ÓPTICA PURA Y APLICADA – Vol. 37, núm. 3, - 2004
Fires in Central Russia 2002 and their effects on optical properties of
atmosphere and solar irradiance in different spectral regions
N.Ye. Chubarova
(1), A.N. Rublev
(2), B.N. Holben
(3)1.
Moscow State University, Faculty of Geography, 119899, Moscow, Russia,
e-mail:
[email protected]
2.
Russian Research Center “KURCHATOV INSTITUTE”, Kurchatov sq., 123182 Moscow
3.
Laboratory for Terrestrial Physics,NASA Goddard Space Flight Center,
Greenbelt, Maryland
REFERENCES AND LINKS.
[1]. B. N.Holben, T.F.Eck, I.Slutsker, D.Tanre, J.P.Buis, A.Setzer, E.Vermote, J.A.Reagan, Y.Kaufman, T.Nakajima, F.Lavenu, I.Jankowiak, and A.Smirnov, “AERONET - A federated instrument network and data archive for aerosol characterization”, Rem. Sens. Environ., 66, 1-16 (1998)
[2]. N. Chubarova and Ye.I. Nezval', “Thirty year variability of UV irradiance in Moscow”, J. Geophys. Res., Atmospheres,
105, 12529-12539 (2000).
[3]. N. Chubarova, “Monitoring of biologicaly effective UV radiation in Moscow region”,. Izvestiya, Atmos. and Oceanic Phys.,38, No 3, 312-322 (2002).
[4]. Uliumdzhieva N., N. Chubarova, B.Holben “Aerosol optical properties during forest fires 2002 in Moscow region”, “Meteorology and Hydrology” 2005, in print
[5]. O. Dubovik, B.N.Holben, T.F.Eck, A.Smirnov, Y.J.Kaufman, M.D.King, D.Tanre, and I.Slutsker, “Variability of absorption and optical properties of key aerosol types observed in worldwide locations”, J. of Atm. Sciences, 59, 590-608 (2002).
[6]. O.Dubovik, King M. D. “A flexible inversion algorithm for retrieval of aerosol optical properties from Sun and sky radiance measurements”, J. of Geophys. Res.,105, D16, 20673-20696 (2000).
[7]. N. Chubarova, “Influence of Aerosol and Atmospheric Gases on Ultraviolet Radiation in Different Optical Conditions Including Smoky Mist of 2002”, Doklady Earth Sciences, 394, No. 1, 62–67. (2004)
[8]. N. N. Uliumdzhieva, N. Ye. Chubarova, A. N. Smirnov, “Aerosol characteristics of the atmosphere in Moscow based on Cimel sun photometer data”, “Meteorology and Hydrology” 2005, in print
[9]. R. Martin, personal communication
[10]. N. Chubarova and O. Dubovik, “The sensitivity of aerosol properties retrievals from AERONET measurements to NO2 concentration over industrial region on the example of Moscow”, this issue.
[11]. N. Chubarova N. A. Rublev, A. Trotsenko, V. Trembach, “The comparisons between modeled and measured surface shortwave irradiances in clear sky conditions”, Izvestiya, Atmospheric and Oceanic Physics, 2, 201-216 (1999) [12]. http://www-litms.imp.kiae.ru/index7.htm
1.
Introduction.
Summer-fall period 2002 in Central Russia was characterized by severe fire smoke conditions that led to high concentrations of atmospheric aerosols
and gaseous species. During the smoke episode, a complex program of aerosol, gas species, and broadband solar radiation measurements was in operation at Meteorological Observatory of Moscow State University (MOMSU) and at its site in
ABSTRACT:
July through September 2002 in Central Russia was characterized by severe fire smoke conditions that led to high concentrations of atmospheric aerosols and gaseous species, and strong attenuation of solar irradiance. The properties of smoke aerosol differ from typical Moscow aerosol characteristics, but correspond to those observed in boreal forest fires of North America. The impact of aerosol and several gaseous components on solar irradiance has been revealed due to model and experimental data. In visible spectral range there is a good agreement between model and observed irradiance, but in UV region model fluxes are substantially higher. The solar absorption due to fire smoke aerosol comprises about 30%, the absorption due to NO2 reaches
2-4%. Maximum radiative forcing at ground in the fire conditions increased up to 305 W/m2.
Moscow suburb (50km to the west) at Moscow State University Zvenigorod biostation (ZB MSU).
In this study we examine several aspects of the observed fires: the changes in aerosol properties and gas species composition of the atmosphere, spectral features of radiation attenuation on the base of experimental and model studies. In addition, using Monte-Carlo calculations we show the changes in radiative forcing and radiative forcing efficiency at the top and at the bottom of the atmosphere as well as in the solar absorption within the atmosphere due to aerosol and gas species content.
2.
Results.
2.a. Description of instrumentation
In addition to standard meteorological measurements and automatic weather stations installed at both sites, we use a complex program of aerosol measurements (Cimel sun photometer of AERONET network1 and hand-held Hasemeters) as
well as global solar irradiance observations in several spectral ranges. The measurements of solar irradiance in shortwave UV region (Qer) were run by the instrument UVB-1 (YES, Inc), which has the spectral sensitivity curve close to the erythemal action curve with the maximum in the region of 300-310 nm. In addition, there were the observations of UV irradiance in the range of 300-380nm (Q380) by MO MSU UV-meter with the effective wavelength in the longwave UV range 340-350nm. The description of UV instruments and the methods of measurements are given in Chubarova and Nezval’2
and Chubarova3. The measurements of visible
irradiance (400-700nm) were run by LI-COR PAR sensor and integral irradiance (340-4,000nm)– by Russian pyranometer M-80. The measurements of gaseous composition of atmosphere have been also used. They were obtained at MO MSU and at the ecological laboratory of MO MSU with the participation of A.M. Obukhov’s Institute of Atmospheric Physics as well as of “Mosecomonitoring” State Environmental Agency. We analyzed the atmospheric gases, which play vital role in the absorption of visible and UV irradiance: nitrogen dioxide (NO2), formaldehyde (HCHO),
sulphuric dioxide (SO2) and surface ozone (O3). The
capture of clear sky periods, which were necessary for the analysis of gaseous and aerosol attenuation, was made using the hourly visual observation of sky conditions.
2.b. Meteorological conditions and characteristics of fire intensity.
The specific meteorological regime with the prevalence of anticyclonic weather over the Central European plain created favorable conditions for forest and peatbog fires expansion over the vast territory. According to the data of Avialesookhrana (AirForest Protection Agency) the distinct growth of fires began in the second part of July when the
increase in an area occupied by fires reached 53 hectares per day (Fig.1). At that time the main fires were observed at the eastern part of Moscow region where large area is covered by peatbogs, which had been drained up in the 30s. In the first half of August a 50% decrease in number of fires was observed due to rainfalls associated with the atmospheric fronts which crossed Moscow region. In Moscow precipitation, P, comprised 35mm during August 2 - August 3, and 12,4mm - on August 8-9. Later in the second part of August and September there was a second peak of fires when the increase of more than 255 fires per day was observed only in Moscow region and the area of their expansion reached 320 hectares per day. During this period the fires has also been intensified in the western and northern regions of the Central European plain. The significant decrease in fire activity was observed only after the middle of September when the Central European plain was under the influence of an intense cyclone (P=68mm on September 17- September 18). It is worth mentioning that the deep peatbog fires continued up to the spring 2003 but their effect on the atmosphere was negligible. Figure 1 shows the intensity of fires over Moscow region as well as the classes of fire indices. Fire indices are calculated according to Nesterov’s equation: G=∑(T*d), which takes into account the day-to-day accumulation of the joint effect of maximum temperature at midday (T) and dew point depression (d=T-Td) in the absence of precipitation higher than 3mm. Fire classes K vary within 0 - 5 so that K=3, K=4, K=5 relate correspondingly to dangerous, high dangerous and extremely dangerous fire conditions.
Figure1. Classes of Nesterov’s flammability indices and the parameters of fires in Moscow region.
In typical weather conditions during June-September period the K values reach 3 in about 50% of cases, while the probability of K=4 is less than 2% and it is practically zero for K=5. In 2002 the occurence of K=4 was about 30%, and it was about 10% for K=5. Furthermore, at MO MSU, due to the intensifying of precipitation in conditions of large city the flammability indices were less than those observed at Zvenigorod. During the June-September
0 50 100 150 200 250 300 350
30.06
.02
10.07
.02
19.07
.02
27.07
.02
05.08
.02
14.08
.02
25.08
.02
01.09
.02
08.09
.02
15.09
.02
22.09
.02
29.09
.02
In
cr
e
a
se
in
sq
ua
re
o
f f
ir
e
s (
S
)
a
nd
th
e
ir
num
be
r (
N
) pe
r da
y
0 1 2 3 4 5 6
K
la
ss o
f N
e
st
e
ro
v'
s
fla
m
m
a
bi
lit
y i
ndi
ce
s,
K
period, the number of days with K≥4 comprised 48 and 98 respectively at MO MSU and at ZB MSU.
2.c. Optical properties and radiative effects of fire smoke atmospheric conditions
Smoke from forest fires leads to significant changes in aerosol and gaseous concentrations in the lower atmosphere. Figure 2 shows an agreement between aerosol optical thickness and the peaks in forest fire numbers increase. We observed extremely high monthly mean aerosol optical thickness (AOT500=0.4-0.84) that is 2-4 times greater than the mean values obtained for typical conditions (AOT500=0.21)4. It is necessary to emphasize that
the temporal changes in aerosol optical thickness at MO MSU and Zvenigorod correlate well (r=0.91 for time synchronized dataset, see also Fig.2), indicating the expansion of the homogeneous smoke cloud over a large distance. High aerosol optical thickness observed in Moscow region was of the same range as those values measured by Cimel instruments in other forest fire conditions: boreal forest fires in Canada, etc5. Volume aerosol size distribution
retrieved by the method described in6 was
characterized by a distinct bimodal character (see Fig.3) with an increase in fine aerosol mode compared with typical Moscow conditions (61% and 56%, respectively).
Figure 2. Aerosol optical thickness (AOT500) obtained by Cimel and Hasemeter sun photometers and fire numbers
increase per day (N) in Moscow region. 2002.
Figure 3. Aerosol size distribution in clear conditions (lines) and during fires (dotted lines).
Figure 4 a,b shows monthly mean values of real and imaginary parts of aerosol refractive index
during typical and so called “fire” months (July, August, September). It is clearly seen that during fire conditions the real part of refractive index tends to be higher and imaginary part – to be lower than those in typical situations. This leads to higher values of aerosol single scattering albedo (SSA) during fire events that can be seen from Table I. These high values of SSA correlate well with the results obtained in other regions (fires in North America, Brasilian fires) where the smouldering phase prevailed over the open fire phase as it was shown in5.
A/
B/
Figure 4a,b. Real (a) and imaginary (b) part of refractive index in typical and fire conditions in Moscow.
TABLE I.
Mean* optical characteristics of aerosol at 440nm for months with fire and with typical conditions, 2001-2002.
Moscow
Typical conditions Fire conditions Aerosol optical
thickness
0.16 0.58
Single scattering albedo
0.90 0.94
Factor asymmetry 0.68 0.68 real_part of
refractive index
1.36 1.44
Imaginary part of refractive index
0.01 0.007
*. In this Table mean values were obtained as the average from monthly means not taking into account for the difference in case numbers. An additional cloud filter described in8 has been also
used for evaluation of monthly mean values.
0 0.5 1 1.5 2 2.5 3 3.5
29/6 9/7 19/7 29/7 8/8 18/8 28/8 7/9 17/9 27/9
τ5 0 0
0 50 100 150 200 250 300 Ν Zv enigo rod Bio sta tio n
(HAS EM ET ER)
M os co w M OM SU( CIM EL )
case numbe r o f the fires (N)
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
0.01 0.1 RADII,MKM1 10 100
d
V
/d
ln
r
SMALL_FIRE_08.07.02 SMALL_FIRE_13.07.02 FIRE-30.07.02 FIRE_31.07.02 FIRE_1.08.02 FIRE_02.09.02 FIRE_05.09.02 FIRE_07.09.02 CLEAR_02.07.02 CLEAR_21.08.02 CLEAR_12.09.02
1.30 1.35 1.40 1.45 1.50
400 500 600 700 800 900 1000 1100
nm
real part of refractive index, n
IX - 01 III - 02
IV - 02 V - 02
VI - 02 VII - 02
VIII - 02 IX - 02
0.000 0.002 0.004 0.006 0.008 0.010 0.012 0.014 0.016
400 500 600 700 800 900 1000 1100
nm
imaginary part of refractive index
III - 02 IV - 02
V - 02 VI - 02
VII - 02 VIII - 02
High concentration of nitrogen dioxide, formaldehyde and sulphuric dioxide were also observed in days with high aerosol loading due to advection of “fire cloud” over Moscow. The correlation of aerosol optical thickness with the NO2
and HCHO concentrations exceeds 0.5 while the correlation with SO2 is much less (0.37), indicating
the existence of another anthropogenic sources of this gas. There was a distinct correlation between AOT500 and daily maxima of surface ozone concentration, which is described in7. Evidently, the
distinct correlation between the concentration of optically active gases and aerosol causes an additional attenuation of solar irradiance during the fires and leads to additional climate effects.
The mean and extreme values of gaseous concentration for cloud-free conditions during warm period 2002 are shown in Table II. It is necessary to mention that the maximum value of gas concentration (except for SO2) was 1.5-3 times
higher than maximum allowable concentrations, and during the “smoke” period even the mean concentration of NO2 was close to maximum
allowable concentration.
Table II.
The concentration of gaseous species in cloud-free conditions in Moscow. May-September 2002. In the heading of the columns the sample maximum allowable concentrations (MAC) are shown according to the Russian
standard.
Concentration of gaseous species NO2,
µg/m3
(MAC= 85
µg/m3)
O3,
µg/m3
(MAC= 160
µg/m3)
SO2, µg/m3
(MAC=500
µg/m3)
HCHO,
µg/m3
(MAC=35
µg/m3)
Mean 75 106 6 16
Min 15 42 1 1
Max 251 252 96 91
Solar irradiance has been also changed significantly in conditions of forest fires. To facilitate the analysis, all radiation measurements were adjusted to the solar elevation of h=40° and the
Qer values were also adjusted to the total ozone content of 350 matm.cm. Then the data were normalized to the measurement at relatively clear day 1.06.02, when the aerosol content was low (AOT500=0.1). Figure 5 shows the losses of solar irradiance in different spectral ranges for cloud-free conditions normalized in this way during the whole warm period 2002. For typical smoke-free optical conditions the attenuation of solar irradiance varied within 20% due to the changes in aerosol and gaseous composition of the atmosphere. At the end of July 2002 and in the first decade of September we can see a significant loss of solar irradiance and, especially, UV irradiance (up to 70% for UV and up to 45% for visible and total solar irradiance). We
assume that the strong attenuation can be caused by joint effect of gas and aerosol attenuation. Using fire smoke atmospheric properties as input parameters to model simulations, we studied the impact of aerosol and several gaseous components (O3, SO2, NO2, and
HCHO) on attenuation of solar irradiance in different spectral ranges and compared the results of our calculations with experimental radiative measurements. It should be noted that to calculate the total content of certain gas species we use direct measurements of ground gas concentration and a model profile from GEOS-CHEM 3-D model9 for
Moscow conditions (see the details in) 10.
The sensitivity of solar fluxes to the account of gas concentration is shown in Table III for UV and visible spectral ranges. Even for Qer the effects of HCHO and SO2 plays not very important role, while
the account of NO2 and surface ozone concentration
in the low troposphere lead to 1-2% changes in Qer
in the urban atmosphere. For longwave UV irradiance the effect of NO2 is more pronounced.
During fire events due to higher gas concentration and higher aerosol content the significant attenuation can be observed: up to 10-15% in UV and up to 5-6% in visible region. The comparison between measurements and calculations has revealed that application of real gas species concentration in the model has led to the improvement of an agreement between model and experimental data. However, for shortwave UV region we still have some overestimation of modele values possibly due to non accounting for aerosol properties directly in UV spectral region that was described in details in7.
Figure 5 illustrates the better agreement between measurements and calculations in visible spectral region with consideration of NO2.
TABLE III
Changes in solar irradiance in different spectral ranges due to account for concentration of different gases observed in Moscow in 2002 relative to the values calculated using
concentration given in standard models (WMO 1986, Lowtran7) in typical and fire conditions*.
Typical conditions,%
Fire conditions,% Erythemally-active irradiance, Qer Surface O3 Up to 1 Up to 3
SO2 <0.2 Up to
1.5
NO2 2 4-6 up
to 10 HCHO <0.2 Up to
0.5 UV 300-380nm, Q380
NO2 2-3 6-8 (up
to 15) Visible 400-700nm, Qvis
NO2 1-2 (2-4)
up to 5-6
Figure 4. Solar irradiance loss in different spectral regions in cloud-free conditions for the warm period 2002.
Figure 5. The ratio between calculated and measured visible irradiance (400-700nm) as a function of aerosol
optical thickness without and with account of NO2.
2.d. Radiative forcing and atmospheric absorption during fire event.
Changes in radiative characteristics due to changes in aerosol and gas composition can lead to significant pertubation of radiative budget and, hence, to the changes in climate system on the whole. Every year territories occupied by different types of fires comprise thousands of square kilometers. Therefore it is very important to evaluate the consequences of this event on radiative budget. For this purpose we use the Monte-Carlo radiative transfer model described in11 which has been
accurately verified against different radiative measurements in various optical regimes. In this study we evaluate the atmospheric absorption in typical and fire conditions, and make some assessments of radiative forcing at the top of the atmosphere and at the ground.
In this study we use the term “radiative forcing” taking into consideration also the effects of gas species absorption. In this case radiative forcing (R) is calculated as
R= Qnet(gas,aerosol) –Qnet(no gas,no aerosol) (1)
In addition, we use radiative forcing efficiency characteristic, which is calculated as:
Reff=R/dAOT_500 (2)
Therefore, Reff characterizes relative changes in solar irradiance without and with gas and aerosol content for dAOT_500=1.
We chose the cases which described most typical situations, and when we were able to recalculate the optical properties of aerosol taking into account for the gas absorption10. Using these characteristics we show the radiative effect of gas and aerosol changes during fire events.
Figure 6 shows the results of calculation of radiative forcing efficiency at ground as well as mean and maximum radiative forcing in typical (clear) and fire conditions. Due to higher SSA values the Reff at ground during fires is less: -88±8 W/m2
against −155±8 W/m2 in typical conditions. At the
same time maximum radiative forcing during fires is very large and reaches -258 W/m2 for the extreme
aerosol loading (AOT_500=2.94) observed in clear sky conditions. This value is twice higher than that obtained in typical conditions with the observed extreme AOT_500=0.98. At the same time mean
radiative forcing over Moscow is not very large in typical conditions (–25 W/m2)compared with that
observed during fires (–51 W/m2).
Figure 6. Mean and maximum radiative forcing and radiative forcing efficiency at ground in different conditions. CosZ≈0.6. Mean R values were calculated
according to AOT500 in Table I.
Radiative forcing at the top of the atmosphere is demonstrated in Figure 7. Aerosol optical thickness higher than 0.5 corresponds to the fire conditions. This characteristic is also a function of aerosol optical thickness and is characterized by significant negative values (up to –80 W/m2)
compared with –10- –20 W/m2 in typical conditions.
Figure 7. Radiative forcing at the top of the atmosphere. cosZ≈0.6.
Absorption in the atmosphere is calculated according to the following equation:
-30% -20% -10% 0% 10% 20% 30%
0 0.5 1 1.5 2 2.5 3
AOT_500nm ch an g es in visib le rad iat io n
with NO2, mean=0%
without NO2, mean=2%
-100 -80 -60 -40 -20 0
0 0.5 1 1.5 2 2.5
AOT_500 R at t op, W /m 2 -80% -70% -60% -50% -40% -30% -20% -10% 0% 10% 20% 02 .0 5. 02 07 .0 5. 02 12 .0 5. 02 17 .0 5. 02 22 .0 5. 02 27 .0 5. 02 01 .0 6. 02 06 .0 6. 02 11 .0 6. 02 16 .0 6. 02 21 .0 6. 02 26 .0 6. 02 01 .0 7. 02 06 .0 7. 02 11 .0 7. 02 16 .0 7. 02 21 .0 7. 02 26 .0 7. 02 31 .0 7. 02 05 .0 8. 02 10 .0 8. 02 15 .0 8. 02 20 .0 8. 02 25 .0 8. 02 30 .0 8. 02 04 .0 9. 02 09 .0 9. 02 ir ra di an ce lo ss
Qer Q_380 visible Qir
-139 -25 -258 -51 -88 -155
-300 -250 -200 -150 -100 -50 0
W/m2 R_aver_max_clear
Qa=S0 cosθ-F↑λ-∫F↓λ(1-Aλ)dλ (3)
where S0 =1365 W/m2, θ- zenith angle, F↑λand F↓λ -upward and downward fluxes, Aλ-surface albedo.
Figure 8 illustrates the changes in Qa as a
function of aerosol optical thickness. We also show several variants of calculations taking into account for gaseous absorption and making calculations without this correction. The recent results of our calculations in clear (typical) conditions are close (but less) than that obtained for continental aerosol model12 due to higher SSA in these particular cases
and difference in water vapor. On the average the solar absorption due to both aerosol and NO2 during
the fire conditions comprises 27-32%. The absorption of NO2 can reach 4%, corresponding
mainly to the fire conditions with high concentration of this gas species while for typical conditions it comprises about 1-2% for Moscow urban area.
On the whole, there is a distinct redistribution of radiation budget in fire conditions with increase in absorption from 74 W/m2 to 103 W/m2 , decrease
in downward (188 W/m2 compared with 243 W/m2 )
and increase in upward irradiance (96 W/m2
compared with 80W/m2 ).
Figure 8. The solar absorption in the atmosphere in different conditions, Monte-Carlo simulations for several typical cases in clear and fire conditions. The calculations
for continental model atmospheric parameters were fulfilled via12. cosZ≈0.6
Conclusions.
The conditions with severe fires over Central Russia observed in 2002 led to significant changes in aerosol and gas composition of the atmosphere, to the losses of total and visible solar radiation (up to 45%) and, especially, of UV radiation (up to 70%).
The analysis of solar flux attenuation in visible spectral region has revealed a good agreement between model and experimental data that has indirectly approved that smoke aerosol has a very slight absorption (SSA>0.95) in visible region but possibly is larger in the UV range.
The effects of gaseous absorption (mainly by NO2) should be taken into account in highly polluted
conditions and during intensive fires.
There is a distinct redistribution of radiation budget in fire conditions with increase in absorption from 74 W/m2 to 103 W/m2 , decrease in downward
(188 W/m2 compared with 243 W/m2) and increase
in upward irradiance (96 W/m2 compared with 80
W/m2). The radiative forcing efficiency at ground
during fires is less than that in typical Moscow conditions: -88±8 W/m2 against -155±8 W/m2 but
maximum radiative forcing during fires has reached
−258 W/m2.
Acknowledgements.
The authors would like to thank Moscow Ecological Monitoring Agency and Institute of Atmospheric Physics for providing the data on concentration of some gas species;
The “Avialosoohrana” Agency for providing the data on fire characteristics.
Dr. Wei Min Hao from USDA Forest Service for providing the Hasemeters for aerosol studies at Moscow suburbs.
The staff of Meteorological observatory of MSU and, personally, V. Rozental’ for the technical help and N. Uliumdzhieva, A. Yurova, Ye. Stolyarova.
This work was partially fulfilled under the support of the USDA Forest Service in the frame of the project ‘‘Solar Radiation and Weather Variability Influences on Russian Sub-Boreal Forest Phenology.’’
0% 5% 10% 15% 20% 25% 30% 35%
0 1 2 3
AOT_500
absorpt
ion,
% NO2+aerosol
no NO2 + aerosol
continental_model
