Download Regional climate change and its impact on photooxidant

Click
Here
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 111, D12302, doi:10.1029/2005JD006748, 2006
for
Full
Article
Regional climate change and its impact on
photooxidant concentrations in southern Germany:
Simulations with a coupled regional
climate-chemistry model
Renate Forkel1 and Richard Knoche1
Received 7 October 2005; revised 30 January 2006; accepted 3 March 2006; published 17 June 2006.
[1] In order to investigate possible effects of global climate change on the near-surface
concentrations of photochemical compounds in southern Germany, nested regional
simulations with a coupled climate-chemistry model were carried out. The simulations
with a horizontal resolution of 60 km for Europe and 20 km for central Europe were driven
by meteorological boundary conditions provided by a long-term simulation of the global
climate model ECHAM4. Two time slices of about 10 years were compared, one
representing the 1990s and one representing the 2030s. For the region of southern
Germany the simulations show an increase of the mean summer temperature by almost
2 along with a decrease of cloud water and ice and a corresponding increase of the
photolysis frequencies and the emissions of biogenic hydrocarbons. Under the model
assumption of unchanged anthropogenic emissions this leads to an increase of the mean
mixing ratios of most photooxidants. Because of the complex topography and the
heterogeneous distribution of precursor emissions all parameters show pronounced
regional patterns. The average daily maximum ozone concentrations in southern Germany
increase for the considered scenario by nearly 10% in the summer months. Depending on
the region, the increase of the mean daily maximum ranges between 2 and 6 ppb. As a
consequence, the number of days when the 8-hour mean of the ozone concentration
exceeds the threshold value of 120 mg m3 increases by 5 to 12 days per year.
Citation: Forkel, R., and R. Knoche (2006), Regional climate change and its impact on photooxidant concentrations in southern
Germany: Simulations with a coupled regional climate-chemistry model, J. Geophys. Res., 111, D12302,
doi:10.1029/2005JD006748.
1. Introduction
[2] The change of global climate due to anthropogenic
emissions of greenhouse gases is expected to result in
higher summer temperatures and less precipitation in Germany. Though the results of global climate simulations
show some differences depending on the assumed greenhouse gas emission scenarios and the global climate model
used, all modeling studies indicate a similar general trend
[Hulme and Sheard, 1999].
[3] Climate change can increase the formation of tropospheric ozone and other photooxidants, which are suspected
to have harmful effects on human health, agricultural crops,
and natural vegetation. Decreased cloudiness will result in
an increase of UV radiation and subsequently in enhanced
photochemical smog production. This effect may still be
enhanced as higher summer temperatures will lead to higher
reaction rate constants. Additionally, higher surface temper1
Forschungszentrum Karlsruhe GmbH, Institute for Meteorology and
Climate Research, Atmospheric Environmental Research, GarmischPartenkirchen, Germany.
Copyright 2006 by the American Geophysical Union.
0148-0227/06/2005JD006748$09.00
atures and increased radiation will result in higher emissions
of biogenic ozone precursors such as isoprene.
[4] A decrease in precipitation and a higher number of
cloud-free days may indicate an increase of high-pressure
situations and a decline in the frequency of cyclones, and a
reduced exchange of air masses. Finally, lower precipitation
and wet removal of soluble constituents, such as H2O2
or HNO3, may further contribute to an increase of the
concentrations of these compounds under future climate
conditions.
[5] On the global scale, the development of tropospheric
ozone under changed climate conditions has been investigated by several modeling studies [e.g., Roelofs et al., 1998;
Dameris et al., 1998; Johnson et al., 1999, 2001; Lelieveld
and Dentener, 2000; Hauglustaine and Brasseur, 2001].
The production of photooxidants depends strongly on the
regional distributions of precursors. Therefore the coarse
horizontal resolution of current global climate-chemistry
simulations does not permit an estimate of the effects of
climate change on tropospheric photooxidant distributions
on the regional scale. Regional climate modeling studies
focus mostly on meteorological applications [e.g. Jones et
al., 1997; Grell et al., 1998; Giorgi et al., 1998; Giorgi and
D12302
1 of 13
D12302
FORKEL AND KNOCHE: REGIONAL CLIMATE-CHEMISTRY SIMULATIONS
Mearns, 1999; Giorgi et al., 2001; Räisinen et al., 2004]. So
far, only a few investigations were carried out on the
potential effect of future global climate change on the
distribution of photooxidants on the regional scale [Ehhalt
et al., 2001]. Hogrefe et al. [2004] have shown that climate
change has a significant impact on the occurrence of high
ozone concentrations in the eastern United States by
performing scenario simulations with the meteorological
model MM5 and an offline chemistry transport model for
5 present-day and future summer seasons. In Europe,
numerical studies on the effect of changed climate conditions on tropospheric chemistry on the regional scale have
been limited to sensitivity studies or to the simulation of
episodes.
[6] In this paper, we investigate the potential effect of
climate change on tropospheric photooxidant concentrations
in southern Germany. Regional climate-chemistry simulations with the coupled three-dimensional meteorologychemistry model MCCM [Grell et al., 2000] were carried
out for two time slices of about 10 years each, one
representing the 1990s and one representing the 2030s. To
our knowledge, this is the first time that an online coupled
meteorology-chemistry model has been applied for climate
effect studies over a period of this length. The simulations
form a link between regional climate change and air quality
and permit a first quantitative estimate of the effects of
changed climate conditions on photochemistry for a region
in Europe.
2. Methods
[7] The investigation of the effect of global climate
change on regional distributions of meteorological variables
and photooxidants was performed by dynamical downscaling using the coupled meteorology-chemistry model
MCCM. On the basis of the concept of a one-way nesting
strategy, the lateral boundary conditions for the regional
limited-area model are derived from the output of a global
climate model. With respect to meteorological modeling
alone, dynamical downscaling of global climate scenarios
with limited area models can nowadays be considered as a
standard procedure [Giorgi and Mearns, 1999]. In this
study, a limited area model with online coupled photochemistry model was applied in order to account for the development of the concentrations of tropospheric ozone and
other compounds of photochemical smog.
2.1. Global Climate Scenario
[8] The time-dependent meteorological boundary conditions for the regional simulations for Europe as well as sea
surface temperatures were derived from the output of a 240
years run of the global climate model ECHAM4 [Roeckner
et al., 1996] coupled with the ocean general circulation
model OPYC [Oberhuber, 1993]. The ECHAM4/OPYC
model run is based on historic greenhouse gas concentrations between 1860 and 1990 and on the Intergovernmental
Panel on Climate Change (IPCC) emission scenario IS92a
[Intergovernmental Panel on Climate Change (IPCC),
2001] between 1990 and 2100.
[9] The horizontal resolution of the global simulation
used for the present study is T42. The output is available
for the corresponding Gaussian grid with 128 grid points in
D12302
east-west direction and 64 grid points in south-north direction and for a vertical resolution of 19 layers.
[10] For the regional simulations two time slices are
selected, 1991 – 2000 (denoted as ‘‘present’’) and 2031–
2039 (denoted as ‘‘future’’). The time slice from 2031–
2039 has been chosen since the investigation of a scenario
for the near future was intended. On the other hand, it is far
enough in the future to show a significant climate signal.
[11] Until 2035 the course of the CO2 emissions for the
‘‘business as usual’’ scenario IS92a runs in the middle
between the emissions described by the more recent scenarios A2 and B2. According to the scenario IS92a the
atmospheric CO2 volume mixing ratio increases by about
100 ppm between 1990 and 2030, which corresponds to an
increase by 28%. The resulting temperature increase simulated by ECHAM4 over southern Germany is about 2.
2.2. Regional Climate-Chemistry Model MCCM
[12] The coupled meteorology-chemistry model MCCM
[Grell et al., 2000] is based on the NCAR/Penn State
University meteorological model MM5 [Grell et al.,
1994]. In addition to the meteorological variables predicted
by MM5, MCCM includes prognostic equations for the
concentrations of gas phase pollutants and offers the option
to apply different gas phase chemistry mechanisms. The
simulation of atmospheric chemistry with an online coupled
model can be regarded as more consistent than an offline
treatment, as the chemistry part of the model receives all
necessary meteorological information directly from the
meteorological part of the model at each time step. Although the advantages of online coupled meteorologychemistry simulations against an off-line treatment are most
effective for fine horizontal resolutions, effects already
become significant at horizontal resolutions of around
30 km [Grell et al., 2004].
[13] Similar to MM5, MCCM permits the choice between
different physics and chemistry options. The following
options were applied for the simulations: ice phase cloud
microphysics according to Reisner et al. [1998], vertical
turbulent transport according Burk and Thompson [1989],
Grell cumulus parameterization [Grell et al., 1994], and
simple radiation schemes with explicit treatment of clouds
for short-wave and long-wave radiative transfer. Further
details of the physics options are given in the MM5
description [Grell et al., 1994]. The soil-vegetation-snow
processes are described by a multilayer model of Smirnova
et al. [1997]. This model solves prognostic equations for
soil temperature and water content, interception storage on
vegetation, and the mass of a snow layer.
[14] The chemical transformations were calculated with
the RADM2 gas phase chemistry mechanism [Stockwell et
al., 1990]. Evaluation of the RADM2 against numerous
smog chamber experiments has shown that the RADM2
yields maximum ozone concentrations which are generally
in good agreement with the experimental data [Dodge,
2000]. This widely used mechanism was chosen as it can
be applied to a very broad range of chemical scenarios.
It contains 38 predicted and 22 diagnosed species and
157 chemical reactions. With respect to the computational
effort this number of constituents and reactions is near the
upper limit for the simulation of atmospheric chemistry with
a 3-D model for a whole decade. Interactions with the
2 of 13
D12302
FORKEL AND KNOCHE: REGIONAL CLIMATE-CHEMISTRY SIMULATIONS
particulate phase, i.e., aerosols, cloud, and rainwater are not
considered in the present study except for a parameterization
of the conversion of N2O5 to HNO3 on particle surfaces.
[15] The photolysis frequencies necessary for the computation of the 21 photolytic reactions of the RADM2 mechanism are computed according to Madronich [1987] at each
grid point of the model. The simulated photolysis frequencies depend on the stratospheric ozone column, which is
prescribed, and on the profiles of temperature, tropospheric
ozone, and the mixing ratios of cloud liquid water and ice,
which are provided by the model.
[16] Biogenic hydrocarbons, particularly isoprene are
important precursors for tropospheric ozone. As described
by Grell et al. [2000], the emission rates of these species
and the emission of NO from soils are computed online
depending on the land use class assigned to each grid point,
and on the actual simulated values of temperature and
incoming solar radiation [Simpson et al., 1995; Guenther
et al., 1993]. The emission factors for different forest types
were adopted from Guenther et al. [1994] and are summarized by Grell et al. [2000, Table 1]. Anthropogenic
emissions of primary pollutants, like NOx, SO2, and hydrocarbons have to be supplied at hourly intervals.
[17] Various validation studies have shown that MCCM is
able to reproduce observed courses of meteorological variables and pollutant concentrations for different conditions
and regions of the earth [Grell et al., 1998; Grell et al.,
2000; Jazcilevich et al., 2003; Forkel et al., 2004; Suppan
and Schädler, 2004; E. Haas et al., Application and intercomparison of the RADM2 and RACM atmospheric chemistry mechanism including a new isoprene degradation
scheme within the online-coupled regional meteorology
chemistry model MCCM, submitted to International
Journal of Environment and Pollution, 2006, hereinafter
referred to as Haas et al., submitted manuscript, 2006;
D. Kim and W. R. Stockwell, Regional scale complex
terrain effects on the transport and transformation of air
pollutants, submitted to Atmospheric Environment, 2006].
Furthermore, MCCM has been successfully applied to
regional weather forecasts (http://imk-ifu.fzk.de/de/wetter/
index_wetter.htm) for several years.
[18] The meteorological part of MCCM, which is almost
identical with MM5, has been used for downscaling of all
15 years of the ERA15 reanalysis and has been extensively
evaluated against observations [Kotlarski et al., 2005]. For
Germany a bias of the mean temperature of less than 0.5
was found. The simulated precipitation was 12% lower than
observed. A detailed discussion of climate simulations with
MCCM with special regard to hydrological implications is
given by Kunstmann et al. [2004].
[19] Daily ozone forecasts have been performed continuously during the summer of 2005. Overall, the regional
patterns of the maximum temperature and the daily maximum ozone concentrations measured by the environmental
agencies of the German States and the Federal Environmental Agency (http://www.env-it.de/luftdaten/start.fwd?
setLanguage = en) are reproduced very well. However, high
ozone concentrations above 90 ppb are frequently underestimated by up to 10%.
[20] A comparison of the results of a continuous simulation without FDDA for two consecutive months of the
summer of 2001 (Haas et al., submitted manuscript, 2006)
D12302
with observed diurnal cycles at different sites in southern
Bavaria has shown that the model can capture the variability
of the synoptic situation and the duration of photosmog
episodes. Correlation coefficients around 0.85 were found
for temperature, 0.5 –0.6 for ozone, and 0.4– 0.5 for NO2.
Especially in the case of NO2, it must be considered that the
model results represent values for a whole grid cell, whereas
the data from the measurement network, in particular data
from the urban stations, are subject to local influences. A
restriction of the analysis to the observed high-ozone
episodes resulted in correlation coefficients around 0.75
for ozone.
2.3. Scenarios for Stratospheric Ozone and
Anthropogenic Emissions
[21] For the computation of UV radiation and photolysis
frequencies the depth of the stratospheric ozone layer is an
important input parameter. Monthly values of the average
zonal mean for the years 1985 – 1997 [Hein et al., 2001]
were used for the simulation of present-day conditions. The
change in the ozone layer depths between the 1990s and the
2030s was estimated on the basis of the scenario
‘‘PROB2050’’ given by Reuder et al. [2001] that was
derived from a global scenario simulation for the development of stratospheric ozone [Dameris et al., 1998]. According to this scenario the ozone layer depths in summer over
southern Germany can be expected to increase by 10 to
15 Dobson units.
[22] Anthropogenic emission data of primary pollutants
based on an emission inventory for 1998 [Friedrich et al.,
2000] were available in hourly intervals for the summer and
the winter season with a horizontal resolution of 20 km. The
7-day emission data sets were repeated in a weekly cycle
during each of the seasons. In order to isolate the effect of
climate change on the formation of photooxidants, the same
anthropogenic emissions were applied for the present-day
and the future time slice.
2.4. Setup of the Regional Climate-Chemistry
Simulations
[23] To obtain the desired horizontal resolution for the
region of southern Germany the climate-chemistry simulations with MCCM were carried in two consecutive nesting
steps. In the first step, the modeling area (domain D1, 66 59 grid points) has a horizontal resolution of 60 km and
covers entire Europe (Figure 1). In a second step, a model
domain with 20 km resolution (domain D2, 64 64 grid
points) covering the Alps and surrounding regions was
nested into the first domain. In the vertical the atmosphere
between the surface and the 100 hPa level is resolved in
25 layers, the lowest being about 14 m thick. The lower
boundary of the five5-layer soil model is located at a depth
of 3 m.
[24] For domain D1 the lateral boundary conditions for
the meteorological variables were derived from the 12-hourly
output of the ECHAM/OPYC run. As the global climate
simulations do not include chemistry, no consistent boundary
conditions for the chemical constituents were available.
Therefore time constant boundary conditions for the chemical compounds were applied. These boundary values were
derived from observed background concentrations and from
results of global chemistry simulations published in the
3 of 13
D12302
FORKEL AND KNOCHE: REGIONAL CLIMATE-CHEMISTRY SIMULATIONS
Figure 1. Map of the two model domains D1 and D2. The
white rectangle within D2 indicates the study area of
southern Germany.
literature [e.g., von Kuhlmann, 2001; von Kuhlmann et al.,
2004]. Boundary values near the surface and above the
boundary layer are summarized in Table 1 for selected
chemical compounds. Identical boundary conditions for the
chemical constituents were used for present and future
conditions.
[25] The choice of the chemical boundary conditions is
subject to some uncertainty and can affect the results of a
limited-area chemistry transport model. This holds particularly for the choice of the ozone mixing ratio. Ozone levels in
Europe can be affected by pollutant transport from regions
outside of Europe, especially from North America. As shown
by Li et al. [2002] North American emissions enhance
summertime surface ozone in Europe by 2 – 4 ppb on average
and contribute mostly to ozone mixing ratios in the range of
40– 60 ppb. This average transport is implicitly included by
the choice of the boundary conditions for the chemical
constituents. Singular trans-Atlantic transport events, which
can increase European ozone mixing ratios by 5 – 10 ppb are
not considered. These events are associated with low-pressure systems near on the North Atlantic and westerly winds
and hardly contribute to high ozone levels, which are usually
associated with stagnant high-pressure systems. During high-
D12302
pressure episodes, transport of pollutants from outside
Europe is only low and pollutant levels in Europe are mostly
determined by emissions within the regional model domain
[Jonson et al., 2001].
[26] As the impact of stratosphere-troposphere-exchange
on tropospheric ozone concentrations is only small in the
photochemically most relevant summer months [Lelieveld
and Dentener, 2000], its contribution was not explicitly
considered in the simulations.
[27] The large extension of domain D1 with horizontal
boundaries that are either located over the ocean or over
remote regions ensures the development of realistic concentration fields of photochemical pollutants over Europe.
A study by Jonson et al. [2001] with a regional model using
chemical boundary conditions from a global model has
shown that a model domain of the size of D1 is large
enough for the investigation of ozone formation in Europe.
[28] For domain D2 the lateral boundary conditions for
the meteorological as well as for chemical variables were
supplied by the MCCM run for domain D1. Though the
main focus of the simulations was on southern Germany,
domain D2 extends far toward the south, in order to ensure
a free flow around the Alps within D2.
[29] Land use data for the calculation of biogenic emissions and for the provision of soil and surface properties
needed for the meteorological part, such as roughness
length, were specified on the basis of CORINE land use
data. The land use data were interpolated to the model grids
of the two domains. Similar to the treatment in MM5, the
dominant land use class at each grid point was considered as
representative for this grid point. Anthropogenic emissions
were also interpolated from the original data set to the grids
of the two model domains. Figure 2 displays the distribution
of anthropogenic NOx emissions for domain D2.
[30] The simulations start in November of the years 1990
and 2030, respectively. The first two months were considered as spin-up time for the meteorological variables and for
soil temperature and moisture as well as snow depth. The
Table 1. Time-Invariant Boundary Values for Chemical Species
With Significant Mixing Ratios
Mixing Ratio, ppb
O3
NO
NO2
CO
HCHO
H2O2
HNO3
PAN
z = 14 m
z = 5000 m
40
0.1
0.1
150
0.3
0.3
0.3
0.5
45
0.001
0.001
100
0.1
0.4
0.3
0.5
Figure 2. Weekly average of anthropogenic NOx emissions in summer for domain D2.
4 of 13
D12302
FORKEL AND KNOCHE: REGIONAL CLIMATE-CHEMISTRY SIMULATIONS
D12302
Figure 3. Simulated mean ozone and NOx mixing ratios during summertime (June – August) for the
time slice 1991– 2000. The numbers show observed mean summertime mixing ratios for the years 1996 –
2000. The location of the ozone and NOx monitoring sites is indicated by the lower left corner of the
numbers.
simulations were carried out continuously until December
2000 and December 2039, yielding 10 and 9 years of model
output for the subsequent analysis, respectively.
3. Results and Discussion
[31] Compared to the ECHAM4 outputs the downscaling
for domain D1 shows more pronounced regional patterns of
pressure, precipitation, and temperature. Still more distinct
structures of the meteorological variables as well as the
chemical constituents are found for domain D2 due to the
better resolution of topography, and anthropogenic and
biogenic precursor emissions. As this was the main focus
of the research, the presentation of the results will concentrate on the part of D2 covering southern Germany.
3.1. Evaluation of Simulation Results for Present-Day
Conditions
[32] As MCCM is driven by meteorological boundary
conditions from a global climate simulation rather than a
simulation of current weather, the model results cannot be
compared with observations on a day-to-day basis, but only
with patterns of long-term mean values. The following
analysis mainly focuses on an evaluation of the simulated
fields of chemical compounds. As already mentioned in
section 2.2, the performance of the meteorological part of
the regional model has already been asserted in a different
study.
[33] Since the global climate simulation with ECHAM4
produces about 1 too high summertime temperatures for
Europe [Roeckner et al., 1996], a corresponding effect on
the results of the regional simulations can be expected.
Summertime precipitation in the midlatitudes is underpredicted by ECHAM4, while good agreement with observations has been found for the solar irradiance over Germany
[Liepert and Lohmann, 2001]. Since the bias can partly be
considered to be systematic in the model, a smaller bias can
be expected when differences between time slices are
analyzed. The regional distribution of the near-surface
temperature simulated with MCCM for the 1990s was
found to be in good agreement with observed climatological
mean temperatures in southern Germany. In correspondence
with the too low summertime precipitation simulated by
ECHAM4, the regional simulation similarly underestimates
the observed precipitation [Kunstmann et al., 2004].
[34] Figure 3 displays simulated summertime mean values of ozone and NOx for the time slice 1991 – 2000
together with observed mean summertime mixing ratios
(Lufthygienischer Jahresbericht, 1996– 2000, available at
http://www.bayern.de/lfu/luft/index.html). The spatial patterns as well as the absolute values of the mean mixing
ratios are reproduced rather well. In agreement with the
observations mean simulated ozone mixing ratios are high
at mountain sites. The observed mixing ratios of 53 ppb
(Wank mountain, 1776 m above sea level (asl)) and 43 ppb
(Tiefenbach, 750 m asl) near the southern and the eastern
border of Germany are slightly underpredicted by the
model. This can be attributed to the representation of the
topography, which is strongly smoothed for a horizontal
resolution of 20 km. The spatial distribution of NOx reflects
the location of major NO sources (see also Figure 2) like the
city of Munich. There a mean NOx mixing ratio of 71 ppb is
observed, which is very well reproduced by the simulations.
[35] Isoprene plays an important role for the formation of
tropospheric ozone and is expected to be very sensitive to
changed climate conditions. The simulated isoprene mixing
ratio over southern Germany shows a patchy pattern that
reflects mainly the locations of isoprene emitting plants in
the model. Since the dominant land use class at each grid
point was regarded as representative for this grid point,
isoprene emissions may be overestimated at single grid
points. A validation of the simulated present-day isoprene
mixing ratio is difficult since measured concentrations are
hardly available. For a spruce forest in northeastern Bavaria,
Klemm et al. [2006] report mean mixing ratios of 0.5 ppb
5 of 13
FORKEL AND KNOCHE: REGIONAL CLIMATE-CHEMISTRY SIMULATIONS
D12302
D12302
Table 2. Average Summertime (June – August) Values Over Southern Germany for Meteorological Impact
Parameters and Maximum Daily Ozone Mixing Ratios for Present and Future Conditions
1991 – 2000
2031 – 2039
15.6
119
5
488
385
1252
175
28.8
2.25
5.9
51.0
17.5
104
4
625
398
1317
192
34.4
2.56
5.6
54.7
Temperature, C
Vertically integrated cloud water, g m2
Vertically integrated cloud ice, g m2
Cloud free hours
PBL height, m
Maximum PBL height, m
Solar radiation, W m2
Number of days with high insolationa
Mean duration of high insolation episodes,a days
Events with 3-hour pressure tendency >2 hPa
Daily maximum ozone mixing ratio, ppb
a
Daily integral of solar radiation >70% of the daily clear-sky solar radiation.
and 0.27 ppb during the summers of 2001 and 2002,
respectively. The simulated mean mixing ratio for the same
site is 0.5 ppb. Since isoprene emissions from spruce are
comparatively low, higher isoprene concentrations can be
expected at grid points covered with mixed forest. A high
emission factor resulting in isoprene mixing ratios between
1.5 and 3 ppb was assigned to mixed forest [Guenther et al.,
1994; Grell et al., 2000]. However, the composition of
mixed forest shows high variability over Germany. Therefore the chosen emission factor probably results in an
overprediction of the isoprene mixing ratios at locations
where spruce and beech are the predominant species of the
mixed forest.
3.2. Regional Climate Change
[36] The occurrence of high summertime concentrations
of tropospheric ozone and other photooxidants is strongly
determined by meteorological processes within the planetary
boundary layer. The well-known fact that high photooxidant
concentrations are associated with stagnant high-pressure
situations during summer has been confirmed by several
studies that show the close relation between ozone concentrations and the meteorological situation [e.g., Vukovich,
1995; Rao et al., 2003, Ordonez et al., 2005]. Meteorological conditions which are known to be associated with high
ozone concentrations are little daytime cloud cover, low
surface wind speed, high mixing heights [Rao et al., 2003],
high solar irradiation, and high temperatures [Ordonez et al.,
2005]. Maximum ozone concentrations depend also on the
duration of stalled high-pressure systems, which means on
the days elapsed since the last frontal passage [Ordonez et
al., 2005].
[37] High-level photosmog episodes are usually terminated
by synoptic activity, leading to an air mass exchange.
Disturbed weather conditions associated with low-pressure
systems, fronts, high cloudiness, and precipitation usually
prevent the occurrence of high ozone concentrations. Mickley
et al. [2004] made an attempt to identify frontal passages by
analyzing the change of the sea level pressure and found a
decrease from 7.5 to 6.8 events per summer. In the present
study, the 3-hour pressure tendency was used as an indication for synoptic activity. The number of days with a
maximum pressure tendency above 2 hPa decreases slightly
in the future from 5.9 events per summer to 5.6.
[38] For the future conditions, the average number of
days with high insolation (which were here defined by a
daily integral of solar radiation of at least 70% of the daily
clear-sky radiation) increases by 6 to 8 days per summer
season in the western part and by about 4 to 7 days in the
eastern part of southern Germany. For the development of
high maximum ozone concentrations several consecutive
days with sunny weather are necessary. The average number
of sunny episodes with a minimum length of four days
during the months June to August was found to increase by
about 2 in southern Germany. In the Black Forest region and
in northern Bavaria this corresponds to an increase by 60%.
In southeastern Bavaria, where a comparatively high number of sunny episodes is simulated for present-day conditions, the relative increase is only 20%.
[39] Summertime averages over southern Germany of
meteorological parameters associated with high ozone levels and the daily maximum mixing ratio of ozone are
compiled in Table 2 for present and future climate conditions. Table 2 shows that the coherences, which were
identified in the studies mentioned above, also show up
when present and future climate conditions are compared.
[40] Similar to the results of the global simulation, the
regional model calculates a significant rise of the nearsurface air temperature between the 1990s and the 2030s for
all months of the year. During the photochemically most
relevant summer months, June to August, the highest
temperature increase in southern Germany, i.e., 2.1 within
40 years, was found in the upper Rhine valley near the
border to France (Figure 4a). Toward the east the difference
between future and present-day temperature decreases. An
increase of only 1.7 was simulated for the southeastern part
of Bavaria.
[41] The meteorological parameter with the highest impact on photochemistry is cloud cover since it controls the
amount of solar radiation reaching the lower troposphere.
For future climate conditions the simulations yield a 20 to
40% increase of the number of hours with clear-sky conditions. This corresponds to a pronounced decrease of the
vertically integrated cloud water content over southern
Germany during the summer months. The regional distribution shows the strongest decrease in the western part of
southern Germany and lower values in the central part
(Figure 4b). Relative to the simulated cloud cover for
present-day conditions the reduction corresponds to 5 and
10% in the central area and almost to 20% in the west. The
vertically integrated cloud ice content also decreases for
future climate conditions. As cloud ice contributes only 5 to
6 of 13
D12302
FORKEL AND KNOCHE: REGIONAL CLIMATE-CHEMISTRY SIMULATIONS
D12302
Figure 4. Difference of (a) temperature and (b) vertically integrated cloud water content between
2031 –2039 and 1991 –2000 for the months June – August.
10% to the total cloud water, the main impact on incoming
radiation is due to the decrease of the liquid cloud water
content.
3.3. Impact on Photolysis Frequencies and Precursor
Emissions
[42] Photolysis frequencies and biogenic VOC emissions
are quantities which are directly affected by a change in
near-surface air temperature and cloudiness. Figure 5 shows
the difference between present-day and future climate conditions for the photolysis frequencies JNO2 for NO2, and JO3
for the reaction O3 ! O(1D) + O2, which can be regarded as
the two most important photolysis frequencies for tropospheric photochemistry. Furthermore, JNO2 and JO3 reflect
the effect of changed cloud water content on incoming solar
radiation and on UV radiation, respectively.
[43] The regional pattern of the change in JNO2 is almost
identical with the pattern of the change in incoming solar
radiation. The difference of JNO2 between future and present-day conditions during the summer months (Figure 5a)
follows mainly the pattern of the change in cloud water
content. As changes in cloud water content only have a
small additional effect for optically very thick clouds, no
complete match of the patterns can be expected.
[44] In the case of JO3 (Figure 5b) a decrease is found for
large areas, which correspond to the regions where the
difference in cloud water content between future and
present-day conditions is only small (see also Figure 4b).
The expected recovery of the stratospheric ozone layer by
10 to 15 Dobson units until 2030 and, to a minor extent, the
increase of tropospheric ozone results in a 6 to 10% lower
UV-B radiation (not shown) and JO3. Therefore the effect of
decreasing cloud water is partly compensated by the effect
of the higher ozone column depth.
[45] The maximum increase of solar radiation in summer
is up to 17% for future climate conditions while the
Figure 5. Difference of the photolysis frequencies (a) JNO2 and (b) JO3 between 2031 – 2039 and 1991 –
2000 for the months June– August.
7 of 13
D12302
FORKEL AND KNOCHE: REGIONAL CLIMATE-CHEMISTRY SIMULATIONS
D12302
Figure 6. Difference of (a) the isoprene emission and (b) NOx mixing ratio between 2031– 2039 and
1991 –2000 for the months June –August.
maximum increase of UV-B radiation over southern Germany is only 10%. Accordingly the maximum increase of
JNO2 is 11%, whereas the maximum increase of JO3 is only
about 8%.
[46] As the emission of isoprene increases with solar
irradiance and temperature, higher emissions are simulated
for future climate conditions under the assumption that
vegetation remains unchanged. The spatial distribution of
the increase of isoprene emissions is superimposed by the
distribution of forest areas with high isoprene emissions.
Therefore the increase of the isoprene emission shows a
patchy pattern that reflects mainly the locations of isoprene
emitting plants in the model (Figure 6a). The maximum
increase, corresponding to 30– 50% of the simulated values
for 1991 – 2000, was found in the Black Forest area and the
forested Alpine regions. The increase in the emissions of
monoterpenes and other biogenic VOC shows a similar
pattern.
[47] An increase in temperature also results in enhanced
NO emissions from the soil. Since the emission of NO is
strongest for areas with agricultural land use, it shows the
highest values as well as the strongest increase of about
0.01 kg m2 h1 in rural areas without isoprene emission.
Compared to the total emission of NOx, which is mostly of
anthropogenic origin, the higher NO emissions from agricultural soils under future climate condition add only about
1 to 5% to the total emissions of NOx. This increase in NO
emissions results in a slightly increased level of NOx for a
major part of southern Germany (Figure 6b). The increase is
most pronounced in the area along the Danube and in the
upper Rhine valley. However, for the urban area around
Frankfurt a decrease of the near-surface mixing ratio of NOx
was found, which can be ascribed to an increase of mixedlayer height in combination with the absence of significant
NO emissions from soils in this region.
3.4. Impact on Regional Distribution of Photooxidants
[48] The concentration of tropospheric ozone depends on
the regional distribution of sources of precursors such as
NOx and hydrocarbons from anthropogenic and biogenic
sources as well as on the meteorological situation. As
already mentioned, the future meteorological conditions in
southern Germany are more favorable for the formation of
tropospheric ozone. Higher isoprene concentrations as a
consequence of enhanced isoprene emissions, and, to a
minor extent, higher NO emissions from the soil in rural
areas are additional factors to promote the formation of
tropospheric ozone and other photooxidants.
[49] The increase in isoprene concentrations (not shown)
closely follows the pattern of the change in isoprene
emission shown in Figure 6a. In the regions with the
strongest increase near the southern and western border of
Germany the mean isoprene mixing ratio increases by 1 –
1.5 ppb or 35 – 45%. Under the model assumption of
unchanged anthropogenic emissions the simulated change
in the meteorological conditions and the higher isoprene
mixing ratios for future climate conditions result in an
average increase of the mean daily maximum of nearsurface ozone (Figure 7a). The increase ranges between
2 ppb in the northern part of Bavaria and 6 ppb toward the
southwest, which corresponds to an increase by 6 – 10%. As
the highest ozone mixing ratios usually do not occur
directly in the source regions of precursors but downwind
of these regions the increase in the mean ozone maxima
reflects only very roughly the areas with high increase of
solar radiation and isoprene emissions.
[50] Figure 7a shows that the highest increase of the
maximum ozone concentration seems to occur in regions
where a high increase of the isoprene emissions coincides
with high NOx emissions (area of Stuttgart (left of the center
of Figure 7a)), northern Switzerland, and the Milan region
in northern Italy (at the southern margin of Figure 7a). A
comparatively small increase of the maximum ozone concentration is found for the high-isoprene regions near the
border between Germany and Austria, in the Vosges in
eastern France, and some other regions, with low anthropogenic NO emissions. On the other hand, a considerable
increase of the maximum ozone concentration is calculated
8 of 13
D12302
FORKEL AND KNOCHE: REGIONAL CLIMATE-CHEMISTRY SIMULATIONS
D12302
Figure 7. Difference (a) of the mean daily ozone maximum and (b) of the number of days with
exceedances of the threshold value of 120 mg m3 for the 8-hour mean between 2031 –2039 and 1991 –
2000 for the months June– August.
for the polluted upper Rhine valley and a band with low
isoprene emissions between the southern border of Germany and the Danube river.
[51] According to the simulations, the target value of
120 mg m3 (about 60 ppb) for the 8-hour mean of the
ozone concentration will be exceeded by additional 5 to
12 days per year for the given scenario (Figure 7b). As the
8-hour mean of the near-surface ozone concentration under
present-day conditions is already close to the threshold in
the southern part of the study area, a stronger increase in the
number of days with threshold exceedance is found there.
[52] The frequency distribution of the simulated daily
ozone maxima during summer is shifted toward higher
values (Figure 8). This results in a significant increase of
the number of days with near-surface ozone concentrations
higher than 180 mg m3 (90 ppb) in southern Germany. For
the 232 grid points of the considered region, the average
number of events when the threshold of 180 mg m3 is
exceeded increases from 99 to 384 events per summer, i.e.,
by a factor of 3.9.
[53] The changed climate conditions and their effect on
photolysis frequencies affect all major compounds of photochemical smog. Odd hydrogen radicals (HOx = OH +
HO2) play a key role in tropospheric photochemistry. The
simulated increase of the HOx mixing ratio (Figure 9a)
reflects the pattern of isoprene emissions and concentrations. This can be explained by the fact that regions with
high emissions of anthropogenic NO and mostly low HO2
mixing ratios are almost complementary to regions with
high biogenic VOC emissions. The latter represent those
areas where the volume ratio of VOC/NOx is larger than 1,
whereas VOC/NOx is lower in those regions where anthropogenic emissions are dominant.
Figure 8. (left) Frequency distribution of the simulated daily ozone maxima averaged over southern
Germany during summer (June– August) for the years 1991 – 2000 and 2031 – 2039. (right) A zoom of the
high-ozone part.
9 of 13
D12302
FORKEL AND KNOCHE: REGIONAL CLIMATE-CHEMISTRY SIMULATIONS
D12302
Figure 9. Difference of mean volume mixing ratios of (a) odd hydrogen, (b) formaldehyde, (c) nitric
acid, and (d) peroxy actylnitrate between 2031– 2039 and 1991– 2000 for the months June– August.
[54] Carbonyl compounds produced during the oxidation
of hydrocarbons, in particular formaldehyde (HCHO), promote additional production of ozone. The formation of
formaldehyde strongly depends on the concentration of
isoprene, therefore the increase of the HCHO mixing ratio
(Figure 9b) matches the pattern of the isoprene emission in
Figure 6a. This also holds for the increase of H2O2 and
organic acids (not shown). In the areas of maximum change
this corresponds to 25% higher values in the case of
formaldehyde and to 15– 20% for H2O2. As removal of
pollutants by precipitation is not included in the model, the
simulated increase in the mixing ratios of soluble compounds can be considered as a lower limit. The change of
the concentrations of soluble compounds would even be
more pronounced if the simulated decrease in precipitation
and wet removal were taken into account.
[55] In contrast to HCHO and H2O2, where the highest
mean concentrations and the maximum increase are mostly
found in rural regions with high isoprene emissions, the
increase of nitric acid (HNO3) dominates in the regions with
high NOx emissions (see also Figure 2). Consequently, the
highest increase of HNO3 (Figure 9c), which is formed to a
large extent by the reaction of NO2 with OH, is mainly
found in urban regions.
[56] The mixing ratio of PAN (peroxy acetylnitrate)
decreases in most parts of southern Germany (Figure 9d).
This is mainly due to higher temperatures, which result in
an enhanced thermal decomposition of PAN. Apart from the
temperature, the mixing ratio of PAN also depends on the
NO2/NO ratio. While NO2 promotes the formation of PAN,
NO prevents the recombination of NO2 and the acylperoxy
radical. As a consequence, NO2/NO ratios around 10 will
lead to an ‘‘effective lifetime’’ of PAN which is about
3 times longer than the one for a NO2/NO ratio of 1 – 2.
A ratio of NO2/NO close or above 10 was already found in
the mountainous regions for present-day conditions. This
study indicates even a further increase under future climate
conditions. An increase of PAN mixing ratios between 0.05
and 0.15 ppb (corresponding to 4 – 10%) was simulated for
these mountainous regions. The decrease of PAN mixing
ratios is restricted to regions at low altitudes, where NO2/
NO ratios between 1 and 3 are usually found. The simu-
10 of 13
D12302
FORKEL AND KNOCHE: REGIONAL CLIMATE-CHEMISTRY SIMULATIONS
lations show that for future climate conditions a further
decrease of this ratio will be likely in these areas.
4. Conclusions
[57] Online coupled regional climate-chemistry simulations were carried out for two time slices with a length of
about one decade each, in order to obtain regional
distributions of meteorological parameters and photooxidant concentrations for present-day and future climate
conditions. The model results show how a possible
climate change according to the greenhouse gas scenario
IS92a can affect the meteorological conditions and photochemistry in southern Germany within the next
30 years.
[58] For the summer months the simulations show an
increase of the length of clear-sky episodes, an increase in
near-surface temperature, and a decrease of cloud water and
cloud ice. The decrease in cloudiness results in higher
insolation and photolysis frequencies at the surface. For
the underlying scenario assuming a recovery of the stratospheric ozone layer within the next 30 years the impact of
changed stratospheric ozone on UV radiation and JO3 was
found to be small compared to the effect of changed cloud
cover due to climate change.
[59] The higher incoming radiation for future climate
conditions results in higher biogenic emissions of isoprene,
which increase regionally by up to 50%, and in an enhanced
formation of ozone and other photooxidants. Assuming
unchanged anthropogenic precursor emissions, 10% higher
daily maximum values of near-surface ozone concentrations
were simulated.
[60] Despite of existing correlations between meteorological parameters and daily maximum concentrations, the
regions with the highest increase in maximum ozone concentrations are not identical with the regions with the largest
increase in temperature, solar radiation or isoprene emission. This indicates that changes in meteorological variables
resulting from climate change cannot be extrapolated in a
simple way in order to estimate future changes in ozone
concentration.
[61] The number of days when the threshold value of
120 mg m3 for the 8-hourly mean is exceeded were
found to increase by 5 days in northern Bavaria and
12 days in the south. A pronounced increase was also
found for the occurrence of high ozone concentrations.
Considering the average over southern Germany, the
number of events with maximum ozone concentrations
above 180 mg m3 increases by a factor of 3.9, which
demonstrates that there is a higher probability for very
high ozone concentrations in southern Germany under
future climate conditions.
[62] Since identical boundary values for the mixing
ratios of the chemical compounds were used for present
and future conditions, the simulated effect might even
represent a lower limit, as higher boundary values for the
photochemical compounds can be expected in the future.
As unchanged anthropogenic emissions were assumed for
the present study, the increase of maximum ozone concentrations due to the more favorable conditions for photo
smog situations could be compensated by regional emission reduction measures. However, the increase in the
D12302
emission of biogenic compounds under future climate
conditions should be considered when mitigation strategies
are projected.
[63] For the interpretation and a possible further processing of the results it should always be kept in mind that
these simulations are not a forecast, but a scenario simulation. The main assumptions comprehend the future
development of greenhouse gas concentrations specified
by the greenhouse gas emission scenario IS92a, the choice
of the stratospheric ozone scenario, and the assumption of
unchanged anthropogenic emissions of tropospheric ozone
precursors. Further uncertainties can be attributed to unpredictable internal variations of the climate system or
unpredictable events such as volcano eruptions. Additionally, shortcomings of the global and regional model may
have an effect on the results. However, the global climate
model ECHAM4 represents an internationally accepted
standard. An intercomparison of global climate models
shows that the extent and the patterns of global warming
are slightly different for each model, but all models show
agreement for the general trend. The regionalization of
global models by regional models is also known as a
reliable technique. The regional climate-chemistry model
MCCM has proven its ability to regionalize global climate
patterns and to reproduce observed pollutant concentration
fields.
[64] Uncertainties with respect to atmospheric chemistry
can result from uncertainties related with the calculation of
biogenic VOC emissions, i.e., from the distribution of land
cover, the application of a majority based land use, and the
specification of BVOC emission factors. Furthermore, the
results depend on the anthropogenic emission scenario and
on the choice of the lateral boundary conditions for the
chemical constituents for future climate conditions. The
latter problem could be overcome, if output of global
climate-chemistry simulations were available, which will
surely be the case in near future.
[65] With respect to the statistical analysis of climate data,
ensembles made up of only one decade are still near the
lower limit. Therefore the climatological interpretation of
the results can still be limited due to the limited length and
the choice of the time slices.
[66] Nevertheless, the results of the MCCM simulations
presented here can be regarded as meaningful under the
premise of the underlying ECHAM4 scenario simulation
and the given precursor emissions. The availability of
regional distributions of ozone and other photooxidants
as well as primary pollutants for present-day and future
conditions offers the possibility for further impact studies
with respect to ecological, economical, and biological
systems [Kesik et al., 2006], and for the development of
mitigation and adaptation strategies. However, for future
applications the availability of output from global climatechemistry models is highly desirable in order to increase
the significance of the simulation results.
[67] Acknowledgments. This investigation was funded by the Bavarian Ministry for Environment, Health, and Consumer Protection within
the joint project BayForUV. The global climate simulations with ECHAM4
were supplied by the DKRZ (Deutsches Klimarechenzentrum) and the Max
Planck Institute for Meteorology in Hamburg. The anthropogenic emission
data were supplied by IER, University of Stuttgart. The simulations were
carried out on three double-processor AMD 1900+ PCs with LINUX
11 of 13
D12302
FORKEL AND KNOCHE: REGIONAL CLIMATE-CHEMISTRY SIMULATIONS
operating system. The authors are particularly indebted to E. Haas and
J. Werhahn, who insured the permanent availability of the computers. We
thank the anonymous reviewers for helpful comments and suggestions. We
would also like to acknowledge helpful comments by several colleagues, in
particular, W. R. Stockwell and H. E. Scheel.
References
Burk, S., and W. Thompson (1989), A vertically nested regional numerical
weather prediction model with second order closure physics, Mon.
Weather Rev., 117, 2305 – 2324.
Dameris, M., V. Grewe, R. Hein, C. Schnadt, C. Brühl, and B. Steil (1998),
Assessment of the future development of the ozone layer, Geophys. Res.
Lett., 25, 3579 – 3582.
Dodge, M. C. (2000), Chemical oxidant mechanisms for air quality modeling: Critical review, Atmos. Environ., 34, 1435 – 1453.
Ehhalt, D., et al. (2001), Atmospheric chemistry and greenhouse gases, in
Contribution of Working Group I to the Third Assessment Report of the
Intergovernmental Panel on Climate Change, edited by J. T. Houghton et
al., pp. 241 – 287, Cambridge Univ. Press, New York.
Forkel, R., G. Smiatek, F. Hernandez, R. Iniestra, B. Rappenglück, and
R. Steinbrecher (2004), Numerical simulations of ozone level scenarios
for Mexico City, paper presented at 84th AMS Annual Meeting, 6th
Conference on Atmospheric Chemistry: Air Quality in Megacities,
Seattle, Wash., 11 – 15 Jan.
Friedrich, R., B. Wickert, U. Schwarz, and S. Reis (2000), Improvement
and application of methodology and models to calculate multiscale high
resolution emission data for Germany and Europe, in Genemis Annu.
Rep. 1999, pp. 28 – 33, EUROTRAC Int. Sci. Secr., Munich, Germany.
Giorgi, F., and L. O. Mearns (1999), Introduction to special section:
Regional climate modelling revisited, J. Geophys. Res., 104, 6335 –
6352.
Giorgi, F., L. O. Mearns, C. Shields, and L. McDaniel (1998), Regional
nested model simulations of present and 2xCO2 climate over the Central
Plains of the U.S., Clim. Change, 40, 457 – 493.
Giorgi, F., B. Hewitson, J. Christensen, M. Hulme, H. Von Storch,
P. Whetton, R. Jones, L. Mearns, and C. Fu (2001), Regional climate
information: Evaluation and Projections, in Climate Change 2001. Contribution of Working Group I to the Third Assessment Report of the
Intergovernmental Panel on Climate Change, edited by J. T. Houghton
et al., pp. 585 – 638, Cambridge Univ. Press, New York.
Grell, G. A., J. Dudhia, and D. R. Stauffer (1994), A description of the fifthgeneration Penn State/NCAR Mesoscale Model (MM5), NCAR Tech.
Note TN-398 + STR, 122 pp.
Grell, G., L. Schade, R. Knoche, and A. Pfeiffer (1998), Regionale Klimamodellierung, final report, Joint Proj. BayFORKLIM, Subproject K2,
Fraunhofer-Inst. für Atmos. Umweltforschung, Garmisch-Partenkirchen,
Germany.
Grell, G., S. Emeis, W. R. Stockwell, T. Schoenemeyer, R. Forkel,
J. Michalakes, R. Knoche, and W. Seidl (2000), Application of a multiscale, coupled MM5/chemistry model to the complex terrain of the
VOTALP valley campaign, Atmos. Environ., 34, 1435 – 1453.
Grell, G. A., R. Knoche, S. E. Peckham, and S. A. McKeen (2004), Online
versus offline air quality modeling on cloud-resolving scales, Geophys.
Res. Lett., 31, L16117, doi:10.1029/2004GL020175.
Guenther, A. B., P. R. Zimmerman, P. C. Harley, R. K. Monson, and
R. Fall (1993), Isoprene and monoterpene emission rate variability:
Model evaluations and sensitivity analyses, J. Geophys. Res., 98,
12,609 – 12,617.
Guenther, A., P. Zimmerman, and M. Wildermuth (1994), Natural volatile
organic compound emission rate estimates for US woodland landscapes,
Atmos. Environ., 28, 1197 – 1210.
Hauglustaine, D. A., and G. P. Brasseur (2001), Evolution of tropospheric
ozone under anthropogenic activities and associated radiative forcing of
climate, J. Geophys. Res., 106, 32,337 – 32,360.
Hein, R., M. Dameris, C. Schnadt, C. Land, V. Grewe, I. Köhler,
M. Ponater, R. Sausen, and B. Steil (2001), Results of an interactively
coupled atmospheric chemistry: General circulation model: Comparison
with observations, Ann. Geophys., 19, 435 – 457.
Hogrefe, C., B. Lynn, K. Civerolo, J.-Y. Ku, J. Rosenthal, C. Rosenzweig,
R. Goldberg, S. Gaffin, K. Knowlton, and P. L. Kinney (2004), Simulating changes in regional air pollution over the eastern United States due to
changes in global and regional climate and emissions, J. Geophys. Res.,
109, D22301, doi:10.1029/2004JD004690.
Hulme, M., and N. Sheard (1999), Climate change scenarios for Germany,
report, 6 pp., Clim. Res. Unit, Norwich, U. K.
Intergovernmental Panel on Climate Change (IPCC) (2001), Climate
Change 2001: The Scientific Basis, Contribution of Working Group I
to the Third Assessment Report of the Intergovernmental Panel on Climate Change, edited by J. T. Houghton et al., 881 pp., Cambridge Univ.
Press, New York.
D12302
Jazcilevich, A. D., A. R. Garcia, and L. G. Ruiz-Suarez (2003), A study of
air flow patterns affecting pollutant concentrations in the central region of
Mexico, Atmos. Environ., 37, 183 – 193.
Johnson, C. E., W. J. Collins, D. S. Stevenson, and R. G. Derwent (1999),
Relative roles of climate and emissions changes on future tropospheric
oxidant concentrations, J. Geophys. Res., 104, 18,631 – 18,645.
Johnson, C. E., D. S. Stevenson, W. J. Collins, and R. G. Derwent (2001),
Role of climate feedback on methane and ozone studied with a oceanatmospher-chemistry model, Geophys. Res. Lett., 28, 1723 – 1726.
Jones, R. G., J. M. Murphy, M. Nouguer, and A. B. Keen (1997), Simulation of climate change over Europe using a nested regional climate
model. II: Comparison of driving and regional model responses to a
doubling of carbon dioxide, Q. J. R. Meteorol. Soc., 123, 265 – 292.
Jonson, J. E., J. K. Sundet, and L. Tarrason (2001), Model calculations of
present and future levels of ozone and ozone precursors with a global and
a regional model, Atmos. Environ., 35, 525 – 537.
Kesik, M., N. Brüggemann, R. Forkel, R. Knoche, C. Li, G. Seufert,
D. Simpson, and K. Butterbach-Bahl (2006), Future scenarios of N2O
and NO emissions from European forest soils, J. Geophys. Res.,
doi:10.1029/2005JG000115, in press.
Klemm, O., et al. (2006), Experiments on forest/atmosphere exchange:
Climatology and fluxes during two summer campaigns in NE Bavaria,
Atmos. Environ., in press.
Kotlarski, S., A. Block, U. Böhm, D. Jacob, K. Keuler, R. Knoche,
D. Rechid, and A. Walter (2005), Regional climate model simulations
as input for hydrological applications: Evaluation of uncertainties, Adv.
Geosci., 5, 119 – 125.
Kunstmann, H., K. Schneider, R. Forkel, and R. Knoche (2004), Impact
analysis of climate change for an Alpine catchment using high resolution
dynamic downscaling of ECHAM4 time slices, Hydrol. Earth Syst. Sci.,
8, 1030 – 1044.
Lelieveld, J., and F. Dentener (2000), What controls tropospheric ozone,
J. Geophys. Res., 105, 3531 – 3551.
Li, Q., et al. (2002), Transatlantic transport of pollution and its effects on
surface ozone in Europe and North America, J. Geophys. Res., 107(D13),
4166, doi:10.1029/2001JD001422.
Liepert, B., and U. Lohmann (2001), A comparison of surface observations
and ECHAM4-GCM experiments and its relevance to the indirect aerosol
effect, J. Clim., 14, 1078 – 1091.
Madronich, S. (1987), Photodissociation in the atmosphere: 1. Actinic flux
and the effects of ground reflections and clouds, J. Geophys. Res., 92,
9740 – 9752.
Mickley, L. J., D. J. Jacob, B. D. Field, and D. Rind (2004), Effects of
future climate change on regional air pollution episodes in the United
States, Geophys. Res. Lett., 31, L24103, doi:10.1029/2004GL021216.
Oberhuber, M. (1993), The OPYC ocean general circulation model, Rep. 7,
130 pp., Dtsch. Klim. GmbH (DKRZ)/MPI für Meteorol., Gruppe Modelle and Daten. (Available at http://www.mad.zmaw.de/service-support/
documents/reports/)
Ordonez, C., H. Methis, M. Furger, S. Henne, C. Hüglin, J. Staehelin, and
A. S. H. Prevot (2005), Changes of daily surface ozone maxima in
Switzerland in all seasons from 1992 to 2002 and discussion of summer
2003, Atmos. Chem. Phys., 5, 1187 – 1203.
Räisinen, J., U. Hansson, A. Ullerstig, R. Döscher, L. P. Graham, C. Jones,
H. E. M. Meier, P. Samuelson, and U. Willen (2004), European climate in
the late twenty-first century: Regional simulations with two driving global models and two forcing scenarios, Clim. Dyn., 22, 13 – 31,
doi:10.1007/s00382-003-0365-x.
Rao, S. T., J. Y. Ku, S. Berman, D. Zhang, and H. Mao (2003), Summertime characteristics of the atmospheric boundary layer and relationships
to ozone levels over the eastern United States, Pure Appl. Geophys., 160,
21 – 55.
Reisner, J., R. M. Rasmussen, and R. T. Bruintjes (1998), Explicit forecasting of supercooled liquid water in winter storms using the MM5 mesoscale model, Q. J. R. Meteorol. Soc., 124, 1071 – 1107.
Reuder, J., P. Koepke, and M. Dameris (2001), Future UV radiation in
central Europe modelled from ozone scenarios, J. Photochem. Photobiol.,
Ser. B, 61(3), 94 – 105.
Roeckner, E., K. Arpe, L. Bengtsson, M. Christoph, M. Claussen,
L. Dümenil, M. Esch, M. Giorgetta, U. Schlese, and U. Schultz-Weida
(1996), The atmospheric general circulation model ECHAM4: Model
description and simulation of the present-day climate, Rep. 218,
90 pp., MPI für Meteorol., Hamburg, Germany.
Roelofs, G. J., J. Lelieveld, and J. Feichter (1998), Model simulations of the
changing distribution of ozone and ist radiative forcing of climate: Past,
present and future, Rep. 283, 24 pp., MPI für Meteorol., Hamburg, Germany.
Simpson, D., A. Guenther, C. N. Hewitt, and R. Steinbrecher (1995), Biogenic emissions in Europe: 1. Estimates and uncertainties, J. Geophys.
Res., 100, 22,875 – 22,890.
12 of 13
D12302
FORKEL AND KNOCHE: REGIONAL CLIMATE-CHEMISTRY SIMULATIONS
Smirnova, T. G., J. M. Brown, and S. G. Benjamin (1997), Performance of
different soil model configurations in simulating ground temperature and
surface fluxes, Mon. Weather Rev., 125, 1870 – 1884.
Stockwell, W. R., P. Middleton, J. S. Chang, and X. Tang (1990), The
second generation regional acid deposition model chemical mechanism
for regional air quality modeling, J. Geophys. Res., 95, 16,343 –
16,367.
Suppan, P., and G. Schädler (2004), The impact of highway emissions on
ozone and nitrogen oxide levels during specific meteorological conditions, Sci. Total Environ., 334 – 335, 215 – 222, doi:10.1016/j.scitotenv.
2004.04.060.
von Kuhlmann, R. (2001), Tropospheric photochemistry of ozone, its precursors and the hydroxyl radical: A 3d-modeling study considering
non-methane hydrocarbons, Ph.D. thesis, Johannes Gutenberg-Univ.
D12302
Mainz, Mainz, Germany. (Available at http://www.mpch-mainz.mpg.
de/kuhlmann)
von Kuhlmann, R., M. G. Lawrence, U. Pöschl, and P. J. Crutzen (2004),
Sensitivities in global scale modeling of isoprene, Atmos. Chem. Phys., 4,
1 – 17.
Vukovich, F. M. (1995), Regional-scale boundary-layer ozone variations in
the eastern United States and their association with meteorological variations, Atmos. Environ., 29, 2259 – 2273.
R. Forkel and R. Knoche, Forschungszentrum Karlsruhe GmbH, Institute
for Meteorology and Climate Research, Atmospheric Environmental
Research (IMK-IFU), Kreuzeckbahnstr. 19, D-82467 Garmisch-Partenkirchen, Germany. ([email protected]; hans-richard.knocke@imk.
fzk.de)
13 of 13