1. No category

Science of the Total Environment 524–525 (2015) 32–39
Contents lists available at ScienceDirect
Science of the Total Environment
journal homepage: www.elsevier.com/locate/scitotenv
Atmospheric conditions associated with extreme fire activity in the
Western Mediterranean region
Malik Amraoui a,b,⁎, Mário G. Pereira c,b, Carlos C. DaCamara b, Teresa J. Calado b
a
b
c
Universidade de Trás-os-Montes Alto Douro, UTAD, Escola de Ciências e Tecnologia, Quinta de Prados, 5000-801 Vila Real, Portugal
University of Lisbon, Instituto Dom Luiz (IDL), Lisbon, Portugal
Centro de Investigação e de Tecnologias Agro-Ambientais e Biológicas, CITAB, Universidade de Trás-os-Montes e Alto Douro, UTAD, Quinta de Prados, 5000-801 Vila Real, Portugal
H I G H L I G H T S
•
•
•
•
•
West Iberia (WI) and North Africa are the sub-regions most affected by fires.
Fire annual cycle has a main peak in August and secondary peak in March over WI.
Extreme fire activity episodes are identified in both summer and winter seasons.
Extreme fire episodes occur in 5% of days but represent 22% of total fire pixels.
Fire occurs under similar weather conditions but different circulation patterns.
a r t i c l e
i n f o
Article history:
Received 23 February 2015
Received in revised form 10 April 2015
Accepted 10 April 2015
Available online xxxx
Editor: D. Barcelo
Keywords:
Fire pixels
Extreme fire activity
MODIS
Fire weather
Atmospheric circulation patterns
Mediterranean
a b s t r a c t
Active fire information provided by TERRA and AQUA instruments on-board sun-synchronous polar MODIS
platform is used to describe fire activity in the Western Mediterranean and to identify and characterize the synoptic patterns of several meteorological fields associated with the occurrence of extreme fire activity episodes
(EEs). The spatial distribution of the fire pixels during the period of 2003–2012 leads to the identification of
two most affected sub-regions, namely the Northern and Western parts of the Iberian Peninsula (NWIP) and
Northern Africa (NAFR). The temporal distribution of the fire pixels in these two sub-regions is characterized
by: (i) high and non-concurrent inter- and intra-annual variability with maximum values during the summer
of 2003 and 2005 in NWIP and 2007 and 2012 in NAFR; and, (ii) high intra-annual variability dominated by a
prominent annual cycle with a main peak centred in August in both sub-regions and a less pronounced secondary
peak in March only evident in NWIP region. The 34 EEs identified were grouped according to the location,
period of occurrence and spatial configuration of the associated synoptic patterns into 3 clusters (NWIP-summer,
NWIP-winter and NAFR-summer). Results from the composite analysis reveal similar fire weather conditions
(statistically significant positive anomalies of air temperature and negative anomalies of air relative humidity)
but associated with different circulation patterns at lower and mid-levels of the atmosphere associated with
the occurrence of EEs in each cluster of the Western Mediterranean region.
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
Vegetation fires are an ecological disturbance occurring worldwide
with climatological, social and economic impacts at the global, regional
and local scales (Le Page et al., 2008; Oom and Pereira, 2013). Fire disturbance was classified as an “essential climate variable” by the Global
Climate Observing System that has highlighted the need for long data
⁎ Corresponding author at: Universidade de Trás-os-Montes e Alto Douro, Quinta de
Prados, 5000-801 Vila Real, Portugal.
E-mail addresses: [email protected] (M. Amraoui), [email protected] (M.G. Pereira),
[email protected] (C.C. DaCamara), [email protected] (T.J. Calado).
http://dx.doi.org/10.1016/j.scitotenv.2015.04.032
0048-9697/© 2015 Elsevier B.V. All rights reserved.
time series to quantify the links between climate and fire (GCOS-107,
2006).
In terms of accuracy and reliability, satellite information is an especially appropriate tool to monitor fire activity at the global level
(Justice and Korontzi, 2001). This is mainly due to the spatial and temporal homogeneity of remote-sensed data and to its independence
from fire policies of the different countries (Pereira et al., 2011;
Amraoui et al., 2013).
During the last decades several fire detection algorithms were developed using remote sensed data obtained from different sensors on
board polar orbiters (Dwyer et al., 2000; Stroppiana et al., 2000;
Justice et al., 2002; Giglio et al., 2003; Arino et al., 2005) or geostationary
satellites (Prins and Menzel, 1992; Calle et al., 2006; Roberts and
M. Amraoui et al. / Science of the Total Environment 524–525 (2015) 32–39
Wooster, 2008; Amraoui et al., 2010). In this context, the MODIS active
fire product (Justice et al., 2002) is especially appropriate in climatological studies since it provides consistent daily information about fire activity, at 1 km resolution over a period of 10 years (2003–2012)
(Csiszar et al., 2005; Oom and Pereira, 2013).
The Mediterranean basin is struck by a large number of devastating
fire events that burn hundreds of thousands of hectares of forests, shrub
and grasslands every year (Barbosa et al., 2007), with dramatic consequences for ecosystems and population (Pereira, 1999). The western
part of the Mediterranean basin which includes the Iberian Peninsula,
Southern France and Sardinia in Europe and Northern parts of
Morocco, Algeria and Tunisia in North Africa has been the most affected
by vegetation fires in the last decades (Schmuck et al., 2013). In the
Mediterranean region fire ignition is strongly conditioned by human
behaviour and socioeconomic activities (Costa et al., 2010) but natural
factors like the morphology of the landscape, land use, land cover, and
meteorological conditions have also to be taken into account
(Amraoui et al., 2013). In particular, weather and climate have a profound influence at all stages of biomass burning from ignition, spread
and behaviour up to severity and effects (Benson et al., 2009). The
vast majority of burned area in Mediterranean regions is due to a reduced number of extreme events that occur during a short period of
time, and are associated with several atmospheric processes interacting
at different temporal and spatial scales (Pereira et al., 2005). Particular
attention is therefore to be devoted to fire activity associated with
extreme events and to the roles played by climate and weather on
such extreme events.
In two previous studies, the authors used composite analysis of meteorological fields to identify and characterize the synoptic patterns associated with large fire episodes. The first study (Pereira et al., 2005)
focused on summer fire activity in Portugal in 1980–2000 and relied
on the official Portuguese database of burned area based on in situ
observations. The second study (Amraoui et al., 2013) focused on Italy
and Greece in summer 2007–2009 and was based on active fire
detections by Meteosat-8, a geostationary satellite operated by the
European Organisation for the Exploitation of Meteorological Satellites
(EUMETSAT).
As mentioned above, these two studies focused on the extreme fire
activity events that occurred only during the summer season. In addition and regarding to the first case, the study relied on the largest fire
database in Europe based on ground observations which is not usually
available for several regions around the Mediterranean basin, while
the second study is based on geostationary remote sensing data which
present low spatial resolution over the region (Pereira et al., 2011;
Amraoui et al., 2013). In this sense, the novelty of this study is to (i) extend the similar rationale to the detection and characterization of extreme fire episodes throughout the year and to expand the analysis to
North Africa; and, (ii) use the MODIS dataset which is particularly useful
to characterize fire activity at the regional scale (Western Mediterranean basin) thanks to its highest spatial resolution and relatively longterm database.
In this context, the specific aims of the study are to (i) characterize
the spatial and temporal distribution of fire pixels over Western Mediterranean during the 2003–2012 period including the identification of
extreme fire episodes throughout the year; and, (ii) assess the meteorological conditions associated with identified extreme episodes based on
the analysis of the atmospheric fields of mean sea level pressure,
geopotential, wind, air temperature and relative humidity.
2. Data and methods
2.1. Fire and meteorological data
Information about active fires was extracted from the MCD14ML
Collection 5 active fire product which relies on MODIS imagery data
(Justice et al., 2002). The dataset covers the 10-year study period,
33
from January 2003 to December 2012. Fire detection is performed
using a contextual algorithm that exploits the strong emission of midinfrared radiation from fires and is based on brightness temperatures
derived from the 3.9 and 10.5 μm channels (Giglio et al., 2003).
Meteorological information was derived from the ERA-Interim
dataset, the most recent global atmospheric re-analysis product of the
European Centre for Medium-Range Weather Forecasts (ECMWF). Detailed information about the product may be found in Dee et al.
(2011). Extracted information covers the 30-year period (1983–2012)
and consists of daily fields at 12 UTC, with a spatial resolution of
0.75° × 0.75° lat-lon, over the Western Mediterranean sector (30°W–
20°E, 20°N–60°N) of the following meteorological variables:
▪
▪
▪
▪
mean sea level pressure (hereafter MSLP);
air temperature at 2 m height (hereafter T2m);
wind speed and direction at 10 m (hereafter W10m);
air temperature, geopotential height and relative humidity at
850 hPa which corresponds to about 1500 m of altitude (hereafter
T850, Z850 and RH850, respectively);
▪ geopotential height at 500 hPa, in the mid atmosphere, at about
5 000 m of altitude (hereafter Z500).
2.2. Analysis of extreme fire activity episodes
The identification of a given extreme episode (hereafter referred to
as EE) of fire activity in a given region is defined based on the following
criteria: the episode should consist of at least 3 consecutive days, each
day having a minimum of 10 fire pixels over the region and each day
ranking first in number of fires among the respective 10 records for
that day of observations (covering the study period 2003–2012).
The role played by the meteorological conditions during EEs over a
given region is assessed by analysing the associated patterns of meteorological fields at different levels. Analysed patterns are anomaly composites, consisting of arithmetic means (performed over all days
associated with EEs) of daily departures of 12 UTC meteorological fields
from the respective average for that day over the reference period
(1983–2012).
The statistical significance (p-value) of obtained anomalies in each
grid point was evaluated by bootstrapping (Efron and Tibshirani,
1993). Anomaly fields are plotted only when statistically significant at
0.999 level, i.e. when values are higher (lower) than the 99.9th (0.1th)
percentile if positive (negative).
3. Results
3.1. Fire activity in the Western Mediterranean region
During the 10-year study period (2003–2012), a grand total of
126 887 fire pixels (all data used) were detected over the study area
(Fig. 1). The spatial distribution of fire pixels suggests defining two
sub-regions of high fire activity. The first sub-region (hereafter referred
to as NWIP) covers the Northern and Western parts of the Iberian Peninsula and accounts for 57% of the grand total of fire pixels. The second
region (hereafter referred to as NAFR) spreads over Northern Africa and
South-Eastern Iberia and contributes to 40% of the grand total. The vast
majority of fire pixels occur from May to October, and account for 82%
and 91% of the total number observed in NWIP and NAFR, respectively.
The temporal distribution of fire pixels reveals high inter-annual
variability (Fig. 2, upper panel), with 43% of the total number occurring
in three years, i.e. 13% in 2003, 15% in 2005 and 15% in 2012; 2008 is the
year with the lowest fire activity, representing just 5% of the grand total.
The contribution of NWIP to the grand total is larger than that of NAFR
in seven out of the 10 years analysed. The contribution of NWIP to the
total fire pixels each year is especially high in 2003 and 2005, reaching
34
M. Amraoui et al. / Science of the Total Environment 524–525 (2015) 32–39
Table 1
Extreme episodes (EEs) characteristics. The list of features includes: starting date, ending
date (Day/Month/Year), number of fire pixels, and the ratio between observed (OFP) and
expected (EFP) daily mean number of fire pixels in each month. OFP corresponds to the
daily mean number of fire pixels during the EEs. EFP is the daily mean of fire pixels after
removing the number of fire pixels during the EEs. The 34 EEs are located in the
abovementioned two sub-regions: NWIP (31 episodes: 18/13 in summer/winter periods)
and NAFR (3 episodes during summer).
NWIP summer
Month
Starting
date
Ending
date
N. fire
pixels
Fire pixels
ratio
May
11/05/03
29/05/06
02/05/09
19/06/03
13/06/04
07/06/05
04/06/06
25/07/04
08/07/05
20/07/05
02/08/03
19/08/05
06/08/06
12/09/03
26/09/04
02/09/12
11/10/11
20/10/11
01/11/07
22/12/05
27/01/08
24/01/11
08/01/12
30/01/05
13/02/05
17/02/05
21/02/12
19/03/09
09/03/12
26/03/12
02/04/12
28/06/12
15/07/05
27/08/07
14/05/03
31/05/06
05/05/09
21/06/03
17/06/04
10/06/05
06/06/06
28/07/04
13/07/05
22/07/05
04/08/03
23/08/05
12/08/06
16/09/03
30/09/04
04/09/12
18/10/11
22/10/11
14/11/07
24/12/05
29/01/08
26/01/11
11/01/12
03/05/05
15/02/05
19/02/05
02/03/12
25/03/09
12/03/12
01/04/12
06/04/12
01/07/12
17/07/05
30/08/07
98
91
86
277
308
235
177
1546
893
934
2495
3245
4632
1146
616
653
2021
276
651
54
46
44
56
85
103
65
748
831
332
936
211
922
684
2970
5
6
4
9
6
6
5
13
5
10
11
9
10
8
4
7
13
4
19
21
10
9
9
2
4
3
11
8
5
9
5
28
7
13
June
July
Fig. 1. Fire pixels over the Western Mediterranean region from January 2003 to December
2012. Coloured dots represent fire pixels belonging to extreme fire activity episodes
grouped by year of occurrence; grey dots correspond to the remaining fire pixels. Boundaries of NWIP and NAFR sub-regions are also shown.
August
September
71% and 75% respectively, whereas NAFR contributes with 65% and 64%
in 2007 and 2012.
There is also a prominent annual cycle of fire activity, showing a
peak in August that accounts for 36% and 41% of the grand total of fire
pixels detected in NWIP and NAFR, respectively. A secondary peak is
also observed in March in the case of NWIP, which represents about
8% of fire activity in that sub-region.
Despite the similar annual cycles of fire activity obtained (Fig. 2, bottom panel), the two sub-regions are differently affected by EEs, in number of events, time of the year and year of occurrence (Table 1). A total of
31 episodes were identified in NWIP, 18 in May–October and the remaining 13 in November–April; 3 episodes were identified in NAFR,
all of them occurring in summer (June–August). Differences between
NWIP and NAFR are also apparent when restricting the analysis to the
spatial distribution of fire pixels associated with EEs (Fig. 1), and 2003,
2005 and 2006 are conspicuous over NWIP, whereas 2005, 2007 and
2012 dominate over NAFR.
Although representing just 4.5% of the days covered by the study period, the number of days associated with EEs accounts for 28 467 of the
fire pixels, representing 22% of the grand total. The latter amount is distributed as follows by region and time of the year: 69% results from the
contribution of the 18 episodes in NWIP in May–October; 15% from the
Fig. 2. Inter-annual (upper panel) and intra-annual (lower panel) variability of fire activity
over NWIP and NAFR. The used data in the figure refers to all fire pixels detected over the
study region during the 2003–2012 period and includes fire pixels associated with EEs.
October
NWIP winter
November
December
January
February
March
NAFR
April
June
July
August
13 episodes in NWIP in November–April; and the remaining 16% from
the 3 episodes in NAFR that occurred in June 2012, July 2005 and August
2007.
In the case of NWIP, the years of 2005, with 8 events, presents the
largest number of extreme episodes and is followed by 2012 with 5
Fig. 3. Number of EEs per year (upper panel) and per month (lower panel) counted during
the 2003–2012 period for NWIP sub-region.
M. Amraoui et al. / Science of the Total Environment 524–525 (2015) 32–39
EEs (Fig. 3, upper panel). No EEs were observed in 2010 and only 1 episode was observed in 2007. Although EEs in NWIP are evenly distributed between the two considered periods of the year (i.e. 18 in May–
October and 13 in November–April) (Fig. 3, lower panel) they are responsible for a very different amount of fire pixels (i.e. 19 729 and
4162, respectively). However EEs in both periods account for one third
of total fire pixels in the respective periods (i.e. 59 418 and 12 765,
respectively).
3.2. The atmospheric conditions associated with extreme fire activity
episodes
Results obtained in the previous section suggest performing an assessment of the role played by meteorological conditions on the onset
and spreading of fire events in the Western Mediterranean. An exhaustive analysis of the composites and anomaly fields of all meteorological
variables listed in Section 2.2 was performed for each day of the 34 EEs.
A total number of 476 (=7 meteorological variables × 2 statistics × 34
EEs) plots were produced and analysed.
Composites and anomaly composites fields of meteorological variables obtained for all days of extreme fire activity in each sub-region
and season will be presented (Figs. 4–6). Each plot includes results obtained for more than one meteorological field and, for each cluster,
two figures of composites and two other figures of anomalies are presented. Two plots respect to meteorological fields near the surface
(T2m, MSLP and W10m) and the other two at 500 hPa and 850 hPa
levels (Z500 and RH850).
35
3.2.1. NWIP during summer (May to October)
The prevailing synoptic conditions at the surface (Fig. 4, upper left
panel) and at 850 hPa (not shown) are characterized by an amplification of the anticyclone of Azores, extending from the Atlantic to Central
Europe and by a sub-Saharan thermal low centred at South Algeria. This
circulation pattern leads to strong North-Eastern and South-Eastern advections of continental dry and warm air, locally known as Soão winds,
resulting in high average surface air temperatures (above 36 °C in some
episodes) together with extreme values of relative humidity (as low as
15%) over Western Iberia. Associated anomalies (Fig. 4, right panels)
reach values as large as 4 °C in surface temperature (exceeding 10 °C
in some events) and as low as − 20% in relative humidity at 850 hPa
(lower than −40% in several episodes).
Composites of the atmospheric flow at 500 hPa (Fig. 4, lower left
panel) reveal synoptic baroclinic activity with a pronounced ridge
with axes in the Southeast to Northwest direction over the Atlantic
Ocean and West of Iberia, forcing warmer and drier air mass into the
NWIP sub-region. The 500 level composite anomaly patterns (Fig. 4,
lower right panel), exhibit positive departures of 120 gpm enhancing
the subsidence of air into the Troposphere, reinforcing the increase of
air temperature through adiabatic heating at lower levels.
3.2.2. NWIP during winter (November to April)
The overall synoptic configuration presents similarities with the one
of NWIP during summer where due to the high Northwest-Southeast
gradient of geopotential height, both at surface and at upper levels,
the atmospheric circulation is dominated by: (i) North and North-
Fig. 4. Composites (left panels) and anomaly composites (right panels) for the EEs over the NWIP sub-region during summer period (May–October) of air temperature (°C) at 2 m (T2m),
mean air pressure (hPa) at sea level (MSLP), wind speed (m·s−1) and direction at 10 m (W10m) (upper panels), relative humidity (%) at 850 hPa level (RH850) and geopotential height
(gpm) at 500 hPa (Z500) (lower panels). Anomalies are only plotted for statistically significant values at the 99.9% level.
36
M. Amraoui et al. / Science of the Total Environment 524–525 (2015) 32–39
Fig. 5. As in Fig. 4, but respecting the EEs over the NWIP sub-region during winter period (November–April).
Eastern advection of continental dry air over Western France, Western
and Northern Parts of Iberia, and North of Morocco and; (ii) NorthWestern advection of wet air over North-Western Algeria, Tunisia and
Italy. These patterns are a consequence of anticyclonic blocking activity
over Eastern Atlantic defined as persistent and quasi-stationary
synoptic-scale high pressure systems, often accompanied by low pressure systems at lower latitudes, which interrupt the zonal flow and
the eastward progression of synoptic systems (Dunn‐Sigouin and Son,
2013). The blocking pattern is conspicuous in the composite of MSLP
(Fig. 5, upper left panel) and it may be noted that the anticyclone
evolves with height into a predominant omega pattern located west of
Europe, constituted by a very strong high-pressure ridge located on
the Central Atlantic and flanked by two troughs (Fig. 5, lower left
panel). The region covered by the omega block pattern experiences
dry weather for an extended period of time and therefore prolonged
droughts (García-Herrera et al., 2007).
The anomalous circulation patterns at the 500 hPa level are
dominated by positive departures of Z500 reaching the impressive
value of about 210 gpm (Fig. 5, lower right panel) west of the British
Isles. Anomalies of meteorological fields (Fig. 5, right panels) reveal
smaller values of T2m (2 °C) than for summer EEs but higher
values of RH850, lower than − 30% (reaching − 60% in some of the
cases).
3.2.3. NAFR during summer (May to October)
The atmospheric circulation at the surface over Northern Algeria and
Tunisia, is dominated by an intense Southerly/South-Easterly advection
of very hot and very dry Saharan air, brought to the region by the anticyclonic circulation, centred over Italy, and the cyclonic flow, due to the
presence of a Saharan heat low centred over South-Western Algeria and
to the Atlas Low located in Central Morocco (Fig. 6, upper left panel).
The mentioned advection process may be identified as the wellknown Shergui (in Morocco) or Chili (in Algeria) winds that may
occur primarily from February to September with peak occurrences
from July to August. The regions under the influence of these winds
experience extreme dryness and heat with extensive dust and haze
layers aloft (Donahue et al., 1996). The surface circulation patterns
over South-Eastern Spain are dominated by the convergence of: (i)
the abovementioned very hot and dry Southerly winds from North
Africa, named Leveche in Spain; and (ii) Easterly wind steered by the
anticyclonic circulation over the Atlantic, that become warmer and
drier as it crosses the Iberian Peninsula. It may be noted that the more
general term for Shergui, Chili and Leveche winds is the Scirocco. The associated high departure values of air temperature (above 8 °C) obtained
in both T2m (Fig. 6, upper right panel) and T850 (not shown), and very
low anomaly values of relative humidity at 850 hPa reaching the impressive value of − 40% (Fig. 6, lower right panel) are well apparent
over this sub-region.
The flow at the 500 hPa level (Fig. 6, lower left panel) is dominated
by a pronounced ridge whose axis is directed from Southwest to Northeast, affecting Northern Algeria and Tunisia and bringing warmer and
drier air mass to the region. The air subsidence associated with this
ridge further contributes to the increase of air temperature throughout
adiabatic heating at low levels.
M. Amraoui et al. / Science of the Total Environment 524–525 (2015) 32–39
37
Fig. 6. As in Fig. 4, but respecting the EEs over the NAFR sub-region during summer period (June–August).
4. Discussion and concluding remarks
Due to its spatial and temporal homogeneity, accuracy and reliability, remote sensing data obtained by the MODIS radiometer on-board
the polar-orbiter satellites, namely TERRA and AQUA, are especially
suitable for monitoring vegetation fires (Justice and Korontzi, 2001;
Oom and Pereira, 2013). In addition, the relatively long period of operability of the abovementioned satellites allows providing long and precise database at local, regional and even at global scales. Furthermore,
the present study demonstrated that information on fire activity derived from these satellites is particularly useful for studying the extreme
episodes of vegetation fires in the Iberian Peninsula and Northern Africa,
particularly struck by this phenomenon.
The spatial distribution of fire pixels detected by MODIS over the
Western Mediterranean region during the 2003–2012 period (Fig. 1) reveals the existence of two sub-regions particularly affected by this natural hazard: the Western and Northern parts of Iberian Peninsula; and
South-Eastern Iberia and Northern region of Algeria and Tunisia, in
North of Africa.
For each region, EEs were identified based on a criterion that requires a minimum of three consecutive days, each day requires to
have a minimum of 10 fire pixels detected and having to rank first
among the 10 records for that day in the study period (2003–2012). Instead of adopting a simple procedure where extreme events are identified as the largest ones recorded in the entire study period, or even at
the annual scale, the criterion adopted allows identifying extreme
events at daily scale throughout the year. While the former simple procedure would lead to the detection of the largest summer events mainly
in Northwest Iberia, the criterion adopted in this study also allows
unveiling other extreme episodes in other sub-regions and seasons,
namely the existence of a secondary peak of fire activity in NWIP region
at the end of the winter semester and high fire activity in North Africa.
EEs were identified in all months of the year, February and June being
the months that presented the higher frequency. The years of 2005
and 2012 were those presenting the highest number of EEs.
The secondary peak of fire activity observed late winter and early
spring in NWIP is essentially of anthropogenic nature, as fires are caused
by land-use practices, negligence or even arson (Krawchuk and Moritz,
2014). Agricultural burnings are applied world-wide for soil fertilization, to prepare fields for harvest work and to dispose of crop residues
(Yevich and Logan, 2003; Korontzi et al., 2006). Besides these human
pyrogenic behaviours, weather conditions, at intra-seasonal scale, also
play a crucial role on the fire activity by controlling fire ignition, fire
spread and fire extinction (Le Page et al., 2010).
Atmospheric patterns associated with EEs in the different subregions were characterized following a methodology previously validated for Portugal (Pereira et al., 2005; Trigo et al., 2013) as well as for Italy
and Greece (Amraoui et al., 2013). Large fire activity in NWIP mostly results from the contribution of North-western Iberia, encompassing
Northern Portugal and the Spanish region of Galicia. The atmospheric
circulation associated with EEs is characterized by predominant easterlies crossing that advect continental air into the sub-region. During
summer, the circulation is characterized by a predominant hot and
dry flow from continental Europe and north Africa, contributing to
values of air temperature (relative humidity) well above (below) average. During winter, the extremely low values of relative humidity are
38
M. Amraoui et al. / Science of the Total Environment 524–525 (2015) 32–39
the most conspicuous feature, which result from the anomalous Eastward circulation over the cold and dry Western Europe. The main
drivers of the atmospheric circulation are nevertheless different in summer and winter. In the latter season, the mid Troposphere is characterized by a strong omega blocking configuration located west of Europe,
enhancing the effect of the anticyclone over Western and Northern Iberia. In summer, the atmospheric configuration is dominated by a presence of an extended ridge over the North-Eastern Atlantic, associated
with the Azores anticyclone, and to the occurrence of heat waves. In
the case of NAFR, the atmospheric patterns associated with the occurrence of EEs during summer are dominated by South-Easterly very hot
and dry Scirocco winds at the surface and by a pronounced ridge over
Central Mediterranean at 500 hPa, advecting warm and dry air into
the sub-region.
The observed atmospheric circulation and the meteorological variables associated with the NWIP summer season EEs are in close agreement with the results of Pereira et al. (2005) and with the findings of
Trigo et al. (2006). This fact validates the methodology adopted and
proves the usefulness of the MODIS active fire data applied to such studies. Furthermore, this study also demonstrates that these patterns associated with typical summer EEs are not only limited to the three months
of the referred season (June, July and August) but extend over a longer
period of 6 months (May to October), corresponding to the dry season
of the studied region. However, the novelty of this study is the identification of the atmospheric patterns associated with (i) NWIP winter EEs,
extending from November to April; and (ii) North Africa summer EEs. In
the first case, and as mentioned above, the extreme episodes of fire activity are the consequence of the anomalous Eastward circulation over
the region which leads to long periods with no precipitation and extremely low values of relative humidity. In the second case, the occurrence of EEs is the result of the South-Eastern very hot and dry
Saharan winds and a pronounced ridge over Central Mediterranean at
500 hPa. The latter atmospheric configuration is very similar to the
one observed over the Balkan Peninsula and Italy when the two extreme events of fire activity, which struck the referred region during
the summer of 2007, were analysed (Founda and Giannakopoulos,
2009; Amraoui et al., 2013).
As in other regions of the globe, the intra- and inter-annual variability of fire activity is related to the type of climate (Fauria and Johnson,
2008; Benson et al., 2009; Costa et al., 2010; Drobyshev et al., 2012)
and to the occurrence of extreme weather conditions (Fauria and
Johnson, 2006; Cardil et al., 2011; García-Ortega et al., 2011). On the
one hand, the temperate type of climate (Csa: temperate climate with
hot and dry summer, Csb: temperate climate with warm and dry summer and Cfb: temperate climate without dry season and warm summer)
of the Mediterranean basin (Peel et al., 2007) is characterized by wet
and mild winters followed by warm and dry weather during summer,
favouring the growth of vegetation during winter and springtime,
followed by severe hydric and thermal stress during summer months
(Lichtenthaler, 1996; Lindner et al., 2010; Pereira et al., 2013). On the
other hand the most catastrophic fire events and, consequently the
highest annual amounts of burnt area are associated with extreme atmospheric conditions (Pereira et al., 2005; Trigo et al., 2006; Good
et al., 2008; Amraoui et al., 2013).
However, there is not a cause–effect relationship between atmospheric conditions and fire activity in the sense that fire ignition in the
Mediterranean region is mainly due to human activities (Ganteaume
et al., 2012; San-Miguel-Ayanz et al., 2013). Instead, weather conditions
during and before the fire events should be viewed as a primary factor of
fire danger in this region because of their influence on fuel availability
and vegetation hydric and thermal stress, particularly during severe
regional summer heatwaves that are highly correlated with the
occurrence of the most devastating fires in the Mediterranean
(Lichtenthaler, 1996; Ciais et al., 2005; Trigo et al., 2006; Peñuelas
et al., 2007; Lindner et al., 2010). In addition, like in many other ecosystems, a small number of large fires are responsible for the major fraction
of total burned area in the Mediterranean biome (Strauss et al., 1989;
Pereira et al., 2011). For this reason, knowing the spatial configuration
of the meteorological fields associated with highest fire activity is of paramount importance for the planning of fire prevention activities and
management of fire suppression resources. In this sense, the findings
of this study can be used to improve projections of fire activity and forestry for the Western Mediterranean (Mouillot et al., 2002; Maracchi
et al., 2005; Bowman et al., 2009; Dury et al., 2010; Pereira et al., 2013).
Acknowledgments
This work was supported by national funds by FCT — Portuguese
Foundation for Science and Technology, under project PEst-OE/AGR/
UI4033/2014 and by project “SUSTAINSYS: Environmental Sustainable
Agro-Forestry Systems” — NORTE-07-0124-FEDER-0000044.
References
Amraoui, M., DaCamara, C.C., Pereira, J.M.C., 2010. Detection and monitoring of African
vegetation fires using MSG-SEVIRI imagery. Remote Sens. Environ. 114, 1038–1052.
Amraoui, M., Liberato, M.L.R., Calado, T.J., DaCamara, C.C., Pinto Coelho, L., Trigo, R.M.,
Gouveia, C.M., 2013. Fire activity over Mediterranean Europe based on information
from Meteosat-8. For. Ecol. Manag. 294, 62–75.
Arino, O., Plummer, S., Defrenne, D., 2005. Fire disturbance: the ten years time series of
the ATSR World Fire Atlas. In: Lacoste, H. (Ed.), Proceedings of the MERIS (A)ATSR
Workshop. September 26–30. ESRIN, Frascati.
Barbosa, P., Amatulli, G., Boca, R., Camia, A., Kucera, J., Libertà, G., San-Miguel Ayanz, J.,
Schmuck, G., Schulte, E., Dierks, H.-H., 2007. Forest fires in Europe 2006. EUR 22931
EN — Joint Research Centre — Institute for Environment and Sustainability. EUR,
Scientific and Technical Research seriesOffice for Official Publications of the
European Communities, Luxembourg (ISSN 1018–5593).
Benson, R.P., Roads, J.O., Wiese, D.R., 2009. Climatic and weather factors affecting fire occurrence and behaviour. In: Bytnerowicz, A., Arbaugh, M., Riebau, A., Andersen, C.
(Eds.), Developments in Environment Science 8, pp. 37–59. http://dx.doi.org/10.
1016/S1474-8177(08)00002-8 (ISSN: 1474–8177).
Bowman, D.M.J.S., Balch, J.K., Artaxo, P., Bond, W.J., Carlson, J.M., Cochrane, M.A.,
D'Antonio, C.M., DeFries, R.S., Doyle, J.C., Harrison, S.P., Johnston, F.H., Keeley, J.E.,
Krawchuk, M.A., Kull, C.A., Marston, J.B., Moritz, M.A., Prentice, I.C., Roos, C.I., Scott,
A.C., Swetnam, T.W., van der Werf, G.R., Pyne, S.J., 2009. Fire in the Earth system.
Science 324, 481–484.
Calle, A., Casanova, J.L., Romo, A., 2006. Fire detection and monitoring using MSG spinning
enhanced visible and infrared imager (SEVIRI) data. J. Geophys. Res. 111, G04S06.
http://dx.doi.org/10.1029/2005JG000116G04S06.
Cardil, A., Molina, D.M., Ramirez, J., Vega-García, C., 2011. Trends in adverse weather patterns and large wildland fires in Aragón (NE Spain) from 1978 to 2010. Nat. Hazards
Earth Syst. Sci. 13, 1393–1399.
Ciais, Ph., Reichstein, M., Viovy, N., Granier, A., Ogée, J., Allard, V., Aubinet, M., Buchmann,
N., Bernhofer, Chr., Carrara, A., Chevallier, F., De Noblet, N., Friend, A.D., Friedlingstein,
P., Grünwald, T., Heinesch, B., Keronen, P., Knohl, A., Krinner, G., Loustau, D., Manca,
G., Matteucci, G., Miglietta, F., Ourcival, J.M., Papale, D., Pilegaard, K., Rambal, S.,
Seufert, G., Soussana, J.F., Sanz, M.J., Schulze, E.D., Vesala, T., Valentini, R., 2005.
Europe-wide reduction in primary productivity caused by the heat and drought in
2003. Nature 437 (7058), 529–533.
Costa, L., Thonicke, K., Poulter, B., Badek, F.W., 2010. Sensitivity of Portuguese forest fires
to climatic, human, and landscape variables: subnational differences between fire
drivers in extreme fire years and decadal averages. Reg. Environ. Chang. 11 (3),
543–551. http://dx.doi.org/10.1007/s10113-010-0169-6.
Csiszar, I., Denis, L., Giglio, L., Justice, C.O., Hewson, J., 2005. Global fire activity from two
years of MODIS data. Int. J. Wildland Fire 14, 117–130. http://dx.doi.org/10.1071/
WF03078.
Dee, D.P., Uppala, S.M., Simmons, A.J., Berrisford, P., Poli, P., Kobayashi, S., Andrae, U.,
Balmaseda, M.A., Balsamo, G., Bauer, P., Bechtold, P., Beljaars, A.C.M., van de Berg, L.,
Bidlot, J., Bormann, N., Delsol, C., Dragani, R., Fuentes, M., Geer, A.J., Haimberger, L.,
Healy, S.B., Hersbach, H., Hólm, E.V., Isaksen, L., Kållberg, P., Köhler, M., Matricardi,
M., McNally, A.P., Monge-Sanz, B.M., Morcrette, J.-J., Park, B.-K., Peubey, C., de
Rosnay, P., Tavolato, C., Thépaut, J.-N., Vitart, F., 2011. The ERA-Interim reanalysis:
configuration and performance of the data assimilation system. Q. J. R. Meteorol.
Soc. 137, 553–597. http://dx.doi.org/10.1002/qj.828.
Donahue, C.A., Whitney, L.D., Walters, K.R., Cloys, K.P., Louer, J.W., Surls, C.D., 1996. Northwest Africa — a climatological study. Air Force Combat Climatology Center (AFCCC),
Scott AFB IL 62225–5116Technical Study report, AFCCC7TN-96/012 (126 pp.).
Drobyshev, I., Niklasson, M., Linderholm, H.W., 2012. Forest fire activity in Sweden: climatic controls and geographical patterns in 20th century. Agric. For. Meteorol.
154–155, 174–186.
Dunn‐Sigouin, E., Son, S.W., 2013. Northern Hemisphere blocking frequency and duration
in the CMIP5 models. J. Geophys. Res. Atmos. 118 (3), 1179–1188.
Dury, M., Hambuckers, A., Warnant, P., Henrot, A., Favre, E., Ouberdous, M., François, L.,
2010. Responses of European forest ecosystems to 21st century climate: assessing
changes in interannual variability and fire intensity. iForest 4, 82–99.
M. Amraoui et al. / Science of the Total Environment 524–525 (2015) 32–39
Dwyer, E., Pinnock, S., Grégoire, J.-M., Pereira, J.M.C., 2000. Global spatial and temporal
distribution of vegetation fire as determined from satellite observations. Int.
J. Remote Sens. 21, 1289–1302.
Efron, B., Tibshirani, R.J., 1993. An Introduction to the Bootstrap. Chapman & Hall, New
York.
Fauria, M.M., Johnson, E.A., 2006. Large-scale climatic patterns control large lightning fire
occurrence in Canada and Alaska forest regions. J. Geophys. Res. 111, G04008. http://
dx.doi.org/10.1029/2006JG000181.
Fauria, M.M., Johnson, E.A., 2008. Climate and wildfires in the North American boreal forest. Philos. Trans. R. Soc. B 363, 2317–2329.
Founda, D., Giannakopoulos, C., 2009. The exceptionally hot summer of 2007 in Athens,
Greece — a typical summer in the future climate. Glob. Planet. Chang. 67, 227–236.
http://dx.doi.org/10.1016/j.gloplacha.2009.03.013.
Ganteaume, A., Camia, A., Jappiot, M., San-Miguel-Ayanz, J., Long-Founel, M., Corinne, L.,
2012. A review of the main driving factors of forest fie ignition over Europe. Environ.
Manag. 51, 651–662.
García-Herrera, R., Paredes, D., Trigo, R.M., Trigo, I.F., Hernández, E., Barriopedro, D.,
Mendes, M.A., 2007. The outstanding 2004/05 drought in the Iberian Peninsula: associated atmospheric circulation. J. Hydrometeorol. 8, 483–498. http://dx.doi.org/10.
1175/JHM578.1.
García-Ortega, E., Trobajo, M.T., López, L., Sánchez, J.L., 2011. Synoptic patterns associated
with wildfires caused by lightning in Castile and Leon, Spain. Nat. Hazards Earth Syst.
Sci. 11, 851–863.
GCOS-107, 2006. Systematic observation requirements for satellite-based products for climate — supplemental details to the satellite-based component of the “Implementation Plan for the Global Observing System for Climate in Support of the UNFCCC”.
http://www.wmo.int/pages/prog/gcos/Publications/gcos-107.pdf.
Giglio, L., Descloitres, J., Justice, C.O., Kaufman, Y.J., 2003. An enhanced contextual fire detection algorithm for MODIS. Remote Sens. Environ. 87, 273–282.
Good, P., Moriondo, M., Giannakopoulos, C., Bindi, M., 2008. The meteorological conditions associated with extreme fire risk in Italy and Greece: relevance to climate
model studies. Int. J. Wildland Fire 17, 155–165.
Justice, C.O., Korontzi, S.A., 2001. A review of the status of satellite fire monitoring and the
requirements for global environmental change research. In: Ahern, F.J., Goldammer,
J., Justice, C.O. (Eds.), Global and Regional Wildland Fire Monitoring From Space:
Planning a Coordinated International Effort. SPB Academic Publishing, The Hague.
Justice, C.O., Giglio, L., Korontzi, S., Owens, J., Morisette, J.T., Roy, D., Descloitres, J.,
Alleaume, S., Petitcolin, F., Kaufman, Y., 2002. The MODIS fire products. Remote
Sens. Environ. 83, 244–262.
Korontzi, S., McCarty, J., Loboda, T., Kumar, S., Justice, C., 2006. Global distribution of agricultural fires in croplands from 3 years of Moderate Resolution Imaging
Spectroradiometer (MODIS) data. Glob. Biogeochem. Cycles 20, G-2021.
Krawchuk, M.A., Moritz, M.A., 2014. Burning issues: statistical analyses of global fire data
to inform assessments of environmental change. Environmetrics 25, 472–481.
Le Page, Y., Pereira, J.M.C., Trigo, R.M., DaCamara, C.C., Oom, D., Mota, B., 2008. Global fire
activity patterns (1996–2006) and climatic influence: an analysis using the World
Fire Atlas. Atmos. Chem. Phys. 8, 1911–1924.
Le Page, Y., Oom, D., Silva, J.M.N., Jönsson, P., Pereira, J.M.C., 2010. Seasonality of vegetation fires as modified by human action: observing the deviation from eco-climatic
fire regimes. Glob. Ecol. Biogeogr. 19, 575–588.
Lichtenthaler, H.K., 1996. Vegetation stress: an introduction to the stress concept in
plants. J. Plant Physiol. 148 (1–2), 4–14.
Lindner, M., Maroschek, M., Netherer, S., Kremer, A., Barbati, A., Garcia-Gonzalo, J., Seidl,
R., Delzon, S., Corona, P., Kolsträm, M., Lexer, M.J., Marchettiet, M., 2010. Climate
change impacts, adaptive capacity, and vulnerability of European forest ecosystems.
For. Ecol. Manag. 259 (4), 698–709.
39
Maracchi, G., Sirotenko, O., Bindi, M., 2005. Impacts of present and future climate variability on agriculture and forestry in the temperate regions: Europe. Clim. Chang. 70
(1–2), 117–135.
Mouillot, F., Rambal, S., Joffre, R., 2002. Simulating climate change impacts on fire frequency and vegetation dynamics in a Mediterranean-type ecosystem. Glob. Chang.
Biol. 8, 423–437.
Oom, D., Pereira, J.M.C., 2013. Exploratory spatial data analysis of global MODIS active fire
data. Int. J. Appl. Earth Obs. Geoinf. 21, 326–340.
Peel, M.C., Finlayson, B.L., McMahon, T.A., 2007. Updated world map of the Köppen-Geiger
climate classification. Hydrol. Earth Syst. Sci. 11, 1633–1644. http://dx.doi.org/10.
5194/hess-11-1633-2007.
Peñuelas, J., Prieto, P., Beier, C., Cesaraccio, C., De Angelis, P., De Dato, G., Emmett, B.A.,
Estiarte, M., Garadnai, J., Gorissen, A., Láng, E.K., Kröel-Dulay, G., Llorens, L.,
Pellizzaro, G., Riis-Nielsen, T., Schmidt, I.K., Sirca, C., Sowerby, A., Spano, D., Tietema,
A., 2007. Response of plant species richness and primary productivity in shrublands
along a North–South gradient in Europe to seven years of experimental warming
and drought: reductions in primary productivity in the heat and drought year of
2003. Glob. Chang. Biol. 13, 2563–2581.
Pereira, J.M.C., 1999. A comparative evaluation of NOAA/AVHRR vegetation indexes for
burned surface detection and mapping. IEEE Trans. Geosci. Remote Sens. 37,
217–226.
Pereira, M.G., Trigo, R.M., DaCamara, C.C., Pereira, J.M.C., Leite, S.M., 2005. Synoptic patterns associated with large summer forest fires in Portugal. Agric. For. Meteorol.
129, 11–25.
Pereira, M.G., Malamud, B.D., Trigo, R.M., Alves, P.I., 2011. The history and characteristics
of the 1980–2005 Portuguese rural fire database. Nat. Hazards Earth Syst. Sci. 11 (12),
3343–3358.
Pereira, M.G., Calado, T.J., DaCamara, C.C., Calheiros, T., 2013. Effects of regional climate
change on rural fires in Portugal. Clim. Res. 57, 187–200.
Prins, E.M., Menzel, W.P., 1992. Geostationary satellite detection of biomass burning in
Southern America. Int. J. Remote Sens. 13, 2783–2799.
Roberts, G.J., Wooster, M.J., 2008. Fire detection and fire characterization over Africa using
Meteosat SEVIRI. IEEE Trans. Geosci. Remote Sens. 46, 1200–1218. http://dx.doi.org/
10.1109/TGRS.2008.915751.
San-Miguel-Ayanz, J., Moreno, J.M., Camia, A., 2013. Analysis of large fires in European
Mediterranean landscapes: lessons learned and perspectives. For. Ecol. Manag. 294,
11–22.
Schmuck, G., San-Miguel-Ayanz, J., Camia, A., Durrant, T., Boca, R., Libertá, G., Schulte, E.,
2013. Forest Fires in Europe, Middle East and North Africa 2012. Publications Office
of the European Union, Luxembourg.
Strauss, D., Bednar, L., Mees, R., 1989. Do one percent of the forest fires cause ninety-nine
percent of the damage? For. Sci. 35 (2), 319–328.
Stroppiana, D., Pinnock, S., Grégoire, J.-M., 2000. The global fire product: daily fire occurrence from April 1992 to December 1993 derived from NOAA AVHRR data. Int.
J. Remote Sens. 21, 1279–1288.
Trigo, R.M., Pereira, J.M., Pereira, M.G., Mota, B., Calado, M.T., DaCamara, C.C., Santo, F.E.,
2006. The exceptional fire season of summer 2003 in Portugal. Int. J. Climatol. 26,
1741–1757.
Trigo, R.M., Sousa, P.M., Pereira, M.G., Rasilla, D., Gouveia, C.M., 2013. Modelling wildfire
activity in Iberia with different atmospheric circulation weather types. Int.
J. Climatol. http://dx.doi.org/10.1002/joc.3749.
Yevich, R., Logan, J.A., 2003. An assessment of biofuel use and burning of agricultural
waste in the developing world. Glob. Biogeochem. Cycles 17, 1095.