Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

15.1 Introduction

The interaction of solar coronal mass ejections and the solar wind with the interplanetary plasma and processes in the Earth’s magnetosphere accelerate particles as protons, electrons, or He nuclei to energies up to the GeV range. Some of these energetic particles penetrate the atmosphere mainly at high geomagnetic latitudes and lose their kinetic energy by cascades of inelastic collisions. Via ionization, dissociation and ion reactions this energy deposition ultimately produces reactive molecules like NO, NO2, OH and others. This significant enhancement of NOx (=NO+NO2+N) and HOx (=H,OH,HO2) causes additional ozone loss and disturbs other trace gas distributions.

During polar night and the absence of photochemistry the impact of energetic particle precipitation (EPP) on the chemical state of the atmosphere is most pronounced per se. But the interhemispheric diabatic circulation with its downward branch in the polar winter hemisphere in addition allows to propagate the result of this interaction, which mainly takes part in the thermosphere and mesosphere, to the stratosphere. The key processes for the impact of EPP on the middle atmosphere are therefore the combination of disturbed chemistry and transport in the winter polar middle atmosphere.

After its maximum in the year 2000/2002 the solar cycle 23 exhibited prolonged activity which gave rise to several extraordinary manifestations of solar-terrestial connections in the Earth’s middle atmosphere. Several strong flare events and several strong geomagnetic storms were responsible for remarkable chemical disturbances in the middle atmosphere in both hemispheres. Evidence for regular long range NOx descent had already been observed in several satellite experiments [Callis et al., 1996; Randall et al., 1998, 2001; Rinsland, 1996], for example from the Halogen Occultation Experiment (HALOE) instrument on UARS, and from the ATMOS and POAM experiments. The late solar cycle 23 fell in a period where new and most capable instruments on satellites gave a wealth of new information for these processes, i.e. the instruments for the determination of atmospheric composition on the ENVISAT satellite, launched in early 2002, SCIAMACHY, GOMOS and MIPAS, the instrument ACE FTS, MLS on Aura, Odin’s SMR instrument. The new observational results offered the possibility for detailed and comprehensive model studies to test and improve our understanding of chemical and dynamical processes in the middle atmosphere as a whole. Together with the exceptional active sun, the conditions to study EPP related processes were excellent.

Since July 2002, the Michelson Interferometer for Passive Atmospheric Sounding (MIPAS) onboard the satellite ENVISAT of the European Space Agency (ESA) observes the middle atmosphere from the upper troposphere up to the mesosphere. In Fig. 15.1 a schematic view of MIPAS in space is shown, focusing on observation modes. MIPAS is a Fourier transform spectrometer for the measurement of high-resolution gaseous emission spectra at the Earth’s limb [Fischer and Oelhaf, 1996; Fischer et al., 2008]. It allows to retrieve global distributions of trace gases as for example O3, HNO3, CH4, N2O and many other substances. These observations represent one of the most complete data sets for studying the influence of energetic particle induced changes in the middle atmosphere as they cover the middle atmosphere during polar night. This is an important advantage of the MIPAS/ENVISAT dataset compared to other observations, e.g. from UV-VIS instruments or from solar occultation instruments, which both depend on solar light. Reactive and reservoir gases, and tracers for the transport in the polar winter middle atmosphere are included in the dataset. Figure 15.2 shows as an example the time series of NO2 night-time observations derived from an operational MIPAS data-product provided by ESA (see next paragraph).

Fig. 15.1
figure 1

Observation scheme of the MIPAS instrument on the ENVISAT satellite (source ESA)

Full size image
Fig. 15.2
figure 2

NO2 night-time observations of the MIPAS instrument on the European ENVISAT satellite, (a) Northern hemisphere, (b) Southern hemisphere between July 2002 and March 2004. In the NH the weak (2002/3) and strong (2003/4) intrusion from the mesosphere/lower thermosphere are seen together with the NOx enhancement after the Halloween storms in November 2003, in the SH the pronounced intrusion from the MLT in the course of the Antarctic winter can be seen (from Reddmann et al. [2010], reproduced by permission of American Geophysical Union)

Full size image

Data are available for example from the operational ESA level 2 MIPAS/ENVISAT data product. An advanced data record with more substances and also dealing with non-LTE effects, has been generated at the Institute for Meteorology and Climate Research—Atmospheric Trace Gases and Remote Sensing at KIT and the Instituto de Astrofisica de Andalucia (CSIC) (see Lacoste-Francis [2010]). By now, this advanced data record contains more than 20 substances and spans the period from July 2002 to 2010. From July 2002 to March 2004 MIPAS observed with a spectral resolution of 0.025 cm−1, afterwards the spectral resolution was reduced, but the vertical resolution improved [von Clarmann et al., 2009]. MIPAS could observe in detail the solar storms of October/November 2003 known as ‘Halloween’ storms as well as NOx intrusions in the years 2004 and 2009 in the NH, observations which gave rise to the detection of new effects related with EPP as documented in a number of publications [Jackman et al., 2005, 2008; Orsolini et al., 2005; Lopez-Puertas et al., 2005a, 2005b; Funke et al., 2005, 2008; von Clarmann et al., 2005].

In several model studies principal effects of EPP have been studied in the past (e.g. Siskind et al. [2000], see Jackman and McPeters [2004] for an overview of related work). 3D-model studies of the effects of EPPs have been performed with chemistry climate models (CCMs) applying artificial NOx enhancements [Langematz et al., 2005; Rozanov et al., 2005] or modules calculating NOx and HOx production from prescribed ionization rates (e.g. Jackman et al. [2008]). Baumgaertner et al. [2009] adapted results of Randall et al. [2007] to estimate effects of NOx produced by low energy electrons in the middle atmosphere. Many model simulations qualitatively reproduce various effects connected with EPPs, but the necessary tight connection of chemical disturbances and transport is still a challenge for current 3D-models of the middle atmosphere. Model simulations using actual meteorological conditions comparing their results with corresponding observations are therefore a valuable tool to test our understanding of the relevant processes in these events. This project focuses on such comparisons.

The first approach within the project to assess the effects caused by EPP was to use detailed observations, primarily of the MIPAS instrument on the ESA satellite ENVISAT, as a boundary condition for the additional NOx in model simulations. Two chemical transport models took part in this approach, the CLaMS and the KASIMA model. The observations of MIPAS used for the studies cover the period from July 2002 till March 2004 and include the Antarctic winter 2003 with strong NOx enhancements originating in the upper mesosphere and lower thermosphere, the strong SPE event and the following geomagnetic storms (Halloween storms) in October/November 2003, and the NOx intrusion following a stratospheric warming in Arctic mid-winter 2003/2004. From these simulations the effect on ozone chemistry could be deduced, e.g. the additional ozone loss and changes in other chemical substances could be quantified. Some results of this work are presented in Sect. 15.2, details can be found in Vogel et al. [2008] and Reddmann et al. [2010].

The second approach concentrated on the solar proton event in October/November 2003 and had therefore the direct effects of the EPP-atmosphere interaction in the focus, e.g. the amount of produced NOx and HOx, and changes of several substances in a short period during and after the ionization event. The very comprehensive observations of the MIPAS instrument of the species NO, NO2, H2O2, O3, N2O, HNO3, N2O5, HNO4, ClO, HOCl, and ClONO2, CO, CH4, and H2O allow a profound test of the chemistry implemented in models together with their transport properties in the polar winter middle atmosphere. An international model-data intercomparison project (High Energy Particle Precipitation in the Atmosphere, HEPPA) was established including both chemical transport models, and chemistry-climate models [Funke et al., 2011]; of the nine participating models, five were also involved in the CAWSES SPP, as was the derivation of ionization rates with the AIMOS model (see also Chaps. 9 and 13). The HEPPA intercomparison initiative has lately been invited to become part of the SPARC SOLARIS initiative. For these comparison, the KASIMA model was extended to include a module for NOx and HOx production by ionizing particles. Ionization rates have been calculated within the KASIMA model or using the precalculated ionization rates provided by the AIMOS calculations (Wissing and Kallenrode [2009], see also Chap. 13. A short overview of results for the KASIMA model is given in Sect. 15.3, for details see Funke et al. [2011]).

Rather few observations exist which show the effects of EPP for the HOx family [Verronen et al., 2006]. Whereas the long-term effects of EPP produced HOx are probably small, HOx related fast reactions during the particle interaction seem to be the least understood in the models. A new retrieval setup for MIPAS/ENVISAT observations was therefore developed to assess possible enhancements of H2O2 after EPP. H2O2 serves as a reservoir of HOx and it was expected that H2O2 concentrations should be enhanced during and after SPEs, and could indeed be detected by the new retrieval scheme. A climatology of H2O2 and EPP related enhancements is presented in Sect. 15.4. These data were also used as an additional species in the HEPPA intercomparison.

A strong impact on the partitioning within the NOy family has been observed when NOx intrusions reach the upper stratosphere resulting in a secondary maximum of HNO3 distribution in th upper stratosphere [Kawa et al., 1995; Stiller et al., 2005]. The proposed conversion of N2O5 to HNO3 can be explained via a mechanism involving protonized water vapor clusters. A new parameterization was developed within the project for this reaction, and it was found that this conversion may have also an impact in the lower stratosphere. The inclusion of this reaction gives better agreement between HNO3 observations and model results there (see Sect. 15.5).

Finally, the observations and model results clearly showed that not the strong but rare SP events are the main contributors to EPP related NOx in the middle atmosphere, but the seemingly more regular but weaker auroras and geomagnetic storms. Most probably, very efficient transport from the lower thermosphere is the key process for these dramatic NOx enhancements. This efficient transport seems to be related to specific dynamic situations. This finding connects the rather restricted study of chemical EPP effects to the more general question how the MLT region interacts with the lower atmosphere.

15.2 Model Studies with Imposed NOx Disturbance

Within this project two different chemical transport models were applied to study the chemical effects of NOx intrusions in the stratosphere, namely the CLaMS model and the KASIMA model. The models use a quite different model architecture and focus on different aspects. The Chemical Lagrangian Model of the Stratosphere (CLaMS) is a Lagrangian chemical transport model, which uses 3-dimensional deformations of the large-scale winds to parameterize mixing and is very well suited to study horizontal transport and mixing processes as shown in previous studies [McKenna et al., 2002a, 2002b; Konopka et al., 2007a]. KASIMA is a combination of a Euler model using analyzed winds in the lower domain, and a mechanistic model on top, and covers the vertical domain up to the lower thermosphere. It focuses on long-term transport processes including the mesospheric branch of the residual circulation and chemical processes in the middle atmosphere.

Both simulations show that the NOx intrusions can have a significant impact on the ozone budget in the middle stratosphere maximizing in the middle stratosphere where additional ozone loss between 30 %–50 % is deduced. The effect on total ozone was found to be non-negligible but restricted to high latitudes within a range of about 10 DU.

15.2.1 The Arctic Winter 2003/4: The CLaMS Perspective

To study the impact of the downward transport of NOx-rich air masses from the mesosphere into the lower stratosphere on stratospheric ozone loss, CLaMS simulations were performed for the Arctic winter 2003/04. This winter includes the Halloween storms at beginning of the winter in late October and the NOx intrusions descending in late winter from the lower thermosphere/mesosphere to the stratosphere. In late December, a stratospheric warming occurred, and low latitude air masses were transported to high latitudes.

CLaMS is based on a Lagrangian formulation of the tracer transport and, unlike Eulerian CTMs, considers an ensemble of air parcels on a time-dependent irregular grid that is transported by use of the 3d-trajectories. The irreversible part of transport, i.e. mixing, is controlled by the local horizontal strain and vertical shear rates with mixing parameters deduced from observations (e.g. Konopka et al. [2003], Grooß et al. [2005]). CLaMS therefore is especially suited to study horizontal transport processes at the boundary of the polar vortex, and possible loss of the SPE or MLT enriched air masses by the mixing processes occurring during the stratospheric warming.

The simulations cover the altitude range from 350–2000 K potential temperature (approx. 14–50 km). The horizontal and vertical transport is driven by ECMWF winds and heating/cooling rates are derived from a radiation calculation. The mixing procedure uses the mixing parameter as described in Konopka et al. [2004] with a horizontal resolution of 200 km and a vertical resolution increasing from 3 km around 350 K potential temperature to approximately 13 km around 200 K potential temperature according to the model set up Konopka et al. [2007b]. The halogen, NOx, and HOx chemistry mainly based on the current JPL evaluation Sander et al. [2006] is included.

Before the first SPE has occurred, the model is initialized once at 4 October 2003 with mainly MIPAS observations (V3O) from 3–5 Oct 2003 (CH4, CO, N2O, O3, NO, NO2, N2O5, HNO3, H2O, and ClONO2) processed at IMK and at the Instituto de Astrofisica de Andalucia (IAA), Granada, Spain. In the model, the flux of enhanced NOx from the mesosphere is implemented in form of the upper boundary conditions at 2000 K potential temperature (50 km altitude) which are updated every 24 hours. The NOy constituents NOx, N2O5, HNO3 and the tracers CH4, CO, H2O, N2O, and O3 at the upper boundary are taken from results of a long-term simulation performed with the KASIMA model (Karlsruhe Simulation Model of the middle Atmosphere), see next section. In this KASIMA simulation, enhanced NOx concentrations in the mesosphere during disturbed periods have been derived above 55 km from MIPAS observations (provided by the European Space Agency (ESA)).

In addition to this ClaMS reference model run (referred to as ‘ref’ run) a model simulation without an additional NOx-entry at the upper boundary is performed. In addition, to estimate the possible maximum impact of NOx on stratospheric ozone loss, we derive a maximum NOx-entry at the upper boundary condition from several satellite measurements of NO and NO2 by the MIPAS, HALOE, and ACE-FTS instruments. For each day at the upper boundary the NOx mixing ratios for equivalent latitudes greater or equal 60°N are replaced by the maximum NOx value observed by any satellite instruments within 60° and 90°N at 2000 K potential temperature at that day, respectively, by the maximum NO or NO2 value if no NOx measurement exists in this 24 hour period. For days where no satellite observations are available we take the maximum NOx derived by KASIMA at that day. This model simulation is referred to as ‘max NOx’ run.

Clams results of NOx enhancements and resulting ozone loss are shown for polar latitudes in Fig. 15.3. For the Arctic polar region (Equivalent Lat. >70°N) we found that enhanced NOx caused by SPEs in Oct–Nov 2003 is transported downward into the middle stratosphere to about 800 K potential temperature (≈30 km) until end of December 2003. The mesospheric NOx intrusion due to downward transport of upper atmospheric NOx produced by auroral and precipitating electrons [Funke et al., 2007] affects NOx mixing ratios down to about 700 K potential temperature (≈27 km) until March 2004. A comparison of the reference run with a simulation without an additional NOx source at the upper boundary shows that O3 mixing ratios are affected by transporting high burdens of NOx down to about 400 K (17–18 km) during the winter (see Fig. 15.3). Locally, an additional ozone loss of the order 1 ppmv is simulated for January between 850–1200 K potential temperature during the period of the strong polar vortex in the middle stratosphere.

Fig. 15.3
figure 3

Additional NOx (ΔNOx) and ozone reduction ΔO3 poleward of 70°N equivalent latitude due to mesospheric NOx intrusions over the course of the winter 2003–2004 for the reference model run (from Vogel et al. [2008])

Full size image

An intercomparison of simulated NOx and O3 mixing ratios with satellite observations by ACE-FTS and MIPAS shows that the NOx mixing ratios at the upper boundary derived from KASIMA simulations are in general too low (see Fig. 15.4). Therefore a model simulation with higher NOx mixing ratios at the upper boundary derived from satellite measurements was performed. For this model run (‘max NOx’) the simulated NOx mixing ratios and the total area of enhanced NOx mixing ratios are in general higher compared to satellite observations and so provide so an upper limit for the impact of mesospheric NOx on stratospheric ozone chemistry. The comparison between simulated and observed ozone mixing ratios confirm the ‘max NOx’ run is an upper limit which overestimates the NOx entry at the upper boundary and underestimates O3 due to stronger O3 destruction. Moreover in the ‘ref’ run the simulated O3 is in very good agreement with satellite measurements (see Fig. 15.4).

Fig. 15.4
figure 4

Horizontal view of NOx and O3 at 2000 K, 1600 K, and 800 K potential temperature for the reference model run on March 18, 2004. The isolines for the equivalent latitude at 60°N and 70°N are marked by white lines. The model results and satellite observations (IMK/IAA-MIPAS: circles and ACE-FTS: diamonds) are shown for noon time. We note that all NOx mixing ratios beyond the scale are also plotted in red (from Vogel et al. [2008])

Full size image

For the different runs ozone loss was calculated from the simulated O3 values minus the passively transported ozone. Halogen-induced ozone loss plays a minor role in the Arctic winter 2003/2004 because the lower stratospheric temperatures were unusually high. Therefore the ozone loss processes in the Arctic winter stratosphere 2003/2004 are mainly driven by NOx chemistry. Further, in addition to the transport of NOx-rich mesospheric air masses in the stratosphere due to SPEs in Oct–Nov 2003 and the downward transport of upper atmospheric NOx produced by auroral and precipitating electrons in early 2004, the ozone loss processes are also strongly affected by meridional transport of subtropical air masses. Likewise, in this case air masses rich in both ozone and NOx are transported during the major warming in December 2003 and January 2004 into the lower polar stratosphere.

Our findings show that in the Arctic polar vortex (Equivalent Lat. >70°N) the accumulated column ozone loss between 350–2000 K potential temperature (∼14–50 km) caused by the SPEs in Oct–Nov 2003 in the stratosphere is up to 3.3 DU with an upper limit of 5.5 DU until end of November (see Fig. 15.5). Further we found that about 10 DU but lower than 18 DU accumulated ozone loss additionally occurs until end of March 2004 caused by the transport of mesospheric NOx-rich air in early 2004. In the lower stratosphere (350–700 K, approx. 14–27 km) the SPEs of Oct–Nov 2003 have negligible small impact on ozone loss processes until end of November and the mesospheric NOx intrusions in early 2004 yield ozone loss about 3.5 DU, but clearly lower than 6.5 DU until end of March. Overall, the downward transport of NOx-rich air masses from the mesosphere into the stratosphere is an additional and non-negligible variability to the existing variations of the ozone loss observed in the Arctic.

Fig. 15.5
figure 5

Column ozone (in Dobson units) with (red), without (blue), and with a maximum (yellow) mesospheric NOx sources integrated over the entire simulated altitude range from 350 K until 2000 K (∼14–50 km) for equivalent latitudes poleward of 70°N. Middle panel: Ozone loss in DU for these model runs. Bottom panel: ΔO3 caused by the additional NOx-source at the upper boundary. Shown is the ozone column of the ‘no NOx’ run minus the ozone column of the ‘ref’ (red) and the ‘max NOx’ run (green) (from Vogel et al. [2008])

Full size image

15.2.2 KASIMA Studies: The Period 2002–2004

The KASIMA model is a 3D mechanistic model of the middle atmosphere which can be coupled to specific meteorological situations by using analyzed lower boundary conditions and nudging terms for vorticity, divergence and temperature. The model is based on the solution of the primitive meteorological equations in spectral formulation and uses the pressure altitude as vertical coordinate. Its vertical domain spans the upper troposphere up to the lower thermosphere. In the MLT region, winds are calculated using physical parameterizations for heating rates and gravity wave drag based on a Lindzen scheme. The model has been used for investigations of transport and chemistry in the middle atmosphere [Kouker, 1993; Reddmann et al., 1999, 2001; Ruhnke et al., 1999; Kouker et al., 1999; Khosrawi et al., 2009]. The model simulates long-term transport in the stratosphere reasonably well. This has been shown by comparisons with the inert tracer SF6 and derived mean age of air [Engel et al., 2006; Stiller et al., 2008].

The rationale of this model study was to use two model runs, one standard run, and one with a changed upper boundary condition for NOx from EPP to derive its impact on the chemical state in the middle atmosphere. No feedback of the chemistry is included to the ozone field used for the calculation of heating rates in the run, so both simulations use the same dynamics for transport and allow to single out the change esp. of ozone for the disturbed chemical conditions. With a good representation of the mean transport times, the model can therefore be used to study the effects of NOx intrusions on a longer time-scale. Against the background of the undisturbed stratosphere, the NOx enhancements in the observations and in the model can be clearly traced as seasonal events restricted to polar latitudes, leaving the rest of the stratosphere mostly unaffected. This allows to estimate the chemical effects of NOx intrusions by comparing the disturbed and the undisturbed model run, and to analyze the individual events which show different characteristics between intrusions starting in the upper mesosphere and the SPE with enhancements down to the stratosphere. Coupling effects between chemical changes due to EEPs and dynamics take place in the real atmosphere, and are therefore implicitly included in the dynamical fields used in the nudged model. In this sense, the chemical effects are also influenced by the coupling, as the reference dynamical state does not correspond to a real undisturbed one. Nevertheless, for the purpose of quantifying the strength of the NOx intrusions on chemistry this is only an effect of minor importance.

For our study we used reprocessed operational MIPAS NO2 data which had a nearly daily coverage [Ridolfi et al., 2000; Carli et al., 2004]. NOx has a photochemical lifetime of a few days in the mesosphere and during night most of NOx is in the form of NO2. Nighttime observations of NO2 are taken therefore as a representative for the concentration of NOx.

Figure 15.2 shows time-height cross sections of the mean NO2 volume mixing ratio of the northern and southern polar cap (geographical latitudes polewards of 60°) in ppb, from the ESA MIPAS/ENVISAT nighttime observations as described above. NOx values from the MIPAS/ENVISAT observations using nighttime NO2 as a proxy for NOx, are imposed on the model values during times of diagnosed increased levels of NOx in the polar mesosphere and during the SPE in October/November 2003. The ESA dataset has been preferred over other data sets based on MIPAS/ENVISAT observations as it provided a rather complete coverage for the MIPAS/ENVISAT phase I observation period, but it lacks the accuracy of the IMK-IAA data set as it does not include non-LTE effects and does not consider horizontal gradients of mixing ratios in the retrieval (see Wetzel et al. [2007]).

The sum of all inorganic nitrogen-containing species except N2 and the source gases like N2O is called NOy, enhanced after EPP [Brasseur and Solomon, 2005]. Figure 15.6 shows the global excess NOy in mol for the period July 2002 to December 2005. The auroral winter intrusions in the Arctic winter 2002/2003 and Antarctic winter 2003, the SP event in fall 2003 and the strong intrusion in the Arctic winter 2003/2004 are clearly discernible as single events adding to the total NOy background. The characteristic time-scale for the decay of the additional NOy in the model lies in the range of about 2 years.

Fig. 15.6
figure 6

Global excess NOy from energetic particle production (from Reddmann et al. [2010]), modified by permissions of the American Geophysical Union)

Full size image

We estimate the additional NOy in Arctic winter 2002/2003 at 0.4 Gigamol (GM), in Antarctic winter 2003 at about 1.4 GM, the NOy from the SPE at about 1.5 GM and in January 2004 at about 2 GM. As in the Arctic winter both intrusions overlap, they cannot clearly be separated. The value estimated for the Antarctic winter 2003 is about 75 % of the value Funke et al. [2005] deduced from their observation based study with MIPAS/ENVISAT data. Randall et al. [2007] give a range of 1 to 2 GM estimated from HALOE observations for this winter, depending on the use of average or maximum observed NOx values. Globally, the additional NOy in the model which we derive amounts at its maximum to about 5 % of the total 70 GM NOy in the middle atmosphere from the oxidation of N2O.

The effects of the additional NOy in the atmosphere on the formation of particles in the cold polar vortex (for example by changing the saturation vapor pressure) is small, but still significant. The mass of NOy in NAT-particles increases in the Antarctic winter 2004 and the Arctic winter 2004/2005 by about 5 %.

Figure 15.7 shows the effects on ozone for the whole simulation period again for 75°S and 75°N. Additional ozone loss lasts for the following year in the Southern Hemisphere at the level of about 0.2 ppm. At the end of 2004 a small ozone change can be found above the band related to the Antarctic winter 2003. Closer inspection of time-latitude cross sections of NOy at different altitudes show that during summer 2004 NOy is transported from the Northern to the Southern Hemisphere at the stratopause level (not shown) exceeding 1 ppb, causing the additional ozone loss. During the Arctic winter 2002/2003 the minor amounts of additional NOx cause only small ozone changes. The maximum changes occur during spring 2004, exceed 30 % and follow closely the downward transport of the additional NOx. The changes caused by the SPE in fall 2003 can be traced until spring 2004 below 30 km but are lost to lower latitudes. The enormous amounts of additional NOx from January cause ozone changes of several percent even in the following spring at altitudes of about 30 km.

Fig. 15.7
figure 7

Ozone change through EPP effects (from Reddmann et al. [2010], reproduced by permission of American Geophysical Union)

Full size image

The NOx induced additional ozone loss causes also changes in total ozone. In the Southern Hemisphere ozone changes are restricted to the Antarctic region. In the Northern Hemisphere a decrease of total ozone of several Dobson units reach the mid-latitudes. At high latitudes changes of about 5 DU can be found even in spring 2005.

15.3 The Solar Proton Events of Oct/Nov 2003 and Direct Effects

This section deals with the second approach where HOx and NOx enhancements are calculated in the models directly from ionization rates. This allows chemistry climate models to include EPP effects as a function of solar and geomagnetic activity. The role of CTMs is here again to validate the results of such calculations with the help of detailed observations. Within the CAWSES project the KASIMA model was expanded to include these direct effect using particle flux data from satellites as GOES. Again, the very well observed HALLOWEEN storms have been used to make these comparisons using the MIPAS data set. These comparisons were carried out in the framework of an international model measurement intercomparison initiative of CCMs and CTMs [Funke et al., 2011]. For KASIMA as for most of the other models taking part in that inter-comparison the general agreement between observations and model is reasonable but by far not perfect.

As the HOx cycle is not well represented in observations a new retrieval strategy was implemented in the MIPAS analysis to derive H2O2 concentrations in the middle atmosphere. For the first time, a H2O2 climatology could be derived and enhancements during the SPE effect in 2003 could be observed. The comparison with the model shows a qualitative but not a quantitative agreement, hinting at deficits of our understanding of the chemical processes related to H2O2 in general as well as and in particular during and after atmospheric ionization events.

15.3.1 Ionization Rates and HOx and NOx Production

Different kind of particles contribute to the ionization in the atmosphere. Enhanced fluxes of energetic protons can be included in models with reasonable accuracy by an approximate description, for example applying the Bethe-Bloch formula and assuming that the protons fill the geomagnetic polar cap more or less homogeneously, as it is the case for solar proton events. The spectral flux distribution of the protons has to be derived from measurements of particle detectors on satellites (GOES data series) which only give integrated counts for a few channels. In KASIMA we applied a particle flux of the form J(E)=A⋅E −δ, where A and δ have been derived by a fit procedure (see Versick [2011]). We also included a component from cosmic rays using the formalism of Heaps [1978]. Figure 15.8 shows ionization rates calculated as described above. It shows distinct and sporadic solar proton events in the upper stratosphere and, approaching solar minimum conditions, a slowly increasing flux of cosmic rays.

Fig. 15.8
figure 8

Ionization rate from protons and cosmic rays calculated in the KASIMA SPE module (from Versick, 2011)

Full size image

To include ionization by electrons a much more complex calculation is necessary, as electron flux is much more dependent on the state of the magnetosphere and the radiation belts. We therefore used the AIMOS data set to include ionization caused by electrons, see the corresponding Chap. 13 of J.M. Wissing and M. Kallenrode. Additional to precipitating electrons, this data-set also considers protons and α-Particles, based on measured particle fluxes. We applied both data sets for the solar proton event in Oct/Nov 2003, and found better agreement of observed and simulated NOx when applying our module. With the limitations of the observations of the particle flux and other assumptions to derive a energy flux spectrum, this underlines that more detailed observations of incoming particle fluxes are necessary.

15.3.2 Model Intercomparison

The SP event in Oct/Nov 2003 is an ideally suited testbed for simulations of direct effects in the middle atmosphere caused by energetic particles as the detailed observations of the MIPAS instrument allow to validate the response of all the most important chemical species in the models. MIPAS showed changes after the event for species such as NO, NO2, H2O2, O3, N2O, HNO3, N2O5, HNO4, ClO, HOCl, and ClONO2. The HEPPA model data intercomparison initiative brought together scientists involved in atmospheric modeling using state-of-the art CCMs and chemistry-transport models on one hand and scientists involved in the analysis and generation of MIPAS IMK/IAA data on the other hand. The objective of this community effort was (i) to assess the ability of state-of-the-art atmospheric models to reproduce SPE-induced composition changes, (ii) to identify and—if possible—remedy deficiencies in chemical schemes, and (iii) to serve as a platform for discussion between modelers and data producers. This was achieved by a quantitative comparison of observed and modeled species abundances during and after SPEs, as well as by inter-comparing the simulations performed by the different models. The initiative focused in its first phase on the inter-comparison of IMK/IAA generated MIPAS/Envisat data obtained in the aftermath of the October/November 2003 SPE with model results.

We give here a few examples of results of the intercomparison for the KASIMA model which extend the studies presented in Sect. 15.2. We refer to the extensive paper of Funke et al. [2011] for the complete inter-comparison and the discussion of the results.

An important parameter for subsequent chemical effects caused by the EPP is the amount of additional reactive nitrogen compounds which is produced in the event, expressed as additional NOy. Figure 15.9 shows the observed NOy enhancement and the results of the models directly after the SPE. The agreement seems to be satisfactory, but note the logarithmic scale. As most other models, KASIMA overestimates NOy below about 0.5 hPa and shows some underestimation above. This discrepancy hints to difficulties in describing the ionization rate, see discussion in the section above.

Fig. 15.9
figure 9

Area-weighted averages (40–90N) of observed and modeled NOy enhancements during 30 October 1 November with respect to 26 October (left) and relative deviations of modeled averages from the MIPAS observations (right). Thick solid and dashed lines represent model multi-model mean average and MIPAS observations, respectively. WACCMp denotes the WACCM simulation including proton ionization only (excluded from multi-model-mean) (from Funke et al. [2011])

Full size image

Figure 15.10 top shows the development of CO in observations and the model. Obviously KASIMA overestimates the downward transport compared to observations. A too fast descent in the Northern polar winter stratosphere was also found in the analysis discussed in Reddmann et al. [2010]. Figure 15.10 shows an example for substances from the chlorine family. Whereas HOCl changes are well represented in the model, ClONO2 changes are underestimated to a great extent which is probably related to lower ClO values in the models compared to MIPAS even during background conditions. Other differences between models and observations concern HNO3 buildup during and shortly after the event, It was shown by Verronen et al. [2008] that this is most likely due to missing ion chemistry in most of the models.

Fig. 15.10
figure 10

Examples from the HEPPA intercomparison showing KASIMA results. Top row: CO time series derived from MIPAS and from KASIMA (ppm). Time series of HOCl and ClONO2 derived from MIPAS observations and difference to KASIMA (ppb)

Full size image

15.4 Hydrogen Peroxid as an Indicator for HOx Production

Besides the production of NOx, also HOx is produced in the course of energetic particle precipitation. Changes in trace gases like HOCl, HNO4 or HNO3 are dependent on changes in the concentration of the HOx family. It is therefore worth to try to assess members of the HOx family directly by observations and to document the changes caused by the EPP. But only the gas H2O2, which acts as a reservoir gas for HOx shows sufficiently strong spectral features in the spectral range of the MIPAS instrument. Therefore, the MIPAS/ENVISAT full and reduced resolution spectra were analyzed to develop a retrieval approach for H2O2.

The main source of H2O2 is the HO2 self-reaction:

$$ \mathrm{HO}_2 + \mathrm{HO}_2 \longrightarrow \mathrm{H}_2 \mathrm{O}_2 + \mathrm{O}_2 $$
(15.1)

Of minor importance is the three-body reaction:

$$ \mathrm{OH} + \mathrm{OH} + \mathrm{M} \longrightarrow \mathrm{H}_2 \mathrm{O}_2 + \mathrm{M} $$
(15.2)

The main sink in the stratosphere is through photolysis:

$$ \mathrm{H}_2 \mathrm{O}_2 + \mathrm{h} \nu\longrightarrow2 \mathrm{OH} $$
(15.3)

The reactions with OH and atomic oxygen destroy H2O2 to a lesser extent.

15.4.1 Hydrogen Peroxide Spectral Signatures and Retrieval Set-Up

In the mid-infrared spectral region, which is covered by the spectral range of MIPAS, hydrogen peroxide shows weak emission lines between about 1210 cm−1 and 1320 cm−1 (Fig. 15.11). All these lines belong to the H2O2 ν 6 band centered at 1266 cm−1. The spectral signatures are taken from the latest update for H2O2 (based on measurements from Perrin et al. [1995] and Klee et al. [1999]) in the HITRAN 2004 molecular spectroscopic database [Rothman et al., 2005]. The challenge of the hydrogen peroxide retrieval is the very weak signal of the emission lines in comparison to the instrumental noise which is much higher in this spectral region than the H2O2 signal (Fig. 15.11).

Fig. 15.11
figure 11

Spectral range of the H2O2 ν 6 rotations-vibration band; Top: emission of all gases; Bottom: emission of H2O2 (black) and typical noise in MIPAS-spectra (red); black lines mark used spectral windows, lower row for heights below 44.5 km, top row above (from Versick [2011])

Full size image

For the retrieval, 19 narrow spectral regions (microwindows) were selected between 1220 cm−1 and 1265 cm−1, which is the lower end of MIPAS channel AB and includes the R-branch of the H2O2 ν 6 band. Those microwindows are used up to tangent altitudes of 47 km.

The main criterion for the selection was high sensitivity to hydrogen peroxide and low interference by other gases. Unfortunately the P-branch can not be used in the lower stratosphere for our retrieval because the lines of the interfering gases are too dense. Above 47 km an additional microwindow from 1285 cm−1 to 1292 cm−1 is used. The microwindows between 1237 cm−1 and 1265 cm−1 are not used in this altitude regions due to potentially bad absorption cross sections for N2O5 at low pressures.

Since the hydrogen peroxide contribution is so small, the contribution of other gases still needs to be considered and HOCl, CH4, N2O, N2O5 and COF2 have to be retrieved jointly. Other gases had to be retrieved before the hydrogen peroxide retrieval and the results had to be used in our H2O2 retrieval. These gases are: H2O, O3, ClONO2 and HNO3. For all the other gases in this spectral region climatological values were used.

The retrieval procedure follows a scheme analog to that described by Rodgers [2000]:

(15.4)

where x is the retrieval vector, K the partial derivatives of the spectral grid points with respect to the retrieval vector (Jacobian), S y the covariance matrix due to the measurement noise, R the regularization or constraint matrix, y the measurement vector, F the forward model, x a the a priori profile, and i the iteration index.

Due to the very high noise in comparison to the signal we had to choose a rather high regularization strength giving us a low number of degrees of freedom. The regularization is stronger in the upper stratosphere. It is weakest in the middle stratosphere where we expect the maximum volume mixing ratio of H2O2. Lowest measurements used were around 25 km because otherwise oscillations in our profiles occurred due to error propagation from below which caused subsequent faults in other altitudes. The retrieval setup was verified by retrieval of H2O2 from synthetic spectra. These spectra were calculated with the Karlsruhe optimized and Precise Radiative transfer Algorithm (KOPRA) [Stiller, 2000]. The error analysis showed that concentrations of H2O2 can be retrieved between 20–60 km. The corresponding vertical resolution is about 8 km in the lower stratosphere and about 35 km in the upper stratosphere. Comparison of H2O2 MIPAS measurements with models must be done by convolving the model results with the MIPAS averaging kernel.

15.4.2 Distribution of H2O2 Under Normal and Disturbed Conditions

The temporal evolution of the H2O2 stratospheric distribution in KASIMA is very similar to the temporal evolution in the MIPAS measurements (see Fig. 15.12). Both show the highest vmr in the inner tropics shortly after equinox. In the tropics and subtropics the H2O2 volume mixing ratio is following the position of the overhead sun. Higher volume mixing ratios are reached in the summer hemisphere. Also the lower volume mixing ratio in the end of 2003 and beginning of 2004 are represented by KASIMA. The absolute value of the mixing ratio vmr in KASIMA is almost twice that of MIPAS. Sensitivity studies show that this could be related to uncertainties of associated reaction rate-constants.

Fig. 15.12
figure 12

Retrieval (left) and model (right) results of H2O2: top temporal evolution of the daily zonal means at 30 km; KASIMA results are shown on MIPAS geolocations convolved with the MIPAS AK. (Bottom): Time-height cross section of H2O2 during the Halloween storms; MIPAS averaging kernels has been applied on KASIMA results (from Versick [2011])

Full size image

Figure 15.12 shows the temporal evolution of H2O2 during the Halloween storms. The retrieval clearly shows enhancements related to the ionization events. The model grossly overestimates the observed concentrations during the SPE. From the HEPPA inter-comparison similar results have been found for other models. Further details of the retrieval procedure and results can be found in Versick [2011] and Versick et al. [2011].

The reduced resolution mode of MIPAS after March 2004 makes regular observations of H2O2 more difficult, and global distributions could not be derived until now. But after the exceptional SP events in January 2005, H2O2 enhancements could also be derived in the reduced resolution mode. Together with observations of MLS instrument on the AURA satellite, a more complete characterization of the HOx family after SPE events was possible (for details see Jackman et al., 2011).

15.5 HNO3 Enhancements

In the middle stratosphere, the reactive nitrogen compounds NO and NO2 are converted to reservoir gases, of which HNO3 is the most abundant. Orsolini et al. [2005] found much higher HNO3 concentrations observed by MIPAS/ENVISAT when compared to their model for the winter 2003/2004, Stiller et al. [2005] analyzed the Antarctic winter 2003, when a distinct secondary maximum of HNO3 was found in MIPAS/ENVISAT data at about 40 km. These observations are in line with findings from earlier satellite missions [Austin et al., 1986; Kawa et al., 1995]. The latter explained this discrepancy as a result of reactions of N2O5 with water cluster ions. This reaction, originally proposed by Böhringer et al. [1983], had been combined with heterogeneous reactions on sulfate aerosols by de Zafra and Smyshlyaev [2001] in order to understand HNO3 satellite and ground based observations in polar winters.

First comparisons of the HNO3 observations with the KASIMA model also showed a pronounced underestimation of HNO3 in the late polar winter when high NOx concentrations indicate strong NOx intrusions. We therefore included the parameterization of de Zafra and Smyshlyaev [2001] in our chemistry module and got reasonable agreement with the observations (see Reddmann et al. [2010]).

The parameterization of de Zafra and Smyshlyaev [2001] uses a fixed profile for protonized water clusters. Using a formulation for the cluster concentration according to Aplin and McPheat [2005] a variant of this parameterization was developed where the concentration of water clusters is dependent on the actual ionization rate. Figure 15.13 compares versions of the KASIMA model with no additional conversion, the implementation of de Zafra and Smyshlyaev [2001] and the new approach. The conversion to HNO3 is strongest in the new version, and results in too low N2O5 concentrations compared to the MIPAS results. Interestingly, through the ionization by cosmic rays we found a pronounced effect of this reaction in the lower stratosphere, bringing the regular maximum of HNO3 at about 25 km in closer agreement to the observations.

Fig. 15.13
figure 13

N2O5 and HNO3 time-height cross sections for the Arctic winter 2003/4. From top MIPAS observations, KASIMA simulations without reaction with protonized water clusters, KASIMA including the reaction parameterized according de Zafra and Smyshlyaev [2001], and reaction with ionization rate dependent concentration of water clusters

Full size image

15.6 The Question of the Origin of NOx Intrusions

In winter 2003/4 the analysis showed that it is not the SPE which caused most of long-term impacts on ozone but the NOx enhancements transported from above about 80 km. The question where the massive NOx intrusions observed in the Arctic winter 2003/4 (see Sect. 15.2), 2005/6 and 2008/9 have its origin is currently under debate. Figure 15.14 shows output of a KASIMA simulation where the downward transport and ionization via electrons is probed. The model shows especially for the 2003/4 and 2005/6 no pronounced downward transport for CO, here used as a tracer for mesospheric air masses. This agrees with the fact that also an artificial thermospheric NOx source in the model does not contribute to massive NOx enhancements in contrast to the observations. But also the inclusion of auroral electron ionization according precalculated ionization rates through the AIMOS model does not improve this deviation from observations. Many models seem to fail to simulate the massive NOx enhancements but it is not clear for what reasons. It was suggested that due to the limited energy range and resolution of the particle counters, precipitation of highly relativistic electrons might not be detected. The fact that the massive enhancements observed in the past Arctic winters occurred after strong stratospheric warmings and accompanied by an elevated stratopause Smith et al. [2009] strongly suggests however, that dynamical processes as the propagation of gravity waves play an important role, as the propagation of gravity waves. As it is well known that gravity wave drag is represented in models only in a parameterized form, a plausible reason for the failure of the model is some deficit in the representation of processes connected to gravity waves and their deposition of energy and momentum. Further model studies and observations, specifically in the MLT region are necessary to solve this question.

Fig. 15.14
figure 14

KASIMA simulations of the NH winters 2002/3–2005/6. CO serves as an upper mesospheric tracer, the artificial thermospheric NOx probes possible transport from the lower thermosphere, and the simulation including the AIMOS electron component probes possible ionization in the mesosphere from auroral sources. The winter 2003/4 and 2006/7 with observed strong intrusion are marked

Full size image