Carbon emissions in Mediterranean shrubland wildfires: An experimental approach

Carbon emissions in Mediterranean shrubland wildfires: An experimental approach

Atmospheric Environment 69 (2013) 86e93 Contents lists available at SciVerse ScienceDirect Atmospheric Environment journal homepage: www.elsevier.co...

368KB Sizes 0 Downloads 58 Views

Atmospheric Environment 69 (2013) 86e93

Contents lists available at SciVerse ScienceDirect

Atmospheric Environment journal homepage: www.elsevier.com/locate/atmosenv

Carbon emissions in Mediterranean shrubland wildfires: An experimental approach Elisa Garcia-Hurtado a, *, Jorge Pey b, c, M. Jaime Baeza a, Arnaud Carrara a, Joan Llovet a, Xavier Querol b, Andrés Alastuey b, V. Ramon Vallejo a a b c

Instituto Universitario Fundación Centro de Estudios Ambientales del Mediterráneo (CEAM-UMH), 46980 Paterna, Valencia, Spain Institute of Environmental Assessment and Water Research, Spanish Research Council (IDÆA-CSIC), Jordi Girona, 18-26, Barcelona 08034, Spain Department of Epidemiology Lazio Region, via S. Costanza, 53, Roma 00198, Italy

h i g h l i g h t s < The experiments were performed to guarantee the conditions of a real fire. < In situ monitoring measures have allowed estimating the percentage of carbon sampled. < Low combustion efficiencies due to the structural composition of the shrubs. < Experimental emission data will improve the information in emission inventories.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 20 July 2012 Received in revised form 23 November 2012 Accepted 29 November 2012

Forest fire emissions modify the chemical composition of the atmosphere and the earth’s climate system. The Ayoraburning experiment was designed to assess and quantify fire emissions from Mediterranean shrublands. A number of gaseous pollutants and particulate matter metrics (CO2, CO, CH4, PM2.5) were measured during 3 burning replicates by using real-time monitors. Quantification of carbon emissions released during the experiments showed that 71% was CO2, 26% CO, 3% CH4, and only 0.3% was particulate carbon. Emission factors obtained for CO2, CO and CH4 were 1257  40, 453  28 and 46  12 g kg1 dry matter, respectively, and combustion efficiencies ranged from 0.46 to 0.99. The experiments allowed the estimation of carbon emission in the different fire phases. Thus, 25% of carbon was sampled in the flaming phase and 75% of C in the smoldering phase. Current natural greenhouse gas (GHG) emission inventories in Mediterranean countries underestimate the actual emissions from forest fires since they do not consider forest shrub understory and shrublands and since they assume that the CO2 emitted is offset by forest re-growth. Our results may be used to improve current forest-fire emission inventories in southern Europe with special emphasis on shrublands. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: Emission factor Shrubland Combustion efficiency Carbon contribution Emission inventory

1. Introduction Wildfires have significantly increased in the European Mediterranean countries since the last quarter of the 20th century (Pausas et al., 2008; Miranda et al., 2009). Similar trends have been observed in other regions of the world (Bowman et al., 2009; Alves et al., 2010). As a consequence of the increase in wildfire frequencies in Mediterranean-climate regions, forests are being replaced by shrublands, and thus, their burnings are becoming much more

* Corresponding author. E-mail address: [email protected] (E. Garcia-Hurtado). 1352-2310/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.atmosenv.2012.11.063

significant (Vallejo and Alloza, 1998; Acácio et al., 2009). In southern Europe, wildfires in shrublands accounted for 73% in France (Promethee, 2012) of total burned area, 69% in Spain (MAGRAMA, 2012), 59% in Portugal (JRC, 2012), and 53% in Italy (DCCV, 2010). In addition, some species in Mediterranean shrublands also form part of the understory of forests. Fire-prone Mediterranean shrublands may accumulate a high fuel load, up to 3600 g m2, with a large proportion of dead fuel (Baeza et al., 2006), and they host a relatively high diversity of plant species with a diverse biochemical composition affecting the characteristics of fire emissions. Biomass burning is one of the largest sources of green-house gas (GHG) emission to the atmosphere and plays an important role in the

E. Garcia-Hurtado et al. / Atmospheric Environment 69 (2013) 86e93

biogeochemical cycles of carbon (C) and nitrogen (N) (Houghton et al., 1999). The main gaseous compounds released to the atmosphere when combustion is complete are carbon dioxide (CO2) and water (H2O). Real fires produce incomplete combustion leading to the emission of H2O, CO2, carbon monoxide (CO), methane (CH4) and other hydrocarbons (Urbanski et al., 2009). The composition of the smoke is dependent on the type of fuel and on the characteristics of the fire. Generally, re-growth of burned areas compensates for most of the CO2 released during the biomass burning event via photosynthetic activity. However, CO2 assimilation by new vegetation covering the burned areas may last several decades to offset the CO2 released during the forest fire, especially in drought-stricken climates such as the Mediterranean. If current climate change projections (IPCC, 2007) are taken into account, the intensification of extreme environmental conditions (higher frequency of droughts and heat-waves) would increase the probability of forest fires, triggering climate change in a positive feedback process. Forest fire emissions are one of the sources in the inventories of emissions, e.g. Kyoto protocol and they are represented by the emission factor (EF). These emissions are also needed to quantify effects on the air quality (Martins et al., 2012) and therefore impacts of fire emissions on human health (Miranda et al., 2010). In recent decades, a number of individual characterizations of wildfire and experimental forest fire emissions have been carried out, yielding some compilations on EF. The lists of chemical compounds analyzed are fairly similar for various forest types. Some lists only considered particulate matter (PM) (Hays et al., 2005), others considered organic compounds (Leumieux et al., 2004), but the most complete classifications include EF according to the type of biomass burned (Andreae and Merlet, 2001; Urbanski et al., 2009). The data collected by Andreae and Merlet (2001) classified EF according to numerous biomes, i.e. savannah and grassland, tropical forest, and extra-tropical forest. The latest collection published by Urbanski et al. (2009) includes temperate forest, temperate rangeland and boreal forest but not the emissions from Mediterranean forest fires. Experimental burnings in Mediterranean shrublands were conducted in the California chaparral (Hardy et al., 1996), in Corsica (Barboni et al., 2010), and in Portugal (Alves et al., 2010). The aim of these studies was to obtain a detailed characterization of smoke emissions from a Mediterranean shrubland. Alves et al. (2010) made a detailed characterization of the inorganic elements and soluble ions, and Vicente et al. (2011) studied the main gaseous and organic compounds in PM. Barboni et al. (2010) reported emissions of volatile and semi-volatile organic compounds. The information obtained from laboratory experiments of Mediterranean vegetation is provided by Ciccioli et al. (2001), who focused on volatile organic compounds (VOCs). In addition, Weise et al. (1991) analyzed gaseous (CO2 and CO) and PM emissions of common chaparral shrubs, and described the interactions between the emissions and the time of the year and the species burned. This study is a contribution for a geographical area greatly affected by devastating forest fires, and tries to provide experimental data to support emission inventories. Our research sought to determine the EF of major carbonaceous species in gas (CO2, CO, CH4) and fine particle phase organic and elemental carbon (OCPM2.5 and ECPM2.5) released during Mediterranean shrubland fires at ground level. 2. Materials and methods 2.1. AYORABURNING project The AYORABURNING project was a comprehensive field experiment guided by four main goals: (1) evaluation of the impact of

87

wildfires on C balance and soil C dynamics in Mediterranean shrublands; (2) physico-chemical characterization of smoke emissions in typical Mediterranean wildfires; (3) determination of GHG emission factors for the Mediterranean to improve country-based inventories; and (4) analysis of post-fire ecosystem dynamics (not addressed in this paper). The experiment was conducted in Ayora, Valencia (Eastern Spain) 39 7.6090 N; 1 1.4830 O; 705 m a.s.l., on 3 consecutive days, from 22 to 24 April 2009. The experiment was performed in April exclusively for security reasons, to minimize the risk of accidental propagation of fire. Fires were conducted under controlled conditions with the help of firefighters and previously delimited by a 5e6-m-wide fire break in which the vegetation was eliminated through mechanical brushing. The burns took place under safe weather conditions: low wind speed (less than 10 km h1) and high relative humidity, following a week of rain. Therefore, fire intensity was expected to be moderate to low (early spring conditions, relatively high fuel moisture) and the smoke column was expected to be fairly vertical. The duration of each experimental fire was about 50 min. Temperatures at ground level were recorded on-line during the fires using 12 thermocouples per fire. Taking into account that fires were conducted under controlled conditions EFs’ obtained in these experiments represent very specific spring conditions, and cannot be generalized for the more common summertime wildfires. 2.2. Plot characteristics The Ayora region is covered by typical Mediterranean vegetation. Pine forests with a shrubby understory prevailed until 1979, when a wildfire ravaged around 35,000 ha. Since that event, new fires have occurred in the region giving way to a landscape dominated by shrublands and sparse pine woodlands. The experimental site was a pine forest growing in old field, which was affected by the large wildfire in the summer of 1979. At the time of experimental burning, vegetation that regenerated after the 1979 fire was a dense 30-year old shrubland dominated by Mediterranean gorse (Ulex parviflorus), with Kermes oak (Quercus coccifera), rosemary (Rosmarinus officinalis), juniper (Juniperus oxycedrus), rock roses (Cistus albidus), heath (Erica multiflora) and a grass species (Brachypodium retusum). Mean vegetation height was 1.5 m and approximately 1/3 of the standing fuel was dead, typical of a senescent Mediterranean gorse shrubland in transition to rosemary shrubland (Baeza et al., 2007). Three replicated plots, 30  35 m each, with the same structural characteristics were delimited to conduct the experimental fires. Each plot was burned and sampled individually to provide three independent observations. An unburned area was used as reference plot for comparison before the fire. For each plot, fuel load and fuel characteristics were estimated by destructive sampling of both above-ground vegetation and litter in 6 randomly selected squares of 1  1m. Clipped plant material was separated by species. At the same time, litter was sampled in six squares of 0.25  0.25 m per site. The samples were taken to the laboratory, oven-dried at 80  C for 24 h and weighed. Similarly, the plots were sampled for the remaining fuel immediately after the fire. The total fuel consumed for each experimental fire was calculated using the difference between the pre-fire fuel load and the fuel remaining unburned after the fire. A number of meteorological parameters were measured before and during the experiment with a weather station installed at the site. Meteorological conditions during the experiments were characterized by moderate temperatures (around 20  C), low wind speed (less than 10 km h1), and moderate relative air humidity

88

E. Garcia-Hurtado et al. / Atmospheric Environment 69 (2013) 86e93

(around 50%). Soil moisture was determined before each fire at 6 sampling points (0e2.5 cm depth) by the gravimetric method. 2.3. Instruments and sampling procedures of atmospheric emissions A sampling device was designed to convey the emissions released during the fire experiments (Fig. 1) to a chamber. An aluminum tube 300 mm in diameter was suspended from a 7-m high steel tower and was positioned directly above the active combustion plume. This tube was connected to a sampling chamber of 6 m3, which contained a number of instruments for particle sampling: an optical counter GRIMM 1108, a Condensation Particle Counter (CPC), 3 high volume samplers for TSP, PM10 and PM2.5 collection, and cartridge samplers. Furthermore, real-time gaseous monitors were installed outside the chamber, but sampling inlets were placed inside. Behind the chamber was an extractor pump which conveyed the smoke from the plume to the chamber with an average speed of 5 m s1 at the sample line inlet corresponding to a total flow of about 0.35 m3 s1. This high flow rate prevented particle condensation. Sampling inlets of gas analyzers were located as close as possible to the particulate sampler inlets in order to avoid any bias in the results. 2.3.1. Real-time measurements CO2 and H2O were measured each 2 s by an IR-non dispersive technique. An LI-840 gas analyzer based on a single interchangeable optical path and dual wavelength infrared detection system was used at a flow rate of 1 l min1. Concentration ranges measured by this instrument varied from 0 to 6000 ppm for CO2. CO2 and CH4 concentrations were recorded every 2 min with a photo-acoustic Field Gas Monitor (INNOVA 1412). Measurements are based on the photo-acoustic infrared detection method. The pumping rate for the sampling tubes was 30 cm3 s1 and for the measurement chamber was 5 cm3 s1. The detection limit is in the range of 102 ppm. CO concentrations were measured with a Thomson Environmental System model 300E. Concentrations were recorded every 10 s and quantification is based on the comparison with the absorbance of infrared energy in a reference sample. The inflow to this monitor was 800 cm3 min1.

2.3.2. Off-line measurements 2.3.2.1. Particulate sampling. PM samples were collected on filters by using high volume (flow rate 30 m3 h1) samplers fitted with DIGITEL inlets for the selected fractions. PM2.5 was sampled on MUNTKEL quartz fiber filters 150 mm in diameter. Sampling was carried out in triplicate in the three fires: before the fire (background conditions), during the flame phase and during the smoldering phase. A 1.5 cm2 section of each PM2.5 filter was analyzed by a thermal TOT technique (Birch and Carey, 1996) using a Sunset Laboratory OCEC Analyzer to determine the organic and elemental carbon concentrations (OC and EC, respectively) by using the EUSAAR_2 temperature program (Cavalli et al., 2010). 2.4. Calculations The calculation of gaseous or particulate compound emissions from biomass burning is based on the concept of the emission factor (EF). The EF relates the mass of emitted pollutant (expressed in g) to the amount of consumed fuel (expressed in kg). The EF is expressed by the following equation (1):

  ½x EFx g kg1 ¼ P CT  FC

(1)

where [xn] is the concentration of compound x in the smoke, CT is the sum of C masses from CO2, CO, CH4 and total carbon (TC ¼ OC þ EC). Finally, FC is a factor obtained from the literature with a value of 0.50 and represents grams of biomass consumed for each gram of carbon emitted (carbon-mass balance). C accounts for 50% of total biomass in accordance with the FC factor obtained from measurements taken in a wide range of vegetation types, and it is likely accurate within 10% (Lobert et al., 1991). The emission ratio (ERX/Y) is the excess mixing ratio of species (X) normalized to the excess ratio of standardized species (Y), referred to CO2.

ERX=Y ¼

DX DY

(2)

The Combustion Efficiency (CE) (Eq. (3)) allows us to determine the degree of combustion and to evaluate the relative amount of flaming and smoldering combustion in a fire. Ward and Hardy (1991) defined this parameter as the ratio of carbon emitted as

Fig. 1. Schematic view of steel tower and smoke sampling system.

E. Garcia-Hurtado et al. / Atmospheric Environment 69 (2013) 86e93

CO2 to the total carbon emitted (CT). These authors proposed a simplified index, modified combustion efficiency (MCE) (Eq. (4)), considering only CO2 and CO for CT because both gases are the main contributors to total emissions of carbon compounds. D symbol represents the difference between smoke measurements and background levels before the fire.

CE ¼

DCO2 DCT

MCE ¼

DCO2 ðDCO2 þ DCOÞ

(4)

I ¼ HWR

(5)

where H is the fuel heat of combustion in kJ kg1, W is the weight of fuel consumed per unit area (kg m2) and R is the rate of fire spread in m s1. 3. Results and discussion 3.1. Fuel load and burning conditions Fuel was characterized considering three fractions: standing dead and live fuel, and litter. Senescent Mediterranean gorse shrublands constitute very flammable fuel because of the abundant dead fine fuel fraction (Baeza et al., 2011). In this study, between 46 and 50% of aboveground fuel was dead, and 30e40% of total fuel (32e57 Mg ha1) was litter. The values of vegetation moisture (Table 1) were around 35% and were similar to those of the experimental burnings reported by Baeza et al. (2002) for the same vegetation type. Moisture content of aboveground biomass was rather similar in the three experimental fires. Soil and litter revealed significant differences in moisture content, with moisture decreasing from the first to the third fire. In the case of litter this decrease was probably due to the dry meteorological conditions during the experimental campaign. As for soil differences, these could be attributed to variations in the texture and the aspect of each plot. Table 2 shows maximum temperature values which were similar to previous experimental fires carried out in the same area (De Luis et al., 2004). Maximum temperatures during the flaming phase (up to 710  C) only lasted a few minutes before undergoing a sharp decrease to 100  C (Fig. 2). The values of the consumed aboveground biomass (Table 2) were around 70% in fires 1 and 2, and higher in the case of the third fire. Litter biomass consumption values were lower and unrelated to maximum temperature values at ground level. Combustion

Table 1 Moisture in vegetation during the experimental fires.

1 2 3

Table 2 Main fire characteristics. Fire

1 2 3

Maximum T ( C)

% Fuel consumed Aboveground

Litter

670 675 710

72 78 92

26 27 30

MCE

Fire-line intensity (kW m1)

Biomass consumed (kg m2)

0.76 0.72 0.73

1882 1607 2454

2.97 2.62 2.32

(3)

MCE values higher than 0.90 correspond to the flaming combustion phase and MCE values below 0.85 usually correspond to the pure smoldering combustion phase (Urbanski et al., 2009). Fire-line intensity (I) may be one of the best single indicators of fire behavior (Alexander, 1982). It is defined as the rate of heat energy released per unit time per unit length of fire front using Byram’s equation (1959):

Fire

89

%Moisture  s.d. Aboveground

Litter

Soil

37  13.7 32  13.5 35  12.2

67  49.4 21  8.4 12  4.1

37  8.1 27  11.3 13  4.3

efficiency was similar in the three plots regardless of moisture and temperature differences. This represents a characteristic process of pure smoldering with a short period of flaming, as mentioned by Lobert and Warnatz (1993) for brushy vegetation. Aboveground biomass was clearly the factor that most influenced fire characteristics as is usual in Mediterranean wildfires (Rodríguez-Murillo, 1994). The fire-line intensity ranged between 1607 kW m1 and 2453 kW m1 (Table 2). These values fall within the range of moderate shrubland wildfires (Baeza et al., 2002; Santana et al., 2011). Fire-line intensities above 3500e3700 kW m1 are considered high intensity and difficult to control (Chandler et al., 1983).

3.2. Emissions of carbon compounds Percentages of carbon emitted in the form of CO2, CO, CH4 and TC (OC þ EC) are presented in Table 3. CO2 and CO jointly accounted for more than 95% of carbon emitted. Similar values were found in other Mediterranean and tropical fires (Vicente et al., 2011; Neto et al., 2009). The EFs calculated from the three replicated experiments are shown in Table 4. The maximum value for EFCO2 was obtained in the flaming phase, which yielded MCE values higher than 90%. By contrast, maximum values for EFCO and EFCH4 were produced during the smoldering phase attaining MCE values below 90%, which is in agreement with Crutzen and Andreae (1990). In this study, EFCO2 values were very close to those found in the literature for Mediterranean shrublands in the range from 1000 to 1697 g kg1 (Alves et al., 2011), and close to the average 1200  172 g kg1 obtained by Vicente et al. (2011). The EFCO obtained in this study was higher than those available in the literature with respect to the MCE values (Alves et al., 2010), in agreement with the dominant smoldering phase in our experiment. However, the values logged in the monitor were not the maximum concentration because the monitor was saturated between 5 and 15 min depending on the fire, which corresponds to a 10% in average of the CO data. A good correlation was found between EFCO and EFCO2: EFCO2 ¼ 1.44, EFCO þ1639 (r2 ¼ 0.978), which is in line with the findings of Alves et al. (2011), EFCO2 ¼ 1.79, EFCO þ1755 (r2 ¼ 0.92). Despite the saturated monitor, the good correlations obtained for mean EFs indicate that average CO concentration measurements are reliable. Methane EFs reported from the few existing Mediterranean shrublands studies are about 20 times lower than those of our study (Table 4), for example a range from 1 to 23 g kg1 was obtained in Alves et al. (2010). The relatively high values of CO and CH4 and the low MCE result from the long duration of the smoldering phase. In our study 25% of emitted carbon was in the flaming phase and 75% of C in the smoldering phase. Percentages of emitted carbon in each phase are not available in the literature. The maximum emissions of CO2, CO and CH4 per hectare, calculated as an average of the three plots, were: 43.7 Mg ha1 of CO2, 14.9 Mg ha1 of CO and 1.9 Mg ha1 of CH4. These results were

90

E. Garcia-Hurtado et al. / Atmospheric Environment 69 (2013) 86e93

Fig. 2. Recorded temperatures in fires 1 and 3.

calculated considering the maximum emission factor values and the value of biomass consumed shown in Table 2. PM2.5 reached concentrations of up to 4 mg m3 in the smoke plume with EF of 3.4 g kg1. EFs of particulate matter determined in our study are lower than those reported by Alves et al. (2010) for shrublands, however similar concentration to our study, up to 3 mg m3, is reported by Miranda et al. (2005) also in Portuguese shrubland fires as shown in Table 5. These values are highly variable depending on plume dilution (Reid et al., 2005). The same variations occur for OC and EC emission factors (Battye and Battye, 2002). On average, particulate carbon in PM2.5 was distributed between OC (97%) and EC (3%). As a result, the ratio EC/TC (Table 6) is lower than 0.04 for MCE below 0.90, in accordance with Alves et al. (2011). The OC/PM2.5 and TC/PM2.5 ratios are similar to those of earlier studies regardless of MCE values. Nevertheless, the OC/EC ratio is directly related to MCE values since OC increased Table 3 Percentage in carbon emitted in various biomes. Fire

%C emitted %C emitted %C emitted %C emitted in CO in CH4 in OC þ EC in CO2

Mediterranean shrublanda 71.4 68.0 Portuguese forestsb 75.1 Amazonian forestc a b c

This study (average). Vicente et al. (2011). Neto et al. (2009).

25.7 24.0 20.6

2.6 3.4 3.4

0.3 4.5 1.2

significantly under smoldering conditions (McMeeking et al., 2009). 3.3. Contribution to the improvement of emission inventories The GHG emissions from forest fires are included in the chapter Land Use, Land-Use Change and Forestry, and are incorporated annually into the National Greenhouse Gas Inventories Programme (NGGIP) in relation to the UNFCCC. These reports are compiled by each country, and they follow the Good Practice Guidance for Land Use, Land-Use Change and Forestry, 2003 (GPG-LULUCF) elaborated by the Intergovernmental Panel on Climate Change (IPCC, 2003). GHG in Spain, Italy and France follow the basic requirements, whereas Portuguese inventories go beyond these requirements by employing experimental research data for calculating the consumption of biomass in wildfires. The inventory estimates the emissions of CH4, nitrous oxide (N2O), CO and nitrogen oxides (NOx) from biomass consumed in forest fires. The accuracy of these estimations varies depending on the available information. In order to calculate the mass of carbon and mass of CO2 emitted in wildfires in Spain, the necessary information was obtained from Rodríguez-Murillo (1994). This resulted in a satisfactory level of accuracy (Tier 2). The certainty of the estimations of the emissions of non-CO2 gases is lower, falling to the basic level (Tier 1) because of the lack of information about the N/C ratios and specific EFs.

E. Garcia-Hurtado et al. / Atmospheric Environment 69 (2013) 86e93

91

Table 4 Emission factors (g kg1 dry matter). iFire

1 2 3

EFCO

EFCO2

EFCH4

MCE

Range

Average

Range

Average

Range

Average

Range

Average

[1667e815] [1752e814] [1628e781]

1301 1222 1247

[571e51] [559e65] [594e79]

421 474 463

[97e4] [57e4] [61e3]

35 59 44

[0.95e0.47] [0.99e0.47] [0.93e0.46]

0.76 0.72 0.73

Table 5 Concentration of PM2.5 (mg m3) and emission factors (g kg1 dry matter) in this study and in other studies with different vegetation types. Vegetation

Study

PM2.5

EF PM2.5

EF OC

EF EC

Mediterranean

Average this study

4.89

3.4  1.39

1.4  0.34

0.051  0.028

Shrubland

Alves et al. (2010) Miranda et al. (2005)

[0.6e12.5] Up to 3.0

e e

e e

e e

Portugal forest

Vicente et al. (2011) Alves et al. (2011)

[15.1e45.4] e

37  12 19.3  15.1

21  6.7 [0.16e42]

0.44  0.21 e

Amazonian forest

Neto et al. (2009)

[2.69e8.59]

[3.9e7.5]

e

e

Table 6 Ratios between OC, EC, PM2.5 (%) and MCE in this study and in other studies conducted in the Iberian Peninsula. Vegetation

Study

OC/EC

OC/PM2.5

Mediterranean Shrubland Portugal Forest

Average of this study Alves et al. (2010) Vicente et al. (2011) Alves et al. (2011)

31  10.6 88  46 [34e76] e

44 50 53 50

In the last 15 years, 70% of the emitted carbon in wildfires in Spain came from shrublands whereas around 30% was emitted from woodlands, information available at Ministerio de Agricultura Alimentación y Medio Ambiente (MAGRAMA, 2012) web page. However, the Spanish inventory does not take into account the forest understory or the tree-less shrublands. Forest understory generally accumulates large biomass in Mediterranean pine forests. It is therefore a major contributor to fire emissions. Wildfires in dense forests may affect crowns and/or surface fuels although tree canopies are not always burned, the understory being the part that is most affected. It goes without saying that the inclusion of understory and shrubland fire emissions will improve the EFs. The national inventory of forest fires distinguishes burned shrublands (MAGRAMA, 2012) but does not include understory biomass. In order to include shrubland emissions, a change in the emission ratio (ER), which at present only concerns tree species, is needed. In southern Europe, national inventories use ER values from the recommendations of IPCC in Good Practice Guidance (2006) to calculate non-CO2 compounds. These values are ERCO/CO2 ¼ 0.06 (Lacaux et al., 1995) and ERCH4/CO2 ¼ 0.012 (Delmas, 1994) corresponding to savannah and tropical forest fires, respectively. Both values differ considerably from those obtained in our study (ERCO/ CO2 ¼ 0.36 and ERCH4/CO2 ¼ 0.036). Another aspect that should be discussed is that only non-CO2 compounds should be presented in the inventories because the emissions of CO2 associated with forest fires are assumed to be offset by forest re-growth, and therefore need not be reported. This is a clear oversimplification of the actual C cycle for two reasons: 1) the recovery of the pre-fire C accumulation in the ecosystem is not usually completed in the ecological time scale, i.e. a burned forest could give way to a shrubland with a lower biomass and litter load. For example in the Region of Valencia, only 65% of a pinewood regenerates into a pinewood after a wildfire (Alloza, 2003); 2) despite the recovery of the pre-fire biomass, this process would take many years depending on the vegetation and the

   

7.9 8 5.5 18

TC/PM2.5(%)

MCE

45  7.6 e e 52  20

0.73 0.99 0.60e0.86 0.52e0.97

characteristics of the fire, leading to an increase in CO2 in the atmosphere. For example, in shrublands similar to those of our study, post-fire plant coverage reached values between 40% and 80% of pre-fire conditions 4 years after the wildfire (Baeza and Vallejo, 2008). In terms of biomass or carbon accumulation, values would be much lower and recovery time much longer than assumed in emission inventories. 4. Conclusions Our experimental fires in Mediterranean shrublands produced a moderate-intensity flaming phase followed by a longer smoldering phase. The percentage of total fuel consumed (aboveground þ litter) was around 50% and temperatures reached up to 700  C. The percentage of burned fuel was around 80% for standing fuel, and 28% for litter. Maximum values of CO2 were attained in the flaming phase with MCE values around 0.90. After the flaming phase, moderate temperatures led to incomplete combustion and relatively high emissions of CO and CH4, with MCE values around 0.50. In the three fires, EFs and MCE values were typical of moderate fireline intensity, moderate maximum temperatures, and high fuel moisture. The percentage of carbon species released was 71% for CO2, 26% for CO, 3% for CH4, and 0.3% for total particulate carbon. Approximately 89% of the PM2.5 mass was composed of organic carbon whereas EC values were much lower (3%). Relatively low MCE values account for the high OC/EC ratios for the smoldering phase. The OC and EC ratios are similar to those obtained in earlier studies in Mediterranean wildfires. Our study provides new experimental data on shrubland fire emissions that improve the EFs currently used in inventories for Mediterranean countries. At present, national inventories underestimate actual emissions from wildfires because they do not consider shrub understory and shrublands, and because they

92

E. Garcia-Hurtado et al. / Atmospheric Environment 69 (2013) 86e93

assume that CO2 emitted is offset by forest re-growth. Moreover, more than 70% of sampled carbon in our experiment was measured in the smoldering phase, resulting in significantly higher EFs with respect to those reported for similar MCE values regardless of the fire phase. These different approaches should be further investigated to obtain more realistic emissions from forest/shrubland wildfires. Acknowledgments This study was funded by the FEEDBACKS-PROMETEO Programme (Prometeo 2009/006), and the CONSOLIDER-INGENIO 2010 Programme (GRACCIE, CSD2007-00067). The CEAM Foundation is supported by Generalitat Valenciana (GVA). The fires were conducted with the assistance of the Conselleria de Gobierno-D.G. Prevención, Extinción de Incendios y Emergencias; Consorcio Provincial de Bomberos de la Diputación de Valencia and Brigadas Forestales de IMELSA-Diputación de Valencia. References Acácio, V., Holmgren, M., Rego, F., Moreira, F., Mohren, G.M.J., 2009. Are drought and wildfires turning Mediterranean cork oak forests into persistent shrublands? Agroforestry Systems 76, 389e400. Alexander, M.E., 1982. Calculating and interpreting forest fires intensities. Canadian Journal of Botany 60, 349e357. Alloza, J.A., 2003. Análisis de repoblaciones forestales en la comunidad valenciana desarrollo de criterios y procedimientos de evaluación. Ph.D thesis, Universidad Politècnica de Valencia. Alves, C.A., Gonçalves, C., Pio, C.A., Mirante, F., Caseiro, A., Tarelho, L., Freitasd, M.C., Viegas, D.X., 2010. Smoke emissions from biomass burning in a Mediterranean shrubland. Atmospheric Environment 44, 3024e3033. Alves, C.A., Vicente, A., Nunes, T., Gonçalves, C., Fernandes, A.P., Mirante, F., Tarelho, L., Sánchez de la Campa, A.M., Querol, X., Caseiro, A., Monteiro, M., Evtyugina, M., Pio, C., 2011. Summer 2009 wildfires in Portugal: emission of trace gases and aerosol composition. Atmospheric Environment 45, 641e649. Andreae, M.O., Merlet, P., 2001. Emission of trace gases and aerosols from biomass burning. Global Biogeochemical Cycles 15 (4), 955e966. Baeza, M.J., Vallejo, V.R., 2008. Vegetation recovery after fuel management in Mediterranean shrublands. Applied Vegetation Science 11, 151e158. Baeza, M.J., De Luís, M., Raventós, J., Escarré, A., 2002. Factors influencing fire behaviour in shrublands of different stand ages and the implications for using prescribed burning to reduce wildfire risk. Journal of Environmental Management 65, 199e208. Baeza, M.J., Raventos, J., Escarré, A., Vallejo, V.R., 2006. Fire risk and vegetation structural dynamics in Mediterranean shrubland. Plant Ecology 187, 189e201. Baeza, M.J., Valdecantos, A., Alloza, J.A., Vallejo, V.R., 2007. Human disturbance and environmental factors as drivers of long-term post-fire regeneration patterns in Mediterranean forests. Journal of Vegetation Science 18, 243e252. Baeza, M.J., Santana, V.M., Pausas, J.G., Vallejo, V.R., 2011. Successional trends in standing dead biomass in Mediterranean basin species. Journal of Vegetation Science 22, 467e474. Barboni, T., Cannac, M., Pasqualini, V., Simeoni, A., Leoni, E., Chiaramonti, N., 2010. Volatile and semivolatile organic compounds in smoke exposure of firefighters during prescribed burning in the Mediterranean region. International Journal of Wildland Fire 19, 606e612. Battye, W., Battye, R., 2002. Development of Emissions Inventory Methods for Wildland Fire. U.S. Environmental Protection Agency, Chapel Hill, NC, 91 pp. Birch, M.E., Carey, R.A., 1996. Elemental carbon-based method for monitoring occupational exposures to particulate diesel exhaust. Aerosol Science & Technology 25, 221e241. 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, 481e484. http://dx.doi.org/10.1126/science.1163886. Byram, G.M., 1959. Combustion forest fuel. In: Davis, K.P. (Ed.), Forest Fire: Control and Use. McGraw-Hill, New York, pp. 61e89. Cavalli, F., Viana, M., Yttri, K.E., Genberg, J., Putaud, J.-P., 2010. Toward a standardized thermaleoptical protocol for measuring atmospheric organic and elemental carbon: the EUSAAR protocol. Atmospheric Measurement Techniques 3, 79e89. Chandler, C., Cheney, P., Thomas, P., Trabaud, L., Williams, D., 1983. Fire in forestry. In: Forest Fire Behaviour and Effects, vol. I. John Wiley and Sons, New York, 450 pp. Ciccioli, P., Brancaleoni, E., Frattoni, M., Cecinato, A., Pinciarelli, L., 2001. Determination of volatile organic compounds (VOC) emitted from biomass burning of Mediterranean vegetation species by GCeMS. Analytical Letters 34 (6), 937e955.

Crutzen, P.J., Andreae, M.O., 1990. Biomass burning in the tropics: impact on atmospheric chemistry and biogeochemical cycles. Science 250 (4988), 1669e 1678. DCCV, 2010. Direziones Centrale per le Indagini su Condizione e Qualità della Vita Statistiche ambientali. Available at: http://www.ftsnet.it/documenti/820/ Pressioniincendi.pdf. Delmas, R., 1994. An overview of present knowledge on methane emissions from biomass burning. Fertilizer Research 37, 181e190. De Luis, M., Baeza, M.J., Raventós, J., González-Hidalgo, J.C., 2004. Fuel characteristics and fire behaviour in mature Mediterranean gorse shrublands. International Journal of Wildland Fire 13, 79e87. Hardy, C.C., Conard, S.G., Regelbrugge, J.C., Teesdale, D.R., 1996. Smoke Emissions from Prescribed Burning of Southern California Chaparral. Research Paper PNW-RP-486. U.S. Department of Agriculture, Forest Service, Pacific Northwest Forest and Range Experiment Station, Portland. Hays, M.D., Fine, P.M., Geron, C.D., Kleeman, M.J., Gullett, B.K., 2005. Open burning of agricultural biomass: physical and chemical proprieties of particle-phase emissions. Atmospheric Environment 39, 6747e6764. Houghton, R.A., Hackler, J.L., Lawrence, K.T., 1999. The U.S. carbon budget contributions from land-use change. Science 285, 574e578. IPCC, 2003. Good Practice Guidance for Land Use, Land-use Change and Forestry, (GPG-LULUCF). Intergovernmental Panel on Climate Change. Available at: http://www.ipcc-nggip.iges.or.jp/public/gpglulucf/gpglulucf_files/GPG_ LULUCF_FULL.pdf. IPCC, 2007. Climate change 2007. Climate change impacts, adaptation and vulnerability. In: Working Group II. Contribution to the Intergovernmental Panel on Climate Change Fourth. Cambridge University Press, Cambridge, UK. Assessment Report. JRC (Join Research Centre), 2012. Institute for Environment and Sustainability. Forest Fire Information System. Available at: http://forest.jrc.ec.europa.eu/effis/ reports/annual-fire-reports/. Lacaux, J.P., Brustet, J.M., Delmas, R., Menaut, J.C., Abbadie, L., Bonseang, B., Cachier, H., Baudet, J., Andreae, M.O., Helas, G., 1995. Biomass burning in the tropical savannas of Ivory Coast: an overview of the field experiment fire of savannas (FOS/DECAFE 91). Journal of Atmospheric Chemistry 22, 195e216. Leumieux, J.P., Christopher, C.L., Santoianni, D.A., 2004. Emissions of organic air toxics from open burning: a comprehensive review. Progress in Energy and Combustion Science 30, 1e32. Lobert, J.M., Warnatz, J., 1993. Emissions from the combustion process in vegetation. In: Crutzen, P., Goldammer, J. (Eds.), The Ecological, Atmospheric and Climatic Importance of Vegetation Fires. Wiley, Chichester, pp. 15e37. Lobert, J.M., Scharffe, D.H., Hao, W.M., Kuhlbusch, T.A., Seuwen, R., Warneck, P., Crutzen, P.J., 1991. Experimental evaluation of biomass burning emission: nitrogen and carbon containing compounds. In: Levine, J.S. (Ed.), Global Biomass Burning: Atmospheric, Climatic and Biospheric Implications. MIT Press, Cambridge, pp. 289e304. MAGRAMA (Ministerio de Agricultura Alimentación y Medio Ambiente of Spain), 2012. Available at: http://www.magrama.gob.es/es/biodiversidad/temas/ defensa-contra-incendios-forestales/estadisticas-de-incendios-forestales/ default.aspx. Martins, P.C., Valente, J., Papoila, A.L., Caires, I., Araújo-Martins, J., Mata, P., Lopes, M., Torres, S., Rosado-Pinto, J., Borrego, C., Annesi-Maesano, I., Neuparth, N., 2012. Airways changes related to air pollution exposure in wheeling children. The European Respiratory Journal 39 (2), 246e253. McMeeking, G.R., Kreidenweis, S.M., Baker, S., Carrico, C.M., Chow, J.C., Collett, J.L., Hao, W.M., Holden, A.M., Kirchstetter, T.W., Malm, W.C., Moosmüller, H., Sullivan, A.P., Wold, C.E., 2009. Emissions of trace gases and aerosols during the open combustion of biomass in the laboratory. Journal of Geophysical Research 114, D19210. http://dx.doi.org/10.1029/2009JD011836. Miranda, A.I., Ferreira, J., Valente, J., Santos, P., Amorim, J.H., Borrego, C., 2005. Smoke measurements during Gestosa-2002 experimental field fires. International Journal of Wildland Fire 14, 107e116. Miranda, A.I., Marchi, E., Ferretti, M., Millán, M.M., 2009. Forest fires and air quality issues in southern Europe. In: Bytnerowicz, A., Arbaugh, M., Riebau, A., Andersen, C. (Eds.), Developments in Environmental Science, vol. 8. Elsevier B.V (Chapter 9). Miranda, A.I., Martins, V., Cascão, P., Amorim, J.H., Valente, J., Tavares, R., Borrego, C., Tchepel, O., Ferreira, A.J., Cordeiro, C.R., Viegas, D.X., Ribeiro, L.M., Pita, L.P., 2010. Monitoring of firefighters exposure to smoke during fire experiments in Portugal. Environment International 36 (6), 736e745. Neto Jr., T.G.S., Carvalho, J.A., Veras, C.A.G., Alvarado, E.C., Gielow, R., Lincoln, E.N., Christian, T.J., Yokelson, R.J., Santos, J.C., 2009. Biomass consumption and CO2, CO and main hydrocarbon gas emissions in Amazonian forest clearing fire. Atmospheric Environment 43, 438e446. Pausas, J.G., Llovet, J., Rodrigo, A., Vallejo, V.R., 2008. Are wildfires a disaster in the Mediterranean basin? A review. International Journal of Wildland Fire 17, 713e723. Promethee, 2012. La banque de données sur les incedies de forêts en région Mediterraéenne en France. Conservatoire de la Forêt Mediterraéenne. Conseil General. Available at: http://www.promethee.com/prom/basedoc/liste.do. Reid, J.S., Koppmann, R., Eck, T.K., Eleuterio, D.P., 2005. A review of biomass burning emissions part II: intensive physical properties of biomass burning particles. Atmospheric Chemistry and Physics 5, 799e825. Rodríguez-Murillo, J.C., 1994. The carbon budget of the Spanish forest. Biogeochemistry 25, 197e217.

E. Garcia-Hurtado et al. / Atmospheric Environment 69 (2013) 86e93 Santana, V.M., Baeza, M.J., Vallejo, V.R., 2011. Fuel structural traits modulating soil temperatures in different species patches of Mediterranean Basin shrublands. International Journal of Wildland Fire 20, 668e677. Urbanski, S.P., Hao, W.M., Barker, S., 2009. Chemical composition of wildland fire emissions. In: Bytnerowicz, A., Arbaugh, M., Riebau, A., Andersen, C. (Eds.), Developments in Environmental Science, vol. 8. Elsevier B.V (Chapter 4). Vallejo, V.R., Alloza, J.A., 1998. The restoration of burned lands: the case of eastern Spain. In: Moreno, J.M. (Ed.), Large Forest Fires. Backhuys Publishers, Leiden, pp. 91e108.

93

Vicente, A., Alves, C., Monteiro, C., Nunes, T., Mirante, F., Evtyugina, M., Cerqueira, M., Pio, C., 2011. Measurement of trace gas and organic compounds in the smoke plume from a wildfire in Penedono (central Portugal). Atmospheric Environment 45, 5172e5182. Ward, D.E., Hardy, C.C., 1991. Smoke emissions from wildland fires. Environment International 17, 117e134. Weise, D.R., Ward, D.E., Paysen, T.E., Koonce, A.L., 1991. Burning California Chaparral. An exploratory study of some common shrubs and their combustion characteristics. International Journal of Wildland Fire 1 (3), 153e158.