Download Climatological evaluation of Haines forest fire weather index over

METEOROLOGICAL APPLICATIONS
Meteorol. Appl. (2013)
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/met.1367
Climatological evaluation of Haines forest fire weather index
over the Mediterranean Basin
Hasan Tatli* and Murat Türkeş
Department of Geography, Physical Geography Division, Climatology/Meteorology Group, Çanakkale Onsekiz Mart University, Turkey
ABSTRACT: Forest fires are the most crucial natural threat to forests and wooded areas. They destroy many more trees
than all other natural catastrophes such as parasite attacks, insects, extreme weather events and others. Forest fires, especially
in summer and dry autumn/spring periods, are frequent in the Mediterranean basin and represent growing environmental
and ecological problems. The aim of this investigation is to determine a climatic pattern of fire-meteorology over the
Mediterranean basin via the frequency analysis of the forest fires weather index (FFWI) of Haines. The FFWI values were
obtained by using the hourly data derived from reanalysis fields available from the National Centers for Environmental
Prediction/National Center for Atmospheric Research (NCEP/NCAR) for the period 1980–2010. High frequency values of
FFWI, taken to be a sign of moderate-level risk of forest fires, were obtained on the forests, scrubs, succulents and wooded
areas in several countries facing the sea including Greece, Italy, Turkey, Syria, Lebanon, Cyprus, Macedonia, Albania,
Serbia, Slovenia, France, Portugal, Spain, Morocco and Tunisia, almost all of which are characterized by dry summer,
subtropical Mediterranean climates. As expected, the highest-level risk values are found in the arid desert climate regions:
the desert areas of the Sahara and Libya in North Africa and the semi-arid steppe climate regions of the Middle East, as
well as the semi-arid environments near the Caspian Sea basin. Copyright  2013 Royal Meteorological Society
KEY WORDS
fire meteorology; frequency; Haines Index; Mediterranean; dry summer; NCEP/NCAR-reanalysis
Received 10 April 2012; Revised 3 September 2012; Accepted 10 October 2012
1.
Introduction
The Mediterranean basin is heavily affected by forest fires;
every year more than 50 000 fires burn an estimated average
of 600 000–900 000 ha, an area roughly equal to 1.3–1.7%
of total Mediterranean forests (EEA, 2008). Forest fires have
decreased in some regions of the world, for example in Canada
(Gillett et al ., 2004); however in regions of the Mediterranean
basin where data have been available since the 1950s, a large
increase in the number of forest fires has been observed from
the beginning of the 1970s (European Union, 1993, 1996;
EEA, 2008). In a recently published study of Juli and Santiago
(2012), the forest fires in the Western Mediterranean basin
have been becoming larger and more frequent, but according
to the European Forest Fire Information Systems (European
Commission, 2010), the number of fires in the southern member
states of European Union increased from 1980 to 2000, and has
been slightly decreasing since that time. Current climate change
trends are towards longer summer droughts and intensification
of droughts in other seasons (Tatli et al ., 2004, 2005; Flannigan
et al ., 2006; Moriondo et al ., 2006; IPCC, 2007; Tatli, 2007;
Türkeş et al ., 2009; Türkeş and Tatli, 2009, 2011; Tatli and
Türkeş, 2011).
Furthermore, the occurrence of fires could be much higher
recently due to economical, political and social activities of
humans. For instance, the vast majority of fires in the tropics
* Correspondence: H. Tatli, Faculty of Sciences and Arts, Department
of Geography, Physical Geography Division, Çanakkale Onsekiz Mart
University, Terzioglu Campus, 17020 Çanakkale, Turkey. E-mail:
[email protected]
Copyright  2013 Royal Meteorological Society
are set deliberately for land clearing and conversion, and as
such are considered land use fires, not forest fires. In the past
two decades, wide-ranging and frequent droughts, coupled with
increased pressures on land and unsustainable forest use, have
led to an increase in catastrophic fire events (SCBD, 2001).
However, the appropriate meteorological conditions alone
are not enough to start a forest fire. Fire establishes when
oxygen, heat and fuel (a form of glucose (C6 H12 O6 )n ) combined
together in the proper ratio lead to fuel ignition and combustion,
frequently referred to as the fire triangle. Additionally, extreme
weather events, such as periods of high temperatures, strong
air dryness and very strong winds, as well as sudden storms
with heavy rainfall in only few hours, are becoming frequent
(Johnson and Miyanishi, 2001). One day of high temperatures
combined with very low humidity and strong winds could be
enough to activate an abandoned forest fire’s processes, which
can lead to damage of thousands hectares of forest. Wildfires
typically occur during periods of increased temperature and
drought. Therefore, scientists and other users make use of
similar indices for monitoring the forest fires and drought
events. There are many tools for monitoring forest fires. Besides
the remote sensing techniques which are widely used (e.g.,
Paltridge and Barber, 1988; Leblon, 2005), in some cases the
forest fire weather indices are also affectively used. These
indices are used in many parts of the world to integrate
meteorological and fuel information into a single value. This
value can then be applied to regions for the issuing of warnings.
The most widely used indices are the Canadian Forest Fire
Weather Index (CFWI) System (Van Wagner and Pickett,
1985; Van Wagner, 1987; Dimitrakopoulos et al ., 2011), the
US National Fire Danger Ratings System (NFDRS) (Deeming
et al ., 1977), Keetch-Byram Drought Index (KBDI) (Keetch and
H. Tatli and M. Türkeş
Figure 1. The study area, and geographical distributions of climate types over the Mediterranean macroclimate region based on the first, second
and third hand letters classification of the Köppen–Geiger climate system. It was re-drawn and arranged from the data by Peel et al . (2007).
Byram, 1968) and the Haines Index (Haines, 1988). In fact,
some of these techniques, such as KBDI, combine measures
of antecedent precipitation and observations or forecasts of
near-surface wind speed, air temperature and humidity data to
estimate fire danger for fires that are primarily wind-driven, but
the Haines Index does not use wind quantity, which indicates
its disadvantage.
Moreover, the formation of a fire, fire growth and the
progression of fire usually depend on vertical humidity and
stability of the atmosphere. Accordingly, the Haines Index
(Haines, 1988) responds to these situations that measure the
potential for rapid wild forest fire growth. Haines does not
explicitly define the criteria for major fire, but one can find
some suggestion of the size of major fire in the study of
Potter (1996), where the criterion is 400 ha. The Haines Index
is widely applied, and a number of studies for the wild
forest fire-prone countries such as in the United States (e.g.,
Saltenberger and Barker, 1993; Werth and Werth, 1998; Potter
and Goodrick, 2003) have shown significant positive correlation
with the Haines Index. There are also several similar key studies
performed in Australia. For example, the study of Bally (1995)
shows that the Haines Index is influentially efficient to predict
fire activity. Additionally, the studies of Long (2006), McCaw
et al . (2007) and Mills and McCaw (2010) are relevant. The
study by Mills and McCaw (2010) developed a climatology of
the Haines Index for southern Australia, proposed an extension
of the index to give it greater resolution in dry continental
environments, and examined a number of case studies where
the index appeared to provide useful information in identifying
factors contributing to unexpected fire behaviour.
2.
A brief outlook of the Mediterranean Basin Climate
In order to evaluate the results of the present study, a brief
outlook of the general features of the climate of the Mediterranean will be helpful. According to the Köppen–Geiger climate classification system, the mid-latitude warm temperate
(mild) climates (C climates) are mostly located in parts of the
latitude range between approximately 25 and 60◦ in either hemisphere, including the Mediterranean Basin (Peel et al ., 2007).
Furthermore, the Cs climates of the Köppen classification are
principally referred to as the temperate rainy climate with dry
summer (humid meso-thermal), but the more widely used term
Copyright  2013 Royal Meteorological Society
is the dry summer subtropical Mediterranean, which can be
shortened to the Mediterranean climate. In C climates the coldest month has an average temperature below 18 ◦ C, but above
0 ◦ C (0 ◦ C < T cold < 18 ◦ C); at least 1 month has an average
temperature above 10 ◦ C (T hot > 10 ◦ C). Temperate C climates
hence have both an evident summer and winter.
The Mediterranean Cs climates are found on the western
side of continents centred at about latitudes 40 ◦ N and 40 ◦ S.
With one exception, which can be called as the ‘true’ or
‘macro’ Mediterranean climate region, all the Mediterranean
climate regions are small in terms of the dominated area.
The Mediterranean Csa and Csb climates are also found,
for example, in northern Iran, California, Chile, Australia,
New Zealand and South Africa, in addition to the ‘true’
Mediterranean regions of Portugal, Spain, Coastal Northwest
Africa (Morocco, Tunisia, and Algeria), France, Italy, Greece,
Turkey, Lebanon and Israel (Figure 1).
The Mediterranean macroclimate mainly results from the
seasonal alternation between mid-latitude (frontal) cyclones,
associated with polar air masses, during the winter, and subtropical high-pressure systems, from subsiding maritime and
continental tropical air masses, during the summer. Continental tropical air-streams from the northern African and Middle
East/Arabian regions generally dominate, particularly throughout summer by causing long-lasting warm and dry conditions
over Turkey, except in the Black Sea region and the continental northeastern part of the Anatolian Peninsula (Türkeş, 1998,
2003; Türkeş et al ., 2009, 2011). In winter, a well-known combination of the northeastern Atlantic originated mid-latitude and
the Mediterranean cyclones and subtropical anticyclones from
the Azores control the weather and climate in the Mediterranean
Basin.
The Mediterranean Cs climates have the following distinctive
characteristics.
1. The major characteristic of the Mediterranean climate is of
high temporal variability varying from seasonal and interannual to centennial scales due to the following factors:
a. it extends in a transition region between temperate and
cold mid-latitudes and tropics (i.e. subtropical zone);
b. it has been facing significant circulation (associated
pressure and wind systems characterizing mid-latitude
Meteorol. Appl. (2013)
Evaluation of Haines forest fire weather index
and tropical/monsoonal weather and climate, respectively) changes between winter and summer, and,
c. it is closely associated with several atmospheric oscillation and/or teleconnection patterns, such as North
Atlantic Oscillation (NAO), Arctic Oscillation (AO),
Mediterranean Oscillation (MO), El Niño-Southern
Oscillation (ENSO), and North Sea – Caspian Pattern
(NCP) during the course of year.
2. About half of the modest annual precipitation amount falls
in winter, whereas summers are mostly rainless.
3. Winter temperatures are unusually mild for the middlelatitudes except in some eastern and inland regions: summer
temperatures vary from hot to warm.
4. Cloudless skies and intensive sunshine (incoming short
wave solar radiation) are typical, particularly in summer
months.
5. Major influences arise from sea and land distribution and
their interactions, in addition to the ocean–atmosphere
interaction, during the year, particularly in the ‘true’ or
‘actual’ Mediterranean macro-climate region.
The study area also includes the B arid and D cold climates
(Figure 1). In the arid climates, no water supply occurs in the
annual water balance. Consequently, generally no permanent
streams develop in these climate regions. The B arid climate
is divided into two types: S steppe and W desert climates.
The letters S and W are applied only to the arid (dry) B
climates, yielding two important combinations as BS and BW.
BS steppe climate is semi-arid with about average annual
precipitation varying from 350 to 700 mm, and the exact
precipitation boundary is determined by a formula considering
air temperature. The BS steppe climates are found mainly over
the Iberian and Anatolian peninsulas, the large area surrounding
the Caspian Sea, and the areas lying nearby the Csa climate
zones in the north-western Africa and the western part of
the Middle East region (Figure 1). BW desert climate is an
arid climate mostly having annual rainfall amount less than
250 mm, exact boundary of desert climate is also determined by
a formula. The BW desert climate regions dominate evidently
over the northern Africa, the Middle East and the Arabian
Peninsula regions.
The D cold (micro-thermal) climates are found in the
northern regions of the Mediterranean Basin including southern
Europe (excluding France mainly with Cf climates), Balkans,
central and eastern mountainously regions of the Anatolian
Peninsula and northern regions of the Caucasia (Figure 1).
In this climate, the coldest month average temperature is
under 0 ◦ C (T cold ≤ 0 ◦ C), and average temperature of the
warmest month is above 10 ◦ C (T hot > 10 ◦ C); that isotherm
also coincides approximately with pole-ward limit of forest
growth (Türkeş, 2010).
3.
Data and methodology
The Haines index is derived from the stability (temperature
difference between different levels of the atmosphere) and
moisture content (dew point depression) of the lower atmosphere. These data may be obtained from radiosonde observations, but this is problematic because there is an insufficient
amount of radiosonde data. In this study, the Haines Index values were obtained by using hourly temperature data measured
at 1200 GMT derived from reanalysis fields available from the
National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) (Kalnay et al .,
1996; Kistler et al ., 2001) for the period 1980–2010. The data
sets have a resolution of 2.5◦ by 2.5◦ in a pane of 12.5◦ W to
55◦ E by 30–50◦ N. The study area in this window completely
envelops the geography of the Mediterranean basin (Figure 1).
The calculation steps of the Haines Index require temperature
values at the 950, 850, 700 and 500 hPa pressure levels, and
dew point temperatures at the 850 and 700 hPa levels (Table 1).
Indeed, 950 hPa is not a standard pressure level, so the
interpolation method suggested by Potter et al . (2008) has been
used to obtain the required temperature data. Additionally, there
is no derived data set for dew point temperatures in the public
data sets of NCEP/NCAR reanalysis over the Mediterranean
basin. For that reason, the following algorithm is preferred with
other available data values of NCEP/NCAR reanalysis in order
to derive a data set of gridded dew point temperatures (T d ).
The dew point is the air temperature to which air must be
cooled to produce a saturated condition without changes in air
pressure and moisture content of the air, that is, the temperature
for which the actual vapour pressure is equal to the saturated
vapour pressure as defined in the following:
Td = 237.3 × ln
e e / 17.27 − ln
6.1078
6.1078
(1)
where e is the actual vapour pressure at the grid point in
question. However, there are also no vapour pressure values in
the public data sets of NCEP/NCAR. Thus, the required vapour
pressure values were obtained by using the public data values
of relative humidity (RH) and temperature (T ) available in the
data sets. The first step requires estimating saturated vapour
Table 1. Calculating steps of the Haines Index (HI).
Elevation
Low
T A = T 950 − T 850
Mid
T A = T 850 − T 700
High
T A = T 700 − T 500
Stability (A) component
category
Humidity (B) component
category
A = 1 if T A < 4 ◦ C
A = 2 if 4 ◦ C ≤ T A ≤ 7 ◦ C
A = 3 if T A < 7 ◦ C
A = 1 if T A < 6 ◦ C
A = 2 if 6 ◦ C ≤ T A ≤ 10 ◦ C
A = 3 if T A > 10 ◦ C
A = 1 if T A < 18 ◦ C
A = 2 if 18◦ C ≤ T A ≤ 21 ◦ C
A = 3 if T A > 21 ◦ C
B = 1 if T B < 6 ◦ C
B = 2 if 6 ◦ C ≤ T B ≤ 9 ◦ C
B = 3 if T B > 9 ◦ C
B = 1 if T B < 6 ◦ C
B = 2 if 6 ◦ C ≤ T B ≤ 12 ◦ C
B = 3 if T B > 12 ◦ C
B = 1 if T B < 15 ◦ C
B = 2 if 15 ◦ C ≤ T B ≤ 20 ◦ C
B = 3 if T B > 20 ◦ C
T B = T 850 − T d850
T B = T 850 − T d850
T B = T 700 − T d700
T and T d indicate the temperature and dew point temperature values at the pressure level in question respectively. HI = A + B : this is ranging from 2 to 6.
Copyright  2013 Royal Meteorological Society
Meteorol. Appl. (2013)
H. Tatli and M. Türkeş
pressure (e s ) of the air as:
273.15
T
273.15
exp 24.846 1 −
T
es = 6.1078 exp 5.0065 ln
(2)
The actual vapour pressure (e) values are then calculated
using the saturated vapour pressure (e s ) and relative humidity
(RH) values (in percent) as in the following:
e=
RH
es
100
Information Systems (GIS) was used to obtain elevation values
of the grids. The calculation steps of the Haines Index are given
in Table 1 according to Haines (1988) and Potter et al . (2008).
Haines (1988) originally developed the Index for the 0000 GMT
weather conditions in the northwest of the United States, which
is not constructed to identify the extreme conditions in the
Mediterranean basin due to the different temperature lapse-rate
and humidity climatology of the two continents. The FFWI of
Haines at 1200 GMT has been obtained, since the heating and
instability for the Mediterranean basin reaches its maximum
level at about 1200 GMT.
(3)
The dew point temperatures are then obtained by inserting e
into Equation (1).
The Haines index is calculated over three geopotential height
ranges: low elevation (950 to 850 hPa), mid elevation (850
to 700 hPa) and high elevation (700 to 500 hPa). A Haines
index equal to 6 means a high potential for large fire growth,
5 means medium potential, 4 low potential, and anything
less than 4 means very low potential. The thresholds used to
transform the differences to values depend on the elevation
(represented by z ) of the areas of interest. In the study, values of
the low (z ≤ 304.8 m (1000 feet)), mid (304.8 m < z ≤ 914.4 m
(3000 feet)) and high (z > 914.4 m) elevations as suggested
by Haines (1988) were used. Indeed, there is no elevation
information for the grids in the NCEP/NCAR data sets.
Hence, a digital elevation model (DEM) used in Geographic
4.
Results
In order to obtain particular climatic information, the relative frequencies of FFWI of Haines represented by categorical
values of ‘2, 3, 4, 5, 6’, and as well as the merged values of
‘2 + 3’, ‘4 + 5 + 6’ and ‘5 + 6’ were calculated. The geographical distribution map of relative frequencies of Haines Index
for the value equal to 6 is given in Figure 2. This figure shows
the distribution map of relative frequency of where forest fires
have high-level risk for large fire growth. It can also be phrased
as the map of long-term (climatological) probabilities. As seen
in this figure, the probability values greater than 0.2 (or 20%)
are only seen over the Syrian and An Nafud deserts of the
Middle East, and the desert areas of the Sahara and Libya
in north Africa. Nevertheless, these areas depict the poorest
Figure 2. The geographical distribution of the probabilities of Haines Index value equal to 6. These are areas with high potential for large fire
growth. This figure is available in colour online at wileyonlinelibrary.com/journal/met
Figure 3. The geographical distribution of the probabilities of Haines Index value of 5. These are areas with medium-risk potential for large fire
growth. This figure is available in colour online at wileyonlinelibrary.com/journal/met
Copyright  2013 Royal Meteorological Society
Meteorol. Appl. (2013)
Evaluation of Haines forest fire weather index
Figure 4. The geographical distribution of the probabilities of the merged index category obtained by summation of the Haines Index values
equal to 4, 5 and 6. This figure is available in colour online at wileyonlinelibrary.com/journal/met
(a)
(b)
(c)
Figure 5. Probability distribution of the humidity (B) component of the Haines Index: (a) B = 1; (b) B = 2; and (c) B = 3. This figure is available
in colour online at wileyonlinelibrary.com/journal/met
Copyright  2013 Royal Meteorological Society
Meteorol. Appl. (2013)
H. Tatli and M. Türkeş
(a)
(b)
(c)
Figure 6. Probability distribution of the stability (A) component of the Haines Index: (a) A = 1; (b) A = 2; and (c) A = 3. This figure is available
in colour online at wileyonlinelibrary.com/journal/met
regions in terms of forest cover as well. Likewise, remarkable
areas are also observed near eastern and northeastern regions
around the Caspian Sea, where the risk of forest fires is greater
than 0.2. From the point of view of climatology, one reason for
high-level risk of forest fire seen over these regions could be
originated from the effects of the cold winter deserts including
the Garagum (or Karakum), Qyzylqum and Gobi to the eastern
and northern regions near the Caspian Sea basin.
Figure 3 shows the geographical distribution of mid-level
risk of Haines Index equal to 5. According to this figure, the risk
values of greater than 0.2 are seen in the south of Europe and
northwestern Africa, where summer forest fires often occur. In
addition, regions of the Aegean Sea (i.e., Greece and the southern and western parts of Turkey), Albania, Macedonia, Southern
France, Italy, Slovenia, Morocco and Tunisia could be identified
Copyright  2013 Royal Meteorological Society
as the areas having moderate-level risk. These regions are also
relatively rich in forest cover with the Mediterranean climate.
Figure 4 depicts the distribution of relative-frequencies
obtained by summation of the Haines Index values equal to 4,
5 and 6. As shown in this figure, high-level risks greater than
0.3 are seen in almost all areas except the high mountains and
plateaus of Turkey’s interior, northern parts of Europe, and the
Middle East region with high mountainous sections. In addition,
high-level risks greater than 0.5 are concentrated in the western
part of Turkey, Greece, the southern parts of the Balkans, Italy,
Slovenia, Spain, Morocco and Tunisia, all of which are mostly
characterized with the dry summer subtropical Mediterranean
climate (Figure 1).
The probability distributions of humidity and the stability
components of the Haines index are given in Figures 5(a)
Meteorol. Appl. (2013)
Evaluation of Haines forest fire weather index
and 6(a). According to these figures, the stable atmospheric
conditions with low humidity create similar patterns as obtained
for the Haines index value equal to 1. This result reveals a
parallel with the results in the recent study carried out for
Turkey by Tatli and Türkeş (2011): the principal components of
Palmer Drought Severity Index show similar patterns with the
patterns extracted for soil moisture. The physical geographical
controllers (climate, soil, geomorphology) and atmospheric
features have strong inter-relationships.
As a result, the spatial patterns for the Haines Index values
have a clear similarity between the areas covered with forest
and wooded areas (Figures 5 and 6). However, high-level risk
values seen near the Caspian Sea basin could be evaluated, as
these areas are semi-arid steppe climate zones (Figures 3–6).
The probability patterns seen over middle and eastern parts of
north Africa (the desert areas of the Sahara and Libya) can
be assessed by a similar point of view. That is, the reasons
for high-level probabilities seen over these regions could be
due to the fact of these areas being desert, one of the reasons
being that the forest fire weather and drought indices are already
obtained by using the same physical background, since both
indices are the highest when the air reaches high temperature
and low humidity.
5.
Conclusions
In this study, the forest fire weather index developed by Haines
(1988) was applied over the Mediterranean basin and Turkey.
It is an important weather index used for monitoring largescale wild forest fires. The Haines index values were obtained
at 1200 GMT, because the temperature and instability reach
their maximum values at this time over the Mediterranean
basin.
To calculate index values, temperature and dew point temperature values are required at various standard pressure levels.
However, there is no public gridded data set of dew point temperatures in the data sets of NCEP/NCAR reanalysis over the
Mediterranean basin. For that reason, the dew point temperature values were obtained using the relative humidity and air
temperature values available in the public data sets.
After obtaining index values, the maps of relative-frequencies
were analysed by viewing the large-scale climatic aspects.
The results show that the highest-level risks of Haines Index
values were found in semi-arid steppe (BS) and arid desert
(BW) climate regions according to the Köppen–Geiger climate
classification system. On the other hand, considerable risk
values of medium-level for forest fires were found in the
forests and wooded areas of the Mediterranean basin, generally
with the mid-latitude warm temperate (C) climates but mainly
pronounced by the dry summer subtropical Mediterranean
climates (Csa and Csb).
Moreover, the IPCC (2007) suggests that with global warming, forest fire frequency will increase all over the world.
Accordingly, several studies point to the fact that global
warming is likely to increase fire frequency and severity (Flannigan et al ., 2006; Moriondo et al ., 2006). Thus, due to high
temperatures and droughts, an increase in the number of forest fires in future will also be evident. Therefore, the climatic
features of forest fires are evident in the large-scale patterns of
fire meteorology.
The inability to obtain the records of real forest fires in some
areas has a negative impact on scientific studies of this issue,
because the results could not be compared with the statistics
Copyright  2013 Royal Meteorological Society
of real forest fires. This should be viewed as an important
open point of this study. Although the spatial properties of
the Haines index were examined in this study, it is suggested
that researchers examine its dynamics as well. For example, the
persistence and trend in the index values need to be studied by
means of its autoregressive features.
On the other hand, the assessments of growing environmental
problems from forest fires, such as emissions from biomass
burning, are very important issues that should be revisited.
Finally, it is hoped that the large-scale spatial distribution
patterns for the risk of forest fires obtained in this study
will provide significant climatic information to end users,
such as for forest fire managers and officers of the regional
and national forestry services and ministries, and researchers
and scientists dealing with the climate change adaptation and
modelling.
Acknowledgements
We would like to thank Dr Matthew Jones (School of Geography, University of Nottingham, UK) for improving the English
quality of our manuscript.
References
Bally J. 1995. The Haines index as a predictor of fire activity in
Tasmania. Proceedings of Australian Bushfire Conference, Hobart.
Deeming JE, Burgan RE, Cohen JD. 1977. The National fire-danger
rating system, 1978. General Technical Report INT- 39, United
States Department of Agriculture, Forest Service, Intermountain
Forest and Range Experiment Station: Ogden, UT.
Dimitrakopoulos AP, Bemmerzouk AM, Mitsopoulos ID. 2011. Evaluation of the Canadian fire weather index system in an eastern Mediterranean environment. Meteorol. Appl. 18: 83–93, DOI:
10.1002/met.214.
EEA. 2008. European forests — ecosystem conditions and sustainable
use. European Environment Agency (EEA) Report No 3/2008,
European Environment Agency (EEA): Copenhagen.
European Commission. 2010. Forest Fires in Europe 2009 . EUR 24502
EN. Publication Office of the European Union: Luxembourg.
European Union. 1993. Pan-European Conference on Forest Protection
in Europe. Strasbourg 1990 (Resolution No. 3). Report of the twelve
member states (DG-VI FII 2 Ed.). European Union: Brussels.
European Union. 1996. Forest Fires in the South of the European
Union-Pilot Project in Preparation for Setting up the Community
Forest-Fire Information System(DG-VI FII 2 Ed.). European Union:
Brussels.
Flannigan MD, Amiro BD, Logan KA, Stocks BJ, Wotton BM. 2006.
Forest fires and climate change in the 21st century. Mitigat. Adapt.
Strategy Glob. Change 11: 847–859.
Gillett NP, Weaver AJ, Zwiers FW, Flannigan MD. 2004. Detecting
the effect of climate change on Canadian forest fires. Geophys. Res.
Lett. 31: L18211, DOI: 10.1029/2004GL020876.
Haines DA. 1988. A lower atmosphere severity index for wildland
fires. Natl. Weather Dig. 13: 23–27.
IPCC. 2007. IPCC Intergovernmental Panel on Climate Change,
Climate change 2007: the physical science basis. In Contribution of Working Group I to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change, Solomon S, Qin D,
Manning M, Chen Z, Marquis M, Averyt KB, Tignor M, Miller HL
(eds). 996 pp. Cambridge University Press: Cambridge, UK, and
New York, NY.
Johnson EA, Miyanishi K (eds). 2001. Forest Fires: Behavior and
Ecological Effects. Academic Press: San Diego, CA.
Juli GP, Santiago FM. 2012. Fire regime changes in the Western Mediterranean Basin: from fuel-limited to drought-driven fire
regime. Clim. Chang. 110: 215–226, DOI: 10.1007 /s10584-0110060-6.
Kalnay E, Kanamitsu M, Kistler R, Collins W, Deaven D, Gandin L,
Iredell M, Saha S, White G, Woollen J, Zhu Y, Chelliah M, Ebisuzaki
W, Higgins W, Janowiak J, Mo KC, Ropelewski C, Wang J, Leetmaa
Meteorol. Appl. (2013)
H. Tatli and M. Türkeş
A, Reynolds R, Jenne R, Joseph D. 1996. The NCEP/NCAR 40-year
reanalysis project. Bull. Am. Meteorol. Soc. 77: 437–471.
Keetch JJ, Byram GM. 1968. A Drought Index for Forest Fire Control.
Research Paper SE-38, Asheville, NC. USDA Forest Service.
Kistler R, Kalnay E, Collins W, Saha S, White G, Woollen J, Chelliah
M, Ebisuzaki W, Kanamitsu M, Kousky V, Van Den Dool H, Jenne
R, Fiorino M. 2001. The NCEP/NCAR 50-year reanalysis: monthly
means CD-ROM and documentation. Bull. Am. Meteorol. Soc. 82:
247–268.
Leblon B. 2005. Monitoring forest fire danger with remote sensing.
Nat. Hazards 35: 343–359.
Long M. 2006. A climatology of extreme fire weather days in Victoria.
Aust. Meteor. Mag. 55: 3–18.
McCaw L, Marchetti P, Elliott G, Reader G. 2007. Bushfire weather
climatology of the Haines Index in south-west WA. Aust. Meteor.
Mag. 56: 75–80.
Mills GA, McCaw L. 2010. Atmospheric stability environments and
fire weather in Australia – extending the Haines Index. CAWCR
Technical Report No 20, CAWCR. www.cawcr.gov.au (accessed 9
April 2012).
Moriondo M, Good P, Durao R, Bindi M, Giannakopoulos C, CorteReal J. 2006. Potential impact of climate change on fire risk in the
Mediterranean area. Clim. Res. 31: 85–95.
Paltridge GW, Barber J. 1988. Monitoring grassland dryness and
fire potential in Australia with NOAA-AVHRR data. Remote Sens.
Environ. 25: 381–394.
Peel MC, Finlayson BL, McMahon TA. 2007. Updated world map of
the Köppen-Geiger climate classification. Hydrol. Earth Syst. Sci. 11:
1633–1644.
Potter BE. 1996. Atmospheric properties associated with large wildfires. Int. J. Wildland Fires 6: 71–76.
Potter BE, Goodrick S. 2003. Performance of the Haines Index
during August 2000 for Montana. In Proceedings of the 4th
Symposium on Fire and Forest Meteorology, 13–15 November
2001, Reno, NV . American Meteorological Society: Boston, MA;
233–236.
Potter BE, Winkler JA, Dwight FW, Ryan PS, Xindi B. 2008.
Computing the low-elevation variant of the Haines Index for fire
weather forecasts. Weather Forecast. 23: 159–167.
Saltenberger J, Barker T. 1993. Weather related unusual fire behavior
in the Awbrey Hall. Fire. Natl. Wea. Dig. 19: 20–28.
SCBD. 2001. Impacts of Human-Caused Fires on Biodiversity and
Ecosystem Functioning, and Their Causes in Tropical, Temperate and
Boreal Forest Biomes, CBD Technical Series No. 5. Secretariat of
Copyright  2013 Royal Meteorological Society
the Convention on Biological Diversity (SCBD): Montreal.
Tatli H. 2007. Synchronization between the North Sea-Caspian pattern
(NCP) and surface air temperatures in NCEP. Int. J. Climatol. 27:
1171–1187.
Tatli H, Dalfes HN, Menteş ŞS. 2004. A statistical downscaling method
for monthly total precipitation over Turkey. Int. J. Climatol. 24:
161–180.
Tatli H, Dalfes HN, Menteş ŞS. 2005. Surface air temperature
variability over Turkey and its connection to large-scale upper
air circulation via multivariate techniques. Int. J. Climatol. 25:
331–350.
Tatli H, Türkeş M. 2011. Empirical orthogonal function analysis of the
Palmer Drought Indices. Agric. For. Meteorol. 151: 981–991.
Türkeş M. 1998. Influence of geopotential heights, cyclone frequency
and Southern Oscillation on rainfall variations in Turkey. Int. J.
Climatol. 18: 649–680.
Türkeş M. 2003. Spatial and temporal variations in precipitation
and aridity index series of Turkey. In Mediterranean Climate –
Variability and Trends, Regional Climate Studies, Bolle HJ (ed)(ed.).
Springer-Verlag: Heidelberg; 181–213.
Türkeş M. 2010. Climatology and Meteorology, Physical Geography
Series No. 1, 1stPublication No. 63 edn. Kriter Publisher: Istanbul
(in Turkish) ISBN: 978-605-5863-39-6.
Türkeş M, Koç T, Sarış F. 2009. Spatiotemporal variability of precipitation total series over Turkey. Int. J. Climatol. 29: 1056–1074.
Türkeş M, Kurnaz ML, Öztürk T, Altınsoy H. 2011. Climate changes
versus ‘security and peace’ in the Mediterranean macroclimate
region: are they correlated? Proceedings of International Human
Security Conference on Human Security: New Challenges, New
Perspectives, 27–28 October 2011, CPRS Istanbul, Turkey.
Türkeş M, Tatli H. 2009. Use of the standardized precipitation index
(SPI) and a modified SPI for shaping the drought probabilities over
Turkey. Int. J. Climatol. 29: 2270–2282.
Türkeş M, Tatli H. 2011. Use of the spectral clustering to determine
coherent precipitation regions in Turkey for the period 1929-2007.
Int. J. Climatol. 31: 2055–2067.
Van Wagner CE. 1987. Development and structure of the Canadian
forest fire weather index system. Technical Report 35, Canadian
Forestry Service, Ottawa.
Van Wagner CE, Pickett TL. 1985. Equations and FORTRAN Program
for the Canadian Forest Fire Weather Index System. Canadian
Forestry Service: Ottawa.
Werth J, Werth P. 1998. Haines Index climatology for the western
United States. Fire Manage. Notes 58: 8–17.
Meteorol. Appl. (2013)