Innes (altitude 1062 m) also suggests that increased eleva-
tion reduces the risk of forest fire, for non-coastal stations.
Bourke (77%) Glen Innes
(6%)
Coonabarabran
(37%)
Tamworth (39%)
Dubbo
(42%)
Bathurst
(15%)
Hay (58%)
Wagga Wagga
(30%)
Burrinjuck Dam
(19%)
Cooma (17%)
Canberra (12%)
Moruya Heads
(5%)
Merimbula (4%)
Cessnock
(28%)
Nowra (7%)
Sydney (6%)
Nobbys (10%)
Williamtown (8%)
Port Macquarie (2%)
Coffs Harbour (2%)
Yamba (2%)
Lismore (11%)
Legend
1–20%
21–40%
41–60%
61–80%
81–100%
Fig. 3. Percentage of days with ‘high’fire risk (FFDI >12) at each of
the 22 study sites.
This is due to rainfall and humidity levels being increased at
higher altitudes, whilst temperatures tend to be lower (Tapper
and Hurry 1993).
The probability of daily ‘high’fire risk during El Niño and
non-El Niño years is examined to determine the impact that
ENSO has on fire risk at each study sites. The average propor-
tion of days with ‘high’ fire risk during El Niño and non-El
Niño years (i.e. La Nina or Neutral) are compared to deter-
mine whether there is any significant difference between El
Niño and non-El Niño phases. To test this hypothesis a sim-
ple test of proportions is applied (Hogg and Tanis 1988). It
is assumed that the sampling distribution of the proportion
of El Niño days with ‘high’fire danger can be approximated
by a normal distribution with a mean of pand a variance
of p(1p)/n, where pis the proportion of El Niño days
with FFDI greater than 12, calculated using p=y/n, where
yis the number of El Niño days with FFDI greater than 12
and nis the total number of El Niño days for the period
being investigated.The same calculations were performed for
non-El Niño events.
In order to determine whether the probability (P1)of
having a day in an El Niño with FFDI greater than 12 is
significantly different from the probability (P2) of having a
day in a non-El Niño year with FFDI greater than 12, the
following statistical test is used (Hogg and Tanis 1988). Let
y1represent the number of El Niño days with FFDI greater
than 12 that occurred in the n1El Niño days from the period
being investigated. Let y2be the number of non-El Niño days
with FFDI greater than 12 that occurred in the n2non-El Niño
days during the same period.The test statistic used to test the
hypothesis that P1equals P2is:
z=|p1p2|
p(1p)(1/n1+1/n2),(3)
where p1=y1/n1,p
2=y2/n2,p=(y1+y2)/(n1+n2)and
zN(0,1).
168 D. C. Verdon et al.
Table 2. Results obtained when the daily probability of‘high’ fire risk during El Niño years (p1) is compared
to non-El Niño years (p2)
The P-value indicates the probability that p1is equal to p2with significance at the <5% and <1% level represented
by * and ** respectively
Station Probability of ‘high’ fire danger Probability of ‘high’ fire danger p1/p2P-value
in El Niño years (p1) in non-El Niño years (p2)
Bathurst 0.186 0.134 1.387 0.000**
Bourke 0.846 0.748 1.131 0.000**
Burrinjuck Dam 0.236 0.169 1.402 0.000**
Canberra 0.206 0.093 2.212 0.000**
Cessnock 0.374 0.247 1.510 0.000**
Coffs Harbour 0.028 0.020 1.349 0.026*
Cooma 0.228 0.143 1.591 0.000**
Coonabarabran 0.449 0.331 1.356 0.000**
Dubbo 0.524 0.384 1.367 0.000**
Glen Innes 0.069 0.059 1.175 0.057
Hay 0.633 0.562 1.126 0.000**
Lismore 0.134 0.094 1.422 0.000**
Merimbula 0.080 0.030 2.699 0.005**
Moruya Heads 0.069 0.037 1.856 0.000**
Nobbys 0.125 0.094 1.328 0.001**
Nowra 0.100 0.056 1.790 0.000**
Port Macquarie 0.026 0.020 1.292 0.059
Sydney 0.091 0.054 1.698 0.000**
Tamworth 0.487 0.350 1.392 0.000**
Wagga Wagga 0.391 0.274 1.426 0.000**
Williamtown 0.112 0.074 1.519 0.000**
Yamba 0.024 0.022 1.114 0.256
Table 2 shows, for each of the 22 study sites, the proportion
of days with ‘high’ fire risk during El Niño (p1) and non-El
Niño (p2) years, with the probability that these proportions
are equal indicated by the p-value. To further illustrate the
effect of ENSO on fire risk, the ratio of the probability of
‘high’ fire risk during El Niño years to non-El Niño years
is also calculated. This represents the increase in probability
of a ‘high’ fire danger day occurring during an El Niño year
when compared to a non-El Niño year. For example a p1/p2
value of 1.5 demonstrates that the chance of a ‘high’ fire
risk day occurring in an El Niño year is 50% greater than it
is in a non-El Niño year. The ratio of p1/p2for each study
site is also shown in Table 2 and the percentage increase in
the probability of daily ‘high’ fire risk during an El Niño is
displayed in Fig. 4.
It can be seen from Table 2 that the probability of daily
‘high’ fire danger during El Niño years is significantly differ-
ent from non-El Niño years at the <1% level for all stations
except Coffs Harbour, Glen Innes, Port Macquarie and Yamba
(with the difference at Coffs Harbour significant at the <5%
level). Figure 4 demonstrates that the probability of daily
‘high’ fire danger is markedly increased during El Niño years
at all stations investigated. The increase in the probability of
‘high’ fire danger, when El Niño years are compared to all
other years, averaged across the 22 study sites is 51%. The
influence of El Niño appears to be strongest in the south-
eastern corner of NSW and weaker in the west, with Canberra
Bourke (13%)
Dubbo
(37%)
Coonabarabran
(36%)
Bathurst
(39%)
Hay (13%)
Wagga Wagga
(43%)
Burrinjuck Dam
(40%)
Cooma (59%)
Merimbula (170%)
Moruya Heads
(86%)
Nowra (79%)
Sydney (70%)
Cessnock
(51%)
Tamworth (39%)
Glen Innes
(18%)
Lismore (42%)
Yamba (11%)
Coffs Harbour
(35%)
Port Macquarie
(29%)
Williamtown (52%)
Nobbys (33%)
Legend
21–40%
41–60%
61–80%
81–100%
100%
1–20%
Canberra (121%)
Fig. 4. Percentage increase in the probability of daily ‘high’fire risk
during El Niño years when compared to non-El Niño years.
and Merimbula displaying an increase in forest fire risk of
121% and 170% respectively, during El Niño years compared
to 13% at Bourke and Hay.
Table 3 shows, for the 14 study sites whose ‘fire weather’
data records spanned both negative and non-negative IPO
phases, the proportion of days with ‘high’ fire risk during
El Niño years occurring when the IPO is negative (p1) and
during all other El Niño years ( p2), with the probability that
Multi-decadal variability of forest fire risk 169
Table 3. Results obtained when the daily probabilityof ‘high’ fire risk during El Niño years occurring
in the negative IPO phase (p1) is compared to all other El Niño years (p2)
The P-value indicates the probability that p1is equal to p2with significance at the <5% and <1% level
represented by * and ** respectively
Station Probability of ‘high’ Probability of ‘high’ p1/p2P-value
fire danger in negative fire danger in all other-
IPO El Niño years (p1) El Niño years (p2)
Bourke 0.991 0.779 1.271 0.000**
Canberra 0.232 0.198 1.172 0.044*
Coffs Harbour 0.052 0.020 2.630 0.000**
Coonabarabran 0.527 0.423 1.246 0.000**
Dubbo 0.547 0.517 1.059 0.109
Hay 0.676 0.619 1.093 0.008**
Lismore 0.173 0.119 1.456 0.001**
Moruya Heads 0.098 0.059 1.659 0.001**
Nobbys 0.182 0.105 1.730 0.000**
Nowra 0.155 0.080 1.938 0.000**
Port Macquarie 0.042 0.020 2.095 0.002**
Sydney 0.142 0.075 1.897 0.000**
Tamworth 0.514 0.479 1.074 0.077
Yamba 0.044 0.018 2.487 0.000**
Bourke (27%)
Coonabarabran
(25%)
Dubbo
(6%)
Hay (9%)
Canberra
(17%)
Moruya Heads
(66%)
Nowra (94%)
Sydney
(90%)
Nobbys (73%)
Port Macquarie
(109%)
Coffs Harbour
(163%)
Yamba (149%)
Lismore (46%)
Tamworth (7%)
Legend
21–40%
41–60%
61–80%
81–100%
100%
1–20%
Fig. 5. Percentage increase in the occurrence of daily ‘high’ fire risk
associated with El Niño and IPO negative years compared to all other
El Niño years.
these proportions are equal indicated by the p-value.The ratio
p1/p2is again calculated to show the increase in fire risk
associated with El Niño events that occur when the IPO is
negative. Figure 5 displays the percentage increase in the
probability of daily ‘high’ fire risk during IPO negative El
Niño years when compared with all other El Niño years.
Table 3 demonstrates that the proportion of days with
‘high’ fire risk during El Niño years that occur when the IPO
is negative is significantly different (at the <1% level) from
the proportion during all other El Niño years at 11 of the 14
stations studied (with the difference at Canberra significant
at the <5% level). Figure 5 shows that the probability of
‘high’ fire risk increases during IPO negative El Niño years
at all stations studied.
Table 4 shows, for the 14 study sites, the proportion of days
with ‘high’ fire risk during IPO negative El Niño years (p1)
and during all non-El Niño years (p2), with the probability
that these proportions are equal indicated by the p-value. The
ratio p1/p2is again calculated to show the increase in fire
risk associated with El Niño events that occur when the IPO
is negative. Figure 6 displays the percentage increase in the
probability of daily ‘high’ fire risk during IPO negative El
Niño years when compared with non-El Niño events.
Table 4 demonstrates that the proportion of days with
‘high’ fire risk during IPO negative El Niño years is sig-
nificantly different (at the <1% level) from the proportion
during non-El Niño years at all stations studied.
Figure 6 demonstrates that, for all stations studied, the
increase in the probability of a ‘high’ fire danger day
occurring when IPO negative El Niño events and non-El
Niño events are compared is even greater than the increase
observed when all El Niño events are compared to non-
El Niño events. The average increase in the risk of ‘high’
fire danger across the 14 study sites during IPO negative El
Niño events is 95% (which is much greater than the average
increase of 51% that was obtained when all El Niño and non-
El Niño events were compared).The influence of the IPO on
El Niño events appears to be strongest along the east coast
of NSW and in the south-eastern corner, and weaker in the
west.Yamba, Port Macquarie, Sydney, Canberra, Nowra and
Moruya Heads all display an increase in forest fire risk of
greater than 100% during IPO-negative El Niño years. The
average increase in coastal areas is 122% compared with an
average of 58% for the six inland stations.
170 D. C. Verdon et al.
Table 4. Results obtained when the daily probabilityof ‘high’ fire risk during El Niño years occurring
in the negative IPO phase (p1) is compared to non-El Niño years (p2)
The P-value indicates the probability that p1is equal to p2with significance at the <5% and <1% level
represented by * and ** respectively
Station Probability of ‘high’ Probability of ‘high’ p1/p2P-value
fire danger in negative fire danger in non-El
IPO El Niño years (p1) Niño years (p2)
Bourke 0.991 0.748 1.325 0.000**
Canberra 0.232 0.093 2.487 0.000**
Coffs Harbour 0.115 0.065 1.777 0.000**
Coonabarabran 0.527 0.331 1.592 0.000**
Dubbo 0.547 0.384 1.426 0.000**
Hay 0.676 0.562 1.202 0.000**
Lismore 0.173 0.094 1.839 0.000**
Moruya Heads 0.098 0.037 2.644 0.000**
Nobbys 0.182 0.094 1.944 0.000**
Nowra 0.155 0.056 2.781 0.000**
Port Macquarie 0.042 0.020 2.126 0.000**
Sydney 0.142 0.054 2.632 0.000**
Tamworth 0.514 0.350 1.468 0.000**
Yamba 0.044 0.022 2.020 0.000**
Bourke (32%)
Coonabarabran
(59%)
Dubbo
(43%)
Hay (20%)
Canberra
(149%)
Moruya Heads
(164%)
Nowra
(178%)
Sydney
(163%)
Nobbys (94%)
Port Macquarie
(113%)
Coffs Harbour
(78%)
Yamba (102%)
Lismore (84%)
Tamworth
(47%)
Legend
21–40%
41–60%
61–80%
81–100%
100%
1–20%
Fig. 6. Percentage increase in the occurrence of daily ‘high’ fire risk
associated with El Niño and IPO negative years compared to non-El
Niño years.
Conclusions
This paper has sought to determine how observed multi-
temporal scale climate variability influences the risk of forest
fire in NSW, Australia. This was achieved by first analysing
the overall risk (regardless of ENSO or IPO phase) of forest
fire for 22 study sites located within NSW. The impact that
ENSO has on forest fire risk, and how this risk is modulated
by the IPO, was then assessed.
The results showed that the probability of a ‘high’fire risk
day (FFDI >12) occurring is much lower on the coast than
it is inland. The results also showed that the risk of forest
fire is significantly increased during El Niño events at all
sites investigated, with on average a 51% increase in the pro-
portion of days with a ‘high’fire danger rating when El Niño
years are compared to non-El Niño years.The strongest influ-
ence of El Niño appears to be in the south-eastern corner of
NSW, with Canberra and Merimbula displaying increases in
the probability of a ‘high’ risk fire day of 121% and 170%
respectively.
Further investigation into how the forest fire risk during
El Niño years is modulated by the IPO revealed that El Niño
years occurring when the IPO is negative are associated with
an increased risk of forest fire when compared to all other El
Niño years for all 14 stations studied. The average increase in
the probability of daily ‘high’fire risk, across the 14 stations
investigated, when IPO negative El Niño years are compared
with all other El Niño years is 63%. This increase in risk is
statistically significant at 12 of the 14 stations investigated.
This is an important result because, as demonstrated, the risk
of forest fire during El Niño years is already significantly
increased when compared to non-El Niño years.This implies
an extremely high risk of forest fire during IPO negative El
Niño years, especially along the east coast and in the south-
eastern corner of NSW.
To further illustrate the IPO induced enhancement of El
Niño impacts when the IPO is negative, the risk of ‘high’
fire danger in IPO negative El Niño years was compared to
the risk in non-El Niño years. The increase in the probability
of daily ‘high’ fire danger when IPO negative El Niño are
compared with non-El Niño years was found to be statistically
significant (at the <1% level) at all stations investigated with
an average increase across all stations of 95%, which is almost
twice the average increase when all El Niño years and non-El
Niño years were compared.
Multi-decadal variability of forest fire risk 171
A likely example of the enhancement of El Niño impacts
in the negative IPO phase might be found in the severe forest
fires experienced in much of NSW during the spring and
summer of 2002–2003. The 2002–2003 event was the first
El Niño in the current negative IPO phase. Despite this event
being of only ‘moderate’strength, with respect to sea surface
temperature and atmospheric pressure anomalies, the impacts
associated with it were far worse than other ‘stronger’El Niño
events that occurred when the IPO was not negative.
This study reveals the potential to use long-term climate
variability insight to predict forest fire risk in NSW. It is now
possible to predict ENSO events up to 9 months in advance
(Kiem and Franks 2001). The implications of this are par-
ticularly relevant to the Australian climate. NSW has been
plagued with devastating forest fires since colonisation and,
whilst it is impossible to eradicate fires of this nature, it is
possible to be prepared. This preparation is made possible
with the advantage of a 9-month lead-time on an upcoming
serious fire season, and the possibility that the IPO may be a
feature of the climate system with decadal to multi-decadal
persistence.
Acknowledgements
The authors acknowledge the Australian Bureau of Mete-
orology for providing the meteorological data and the UK
Meteorological Office for kindly making the IPO data used
in this study available.
References
Allan RJ (1988) El Niño-Southern Oscillation influences on the
Australasian region. Progress in Physical Geography 12, 4–40.
Allan RJ (2000) ENSO and climatic variability in the last 150 years.
In ‘El Niño and the Southern Oscillation: Multi-scale variability,
global and regional impacts’. (Eds HF Diaz, V Markgraf ) pp. 3–56.
(Cambridge University Press: Cambridge, UK)
Folland CK, Parker DE, ColmanAW, Washington R (1999) Large scale
modes of ocean surface temperature since the late nineteenth century.
In ‘Beyond El Niño: Decadal and interdecadal climate variability’.
(Ed. A Navarra) pp. 73–102. (Springer: Berlin)
Franks SW (2002a) Assessing hydrological change: deterministic gen-
eral circulation models or spurious solar correlation? Hydrological
Processes 16, 559–564. doi:10.1002/HYP.600
Franks SW (2002b) Identification of a change in climate state using
regional flood data. Hydrology & Earth System Sciences 6, 11–16.
Franks SW, Kuczera G (2002) Flood frequency analysis: evidence and
implications of secular climate variability, New South Wales.Water
Resources Research 38(5), 1062. doi:10.1029/2001WR000232
http://www.publish.csiro.au/journals/ijwf
Hogg RV, Tanis EA (1988) ‘Probability and statistical inference (3rd
edn).’ (Macmillan Publishing: NewYork)
Keetch JJ, Byram GM (1968) ‘A drought index for forest fire control.’
USDA Forest Service, Research Paper SE-38. (Asheville, NC)
Kiem AS, Franks SW (2001) On the identification of ENSO-induced
rainfall and runoff variability: A comparison of methods and
indices. Hydrological Sciences Journal—Journal des Sciences
Hydrologiques 46, 715–727.
Kiem AS, Franks SW (2004) Multi-decadal variability of drought risk–
eastern Australia. Hydrological Processes 18. doi:10.1002/hyp.1460
Kiem AS, Franks SW, Kuczera G (2003) Multi-decadal variabil-
ity of flood risk. Geophysical Research Letters 30. doi:10.1029/
2002GL015992
Love G, Downey A (1986) The prediction of bushfires in central
Australia. Australian Meteorological Magazine 34, 93–101.
Luke RH, McArthur AG (1986) ‘Bushfires in Australia (2nd edn).’
(Australian Government Publishing Service: Canberra)
Mantua NJ, Hare SR, ZhangY, Wallace JM, Francis RC (1997) A Pacific
interdecadal climate oscillation with impacts on salmon production.
Bulletin of the American Meteorological Society 78, 1069–1079.
doi:10.1175/1520-0477(1997)078<1069:APICOW>2.0.CO;2
McArthur AG (1967) ‘Fire behaviour in eucalypt forest.’ Leaflet No.
107. (Commonwealth of Australia Forestry and Timber Bureau:
Canberra)
Noble IR, Bary GA, Gill AM (1980) McArthur’s fire-danger meters
expressed as equations. Australian Journal of Ecology 5, 201–203.
Power S, Casey T, Folland C, Colman A, Mehta V (1999) Inter-decadal
modulation of the impact of ENSO on Australia. Climate Dynamics
15, 319–324. doi:10.1007/S003820050284
Power S, Tseitkin F, Torok S, Lavery B, Dahni R, McAvaney B
(1998) Australian temperature,Australian rainfall and the Southern
Oscillation, 1910–1992—coherent variability and recent changes.
Australian Meteorological Magazine 47, 85–101.
Ropelewski CF, Halpert MS (1987) Global and regional scale pre-
cipitation patterns associated with the El Niño/Southern Oscilla-
tion. Monthly Weather Review 115, 1606–1626. doi:10.1175/1520-
0493(1987)115<1606:GARSPP>2.0.CO;2
Stone R, Auliciems A (1992) SOI phase relationships with rainfall in
eastern Australia. International Journalof Climatology 12, 625–636.
TapperN, Hurry L (1993) ‘Australia’s weather patterns—an introductory
guide.’ (Dellasta: Mount Waverly, Victoria)
Williams AA, Karoly DJ (1999) Extreme fire weather in Australia
and the impact of the El Niño Southern Oscillation. Australian
Meteorological Magazine 48, 15–22.
Wolter K, Timlin MS (1993) Monitoring ENSO in COADS with a sea-
sonally adjusted principal component index. In ‘Proceedings of the
17th Climate Diagnostics Workshop’. pp. 52–57. (Norman, OK)
Wolter K, Timlin MS (1998) Measuring the strength of ENSO—how
does 1997/98 rank? Weather 53, 315–324.
... The relationship between fire occurrence and the climate factors in Australia has been explored in the literature (Verdon et al. 2004;Russell-Smith et al. 2007;Mariani et al. 2016;Dowdy 2018). In particular, the most important Australian climate drivers are the El Niño Southern Oscillation (ENSO), Indian Ocean Dipole (IOD) and Southern Annular Mode (SAM), causing spatio-temporal variations of temperature and rainfall (Hendon et al. 2007;Risbey et al. 2009), and therefore affecting the Australian fire behaviour. ...
... The fire weather index is a common methodology used to assess the fire danger and to explore the effects of climatic indices in the Australian fire danger. In particular, several studies have adopted the McArthur Forest Fire Danger Index (FFDI) to find evidences of the variability of fire weather in Australia, and also to assess the linkages between climate drivers and FFDI (Verdon et al. 2004;Clarke et al. 2013;Harris and Lucas 2019). However, there is an important drawback related to the fire weather index methods. ...
... Indeed, since the decadal climate variability and climate change overlap, based on a short observed time series, it is difficult to distinguish between the two effects. Changes in fire weather over longer time scales have been widely discussed in the literature, and it have been associated with anthropogenic climate changes (Dowdy et al. 2016;Harris and Lucas 2019), but also with climate variability (e.g., Interdecadal Pacific Oscillation) (Verdon et al. 2004). On the other hand, it is worth noting that, for longer time series, our proposed model could be considered as an important tool to identify the distinct effects from climate change and climate variability, given the model ability to capture persistent and mean-reverting (seasonal and components with irregular periodicity) terms, taken into account the effects of covariates and the spatial heterogeneity. ...
Article
Full-text available
Fire is one of the most notorious hazards in Australia, with important economic impacts and damage to ecosystems. There is a concern of worsening fire conditions under climate variability, but there is little understanding of the variability in fire occurrence related to climate patterns. We present a statistical decomposition for spatio-temporal analysis of changes in fire occurrence in Australia and its association with climate factors. We found evidence of variability in the trend results for fire occurrence, and also some evidence that this variation is related to climate patterns. Our approach has applicability to other climate-related issues, providing a useful tool to identify possible changes in the intensity of occurrence over time, capturing long-term changes, and also seasonal and cyclical effects.
... T he relationship between Pacific decadal variability (PDV) and climate impacts such as Pacific Basin drought, flood, wildfire and tropical cyclone risk and global temperature has been extensively studied [1][2][3][4][5][6][7][8][9][10][11] . More recently, PDV has been implicated in pronounced changes in high latitude Southern Hemisphere and Antarctic climate, including abrupt declines in annual sea ice extent 12 and regional Antarctic surface air temperature warming 13 and cooling 14 , which has implications for global climate and ice mass balance projections. ...
... First, consistent with the previous literature [3][4][5] , eastern Australian rainfall is significantly lower during positive IPO years than during negative IPO years (Fig. 4). The statistical probability of this difference varies by region, with the states of New South Wales and Queensland having the lowest probability that the distributions are the same. ...
Article
Full-text available
The Interdecadal Pacific Oscillation, an index which defines decadal climate variability throughout the Pacific, is generally assumed to have positive and negative phases that each last 20-30 years. Here we present a 2000-year reconstruction of the Interdecadal Pacific Oscillation, obtained using information preserved in Antarctic ice cores, that shows negative phases are short (7 ± 5 years) and infrequent (occurring 10% of the time) departures from a predominantly neutral-positive state that lasts decades (61 ± 56 years). These findings suggest that Pacific Basin climate risk is poorly characterised due to over-representation of negative phases in post-1900 observations. We demonstrate the implications of this for eastern Australia, where drought risk is elevated during neutral-positive phases, and highlight the need for a re-evaluation of climate risk for all locations affected by the Interdecadal Pacific Oscillation. The initiation and future frequency of negative phases should also be a research priority given their prevalence in more recent centuries.
... A Wilcoxon rank sum test (non-parametric test for randomness) was used to identify statistically significant trends in TC counts over time (Chiew and Siriwardena 2005;Lavender and Abbs 2013). A test of proportions was used to identify the statistical significance of TC occurrence in combined phases of SOI and IPO (Hogg and Tanis 1988;Verdon et al. 2004). Wavelet analysis was used to identify climate anomalies in the TC time series based on a time and frequency relationship (Jänicke et al. 2009). ...
... Less common events occurred during neutral phases, IPO positive and SOI negative (El Niño) combinations. The statistical significance of the probability of TC occurrence during the different phase combinations of SOI and IPO was tested using a test of proportions (Hogg and Tanis 1988;Verdon et al. 2004). For this test the sampling distribution was assumed to be normal with the mean equal to p, the variance equal to p(1 -p)/n, with p being the proportion of TC events that occur during a particular combined SOI/IPO phase and n being the total number of TC events in the TC database. ...
Article
Full-text available
It is well known that severe storms result in some of the costliest natural disasters for New South Wales (NSW), Australia. However, it is not widely acknowledged that some of these events are, in fact, a result of landfalling tropical cyclones (TCs). Indeed, the intense focus of TC research within the tropics generally disregards landfalling TC events in the mid-latitude regions of Australia. This is likely due to the perceived infrequency of these events compared to other more susceptible regions. Therefore, in this study, we review this assumption by developing a 150-year record of TC activity, based on a range of digitised and analogue historical datasets and identify 30 individual landfalling TCs that have impacted NSW. Periods of enhanced and reduced TC activity are observed, with a defined hiatus (absence of landfalling TCs) after approximately 1980. The recent decrease in TC activity is subsequently linked to an increase in El Niño activity and warming of northwest Australian sea-surface temperatures during this time. Importantly, it is possible that a return to enhanced TC activity could occur again in the future if the Pacific conditions align. We also propose that pre-instrumental data on TC activity need to be developed to appropriately quantify TC risk for the study region via the development of local palaeoclimate archives. This study provides a significant contribution to understanding the risks of NSW landfalling TCs and expands upon our knowledge of environmental conditions that influence landfalling TCs in NSW.
... Dryness, on the other hand, strongly increases the availability of the vegetation to burn [28]. It could be shown that drought conditions are associated with major fires in the forested areas of Southern Australia [29,30]. These ecosystems usually feature sufficient litter for propagation of fire at most times. ...
Article
Full-text available
Increased fire activity across the Amazon, Australia, and even the Arctic regions has received wide recognition in the global media in recent years. Large-scale, long-term analyses are required to postulate if these incidents are merely peaks within the natural oscillation, or rather the consequence of a linearly rising trend. While extensive datasets are available to facilitate the investigation of the extent and frequency of wildfires, no means has been available to also study the severity of the burnings on a comparable scale. This is now possible through a dataset recently published by the German Aerospace Center (DLR). This study exploits the possibilities of this new dataset by exemplarily analyzing fire severity trends on the Australian East coast for the past 20 years. The analyzed data is based on 3,503 tiles of the ESA Sentinel-3 OLCI instrument, extended by 9,612 granules of the NASA MODIS MOD09/MYD09 product. Rising trends in fire severity could be found for the states of New South Wales and Victoria, which could be attributed mainly to developments in the temperate climate zone featuring hot summers without a dry season (Cfa). Within this climate zone, the ecological units featuring needleleaf and evergreen forest are found to be mainly responsible for the increasing trend development. The results show a general, statistically significant shift of fire activity towards the affection of more woody, ecologically valuable vegetation.
... Similar results were found for FWI_mean (Fig. S18 online). This differs from previous studies reporting ENSO and IOD as the predominant drivers of interannual fire weather conditions in SE Australia, using correlation-based analysis [20,29,75,76] . The individual single-variable partial dependence plot shows that when the value of SAM is less than around −0.5σ, the FWI90 increases sharply (Fig. 5a). ...
Article
Full-text available
In Australia, the proportion of forest area that burns in a typical fire season is less than for other vegetation types. However, the 2019−2020 austral spring-summer was an exception, with over four times the previous maximum area burnt in southeast Australian temperate forests. Temperate forest fires have extensive socio-economic, human health, greenhouse gas emissions, and biodiversity impacts due to high fire intensities. A robust model that identifies driving factors of forest fires and relates impact thresholds to fire activity at regional scales would help land managers and fire-fighting agencies prepare for potentially hazardous fire in Australia. Here, we developed a machine-learning diagnostic model to quantify nonlinear relationships between monthly burnt area and biophysical factors in southeast Australian forests for 2001−2020 on a 0.25° grid based on several biophysical parameters, notably fire weather and vegetation productivity. Our model explained over 80% of the variation in the burnt area. We identified that burnt area dynamics in southeast Australian forest were primarily controlled by extreme fire weather, which linked to fluctuations in the Southern Annular Mode (SAM) and Indian Ocean Dipole (IOD), with a relatively smaller contribution from the Central Pacific El Niño Southern Oscillation (ENSO). Our fire diagnostic model and the non-linear relationships between burnt area and environmental covariates can provide useful guidance to decision-makers who manage preparations for an upcoming fire season, and to model developers working on improved early warning systems for forest fires.
... These results are complementary to previous studies that link El Niño to fire danger (e.g. Love and Downey 1986;Williams and Karoly 1999;Williams et al 2001;Verdon et al. 2004;Long 2006;Lucas et al. 2007) and fire activity (Felderhof and Gillieson 2006;Nicholls and Lucas 2007;Harris et al. 2008Harris et al. , 2014Mariani et al. 2016) in Australia. The signal is weaker in DJF than in SON over many areas, particularly in South Australia where a decrease in extreme fire weather frequency occurs during El Niño. ...
Article
Full-text available
We assess the ability of the Bureau of Meteorology’s ACCESS-S1 dynamical forecast system to simulate and predict extreme fire weather over Australia during austral spring (SON) and summer (DJF) on subseasonal timescales. Specifically, we focus on the roles of the El Niño-Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), Southern Annular Mode (SAM), Madden–Julian Oscillation (MJO), and two modes of persistent high-pressure in the Australian region characterised as (i) split-flow blocking highs and (ii) subtropical ridge Tasman highs (STRH). The observed likelihood of extreme fire weather increases over most of Australia in association with El Niño, the positive IOD, negative SAM and low split-flow blocking, in both seasons. These increases are generally largest in SON over the southeast. Notable increases in the likelihood of extreme fire weather also occur north of 30° S during low STRH activity, and over the southeast during MJO phase 3. Using retrospective forecasts at lead times of 2–3 weeks for the period 1990–2012, we show that ACCESS-S1 simulates reasonably well the observed modulation of extreme weekly-mean fire weather by each climate driver, however the simulated changes in probabilities are often weaker than those observed. Each climate driver plays an important role in providing predictive skill for regions where ACCESS-S1 captures a high likelihood of experiencing extreme fire weather conditions. The results of this study highlight windows of forecast opportunity during active climate driver phases that can be useful to regional users in fire management, emergency services, health, national park management, and the agriculture and energy sectors.
... El clima de Victoria tiene fuertes influencias remotas. El Niño Oscilación del Sur (ENOS) y el Dipolo del Océano Índico (IOD por sus siglas en Inglés) son importantes impulsores de la variabilidad climática interanual en Victoria (Verdon, Kiem and Franks, 2004) y sus referencias. Estos controladores pueden producir condiciones más húmedas o secas en esta región. ...
Article
Full-text available
La tendencia mundial es una correspondencia entre el aprovechamiento eficiente de recursos naturales y el desarrollo que permita enfrentar los desafíos de proveer alimento, agua y energía para permitirle a toda la población una vida digna. En los últimos años el uso de recursos naturales ha entrado en fuerte debate, en el cual no se ha entendido bien que la tecnología es mucho más que sus aplicaciones prácticas, es el resultado del edificio conceptual de la ciencia; es por lo tanto una obra colectiva cultural que, localmente, aprende del entorno y aprovecha los recursos sin rupturas del tejido social. Es nuestro sentir que las tecnologías aplicadas a los recursos naturales desarrollan un espacio centrado en el conocimiento de los recursos naturales, y el manejo de herramientas técnicas que permite aprovechar estos recursos, dentro de un margen de desarrollo sostenible. Los recursos naturales son los elementos de la naturaleza que el hombre puede utilizar y aprovechar. Estos recursos naturales representan, además, fuentes de riqueza para la explotación económica. Por ejemplo, los minerales, el suelo, los animales y las plantas constituyen recursos naturales que el hombre puede utilizar directamente como fuentes para esta explotación. De igual forma, los combustibles, el viento y el agua pueden ser utilizados como recursos naturales para la producción de energía. Pero la mejor utilización de un recurso natural depende del conocimiento que el hombre tenga al respecto, y de las leyes que rigen la conservación de la megadiversidad de nuestro país.
... Our model correctly predicts high burned area in these regions and years. Furthermore, the extreme fire events observed in Australia in 2011 and 2012 appear to be caused by negative values of the Interdecadal Pacific Oscillation (IPO) coupled with El Niño (negative values of the Southern Oscillation Index) 51,52 . Our model also predicted high burned areas in Australia during these years (Fig. 3J). ...
Article
Full-text available
Fires determine vegetation patterns, impact human societies, and are a part of complex feedbacks into the global climate system. Empirical and process-based models differ in their scale and mechanistic assumptions, giving divergent predictions of fire drivers and extent. Although humans have historically used and managed fires, the current role of anthropogenic drivers of fires remains less quantified. Whereas patterns in fire–climate interactions are consistent across the globe, fire–human–vegetation relationships vary strongly by region. Taking a data-driven approach, we use an artificial neural network to learn region-specific relationships between fire and its socio-environmental drivers across the globe. As a result, our models achieve higher predictability as compared to many state-of-the-art fire models, with global spatial correlation of 0.92, monthly temporal correlation of 0.76, interannual correlation of 0.69, and grid-cell level correlation of 0.60, between predicted and observed burned area. Given the current socio-anthropogenic conditions, Equatorial Asia, southern Africa, and Australia show a strong sensitivity of burned area to temperature whereas northern Africa shows a strong negative sensitivity. Overall, forests and shrublands show a stronger sensitivity of burned area to temperature compared to savannas, potentially weakening their status as carbon sinks under future climate-change scenarios.
... A handful of studies have focused on the effect of slope on fire propagation (Butler et al., 2007;Estes et al., 2017;Liu et al., 2014) and fuel moisture content (Burton et al., 2019). Only a few of these variables (particularly, temperature, relative humidity, wind speed, and meteorological drought) are accounted for in the McArthur Forest Fire Danger Index (FFDI) (Noble et al., 1980), which is widely used to assess the probability of bushfire occurrence in Australian landscapes (Verdon et al., 2004). Apart from these factors, some studies have also found antecedent soil moisture content to be a critical factor governing the bushfire propagation; however, these studies are only conducted in the United States of America (Krueger et al., 2015;Schulte et al., 2019). ...
Article
Full-text available
The recent bushfires (2019-2020) in New South Wales (NSW) Australia were catastrophic by claiming human and animal lives, affecting ecosystems, destroying infrastructure, and more. Recent studies have investigated relationships between hydroclimatic signals and past bushfires, and very recently, a few commentary papers claimed drought and fuel moisture content as the probable causes for the widespread 2019-2020 bushfires. Therefore, in this study, a novel work of encompassing a wide range of factors attributing to the recent bushfires is presented. Empirical evidence-based statistical methods are used to identify the hydroclimatic variables and geomorphic characteristics contributing to the 2019-2020 bushfires. The results highlight that ongoing drought, surface soil moisture (SSM), wind speed (WS10), relative humidity (RH), heat waves (HW), dead and live fuel moisture, and certain land cover types create favorable conditions for fire ignition and aid in fire propagation in different regions of the NSW state. The findings suggest that accounting for the above-identified variables in bushfire prediction and monitoring system are crucial in avoiding such catastrophes in the future. The overarching application of this study is developing robust and more versatile fire protection planning and management. Plain Language Summary Since the 2019-2020 Australian bushfires were catastrophic in terms of burnt area and severity, a detailed analysis of the primary causes is crucial. In this paper, several probable causes are tested statistically to establish their relationship with the burnt area. The results indicate that the ongoing drought, surface soil moisture, wind speed, relative humidity, heat waves, dead and live fuel moisture, and land cover with certain vegetation (particularly native eucalyptus and grazing land) are the primary causes of the widespread bushfire. These results are extremely critical in updating the current bushfire planning and management.
... Refer to Box 2.1 for further details. ENSO causes extreme weather (such as floods and droughts) in many regions of the world, particularly those bordering the Pacific Ocean (e.g., Gallant et al. 2012;Murphy and Timbal 2008), with the effects of ENSO including magnified fluctuations in streamflow volumes compared to rainfall (Chiew et al. 1998;Verdon et al. 2004b;Wooldridge et al. 2001), elevated flood risk during La Niña events (Kiem et al. 2003), and increased risk of drought (Kiem and Franks 2004) and bushfire (Verdon et al. 2004a) during El Niño events. Downloaded by [University of Newcastle] at 20:02 24 July 2014 ENSO Modoki: A pattern of central equatorial Pacific SST variations has also been identified by Ashok et al. (2007) and defined as an ENSO Modoki (Modoki is a Japanese word meaning a similar but different thing). ...
Book
Full-text available
There is increasing evidence of coherent patterns of variability on near quasi-bidecadal time scales in a range of climatic data from many parts of the world. Folland et al. (1984) found peaks at periods of 16 and 21 years respectively in spectra of globally- averaged sea surface temperature (SST) and night marine air temperature (NMAT) for 1856–1981. Newell et al. (1989) found variations near a period of 21 years in global and Southern Hemisphere NMAT for 1856–1986, and to a lesser extent in Northern Hemisphere NMAT and global and hemispheric SST. Ghil and Vautard (1991) drew attention to variations on approximately 20-year time-scales in globally-averaged anomalies of combined land surface air temperature and SST for 1854–1988, though Allen and Smith (1996) question the statistical significance of their results. Mann and Park (1994) found a 15–18 year mode in fields of mainly land surface air temperature anomalies for 1891–1990. They suggested that this mode, which had a pattern similar to that of the thermal signature of the interannual El Niño-Southern Oscillation (ENSO), may be a manifestation of long timescale modulation of ENSO as well as being the reason for Ghil and Vautard’s (1991) global-average result. Latif and Barnett (1996) discussed near bidecadal variations in SST and atmospheric circulation over the North Pacific in both observations and a coupled model, and the consequential variations of temperature and precipitation over North America. In the Southern Hemisphere, Venegas et al. (1996) found a coupled mode in South Atlantic SST and mean sea level pressure (MSLP) data for 1953–1992, with significant variations on near 15-year timescales and provided evidence that the atmosphere was forcing the ocean.
Article
Full-text available
The best available surface temperature (T) and precipitation (P) records for Australia dating back to 1910 have been examined to look for coherent interannual variability. P exhibits a tendency to be out of phase with daily maximum temperature, Tmax, and this results in P tending to be out of phase with both the daily average temperature, Tbar (estimated here as the average of Tmax and the daily minimum, Tmin), and the DTR (diurnal temperature range, Tmax-Tmin). The association between P and Tmin is generally weak. The (expected) increase in P associated with a positive Southern Oscillation Index is (generally) accompanied by reduced average temperatures (Tbar) and a reduced DTR, both of which primarily arise from a reduction in Tmax. When variability in both P and Tmin associated with Tmax is removed, the residual signals (P* and Tmin*) show widespread statistically significant positive correlations, consistent with the hypothesis that clouds help to reduce night-time cooling. These relationships are less clear at near-coastal sites, and absent at the island and exposed coastal sites considered. Results from three separate ten-year integrations of the Bureau of Meteorology Research Centre's atmospheric general circulation model were then examined. The tendency for (a) P to be out of phase with Tmax, Tbar and the DTR and (b) P* to be in phase with Tmin* over Australia on interannual time-scales was also generally evident over land elsewhere, except at high latitudes and over North Africa. An analysis of the model's surface heat budget over land showed that this arises from associated surface short wave radiation and latent heating anomalies. The latter is generally more important over low-latitude regions where deep convection occurs, with the hierarchy reversed elsewhere. Evaporative cooling anomalies appear to be dominated by soil moisture changes. Surface long wave radiation, sensible heating and subterranean heat exchange tend to reduce the temperature change which would otherwise occur. Recent changes in some of the relationships exhibited between observed P, T and the Southern Oscillation Index appear unusual in terms of the interdecadal variability evident in the records prior to 1972, and previous conclusions drawn on the basis of 'all-Australia' P and T indices were found to have broad applicability. Interrelationships between recent changes in the 20-year means of P, T and the SOI do not match the changes that might be expected on the basis of their interrelationship on interannual time-scales. Possible reasons for the changes suggested by the analysis (e.g., global warming and naturally occurring interdecadal climate variability) are discussed.
Article
Full-text available
This study investigates the influence of the El Nino-Southern Oscillation on fire weather in Australia, including its impact on the severity of seasonal fire danger and on extreme daily fire danger, The meteorological parameters of McArthur's forest Fire Danger Index (FDI) are analysed for composites of years of extreme high and low Southern Oscillation Index (SOT) during the period 1960 to 1992 using eight stations in different climatic zones. Results show that in southeast Australia and in central Australia, seasonal fire danger is higher in years of strong negative SOI and that the daily FDI has a significantly different distribution (with many more days with extreme fire danger). The Southern Oscillation has an opposite but small impact on the daily FDI distribution in the southwest of Australia. Daily minimum relative humidity (RH) is the fire weather parameter that is most strongly influenced by the Southern Oscillation. In southeastern Australia RH is significantly lower in years of negative SOI.