A global historical ozone data set and signatures of El Niño and the 11-yr solar cycle

Introduction Conclusions References

spatial distribution of ozone changes on different time scales related to external forcings, a transient ozone boundary condition is essential.While in the third Climate Modelling Intercomparison Project (CMIP-3) some models still used a time-invariant ozone climatology (Miller et al., 2006), current model simulations (CMIP5) mostly use a timedependent ozone data set (Taylor et al., 2012).Since 1979, two-or three-dimensional ozone data sets are available from merging different satellite (or other) observations (e.g.Stolarski and Frith, 2006;Hassler et al., 2009Hassler et al., , 2012)).For CMIP5 simulations starting in 1850, a global monthly mean two-dimensional ozone data set was produced whose temporal variability was described using a regression approach based on stratospheric chlorine and the 11 yr solar cycle (Cionni et al., 2011).This data set was designed for coupled ocean-atmosphere simulations and consequently does not include interannual variability that arises due to non-forced climate variability.For atmosphere global climate models (AGCMs), or for data analysis, a vertically and meridionally resolved ozone data set would be desirable that is as close to observations as possible and hence also reflects interannual variability, which may arise, e.g. from changes in sea-surface temperatures (SSTs).
Here we present an approach for producing a monthly mean zonal mean vertically resolved global historical ozone data set based on chemistry climate model simulations and historical (ground-based and satellite-based) total column ozone observations using data assimilation techniques.The main aim of this paper is to discuss the approach and point to possible improvements.The product, termed HISTOZ.1.0,is presented and analysed with respect to prominent variability prior to 1979, focusing on El Ni ño/Southern Oscillation and the 11 yr solar cycle.
The paper is organised as follows.Section 2 describes the approach.Section 3 describes the data used.In Sect. 4 we report on the ancillary processing and analysis steps, starting with an assessment of the model simulations and followed by technical material related to bias corrections.Section 5 focuses on the specification of observation errors, the construction of error covariance matrices, and covariance localisation.In Sect.6 we present HISTOZ.1.0and discuss validation results in the post-1979 Introduction

Conclusions References
Tables Figures

Back Close
Full period.Some prominent variations in the global ozone field such as those due to El Ni ño/Southern Oscillation (ENSO) and the 11 yr solar cycle are analysed in Sect.7. Conclusions are drawn in Sect.8.

Ensemble Square Root Filter
The basis of our approach (sketched schematically in Fig. S1) is an off-line data assimilation procedure as described by Bhend et al. (2012).The approach aims to find an optimal state (in the model space), termed "analysis" or x a , starting from a "background" model state x b and adding a correction that depends on the observations y.
The observation error covariance R determines the weight given to each observation.
The background error covariance matrix P b specifies how deviations in each element of the model state vector are related to deviations in each other element and thus determines how the information from the observations is spread into model space.
We use the Ensemble Square Root Filter (EnSRF, Whitaker and Hamill, 2002), a variant of the Ensemble Kalman Filter (EnKF, e.g.Evensen, 2003), to determine the correction (see Bhend et al., 2012 for the following).In our application, x b , a vector of length m, denotes the monthly mean ozone at all grid points (latitudes, levels) for a given time from one simulation (one member) in an initial-condition ensemble simulation.Each member, as well as the ensemble mean, are corrected using the observations y (all available monthly means for the same time), a vector of length n.In the EnSRF, x b is decomposed into the ensemble mean x b and the deviations therefrom (x b ).Similarly, the correction is separated into an ensemble mean correction (Eq. 1) which is identical to the EnKF correction, and a correction of the anomalies from the ensemble mean Figures (Eq.2): H, a matrix of size n × m, is the Jacobian matrix of the linear observation operator that extracts the observations from the model state x (see Sect. 2.3).The Kalman gain matrix K (m × n) is identical to the gain matrix in the classical EnKF approach (Eq.3).
The gain matrix for the ensemble anomalies is expressed as follows (Eq.4): P b is the m × m background error covariance matrix (Sect.5.3), R is the n × n observation error covariance matrix (Sect.5.1).We assume that the observation errors are uncorrelated (R is diagonal), which allows a serial incorporation of each observation (i.e.R becomes a scalar).
Note that the analyses are not used as new initial states for the model.Rather, we use existing simulations in an "off-line" approach.Although we lose some of the advantages of data assimilation with this approach, we gain simplicity.The approach can be used a posteriori, i.e. the procedure can be optimised and observations or error information can be updated without the need to repeat the simulations.Since the filter will only be applied once every six months (albeit the resolution of the product is monthly, see Sect.2.2), the role of initial conditions is small and that of the boundary conditions becomes more important.Introduction

Conclusions References
Tables Figures

Back Close
Full

State vector
In most data assimilation approaches, the vector x describes the model state at a given time.In this study the product is generated at monthly resolution.However, as ozone at middle to high latitudes has strong month-to-month correlation during the summer months of the respective hemisphere (Fioletov and Shepherd, 2003), observations in a given month were allowed to affect neighbouring months in the same season by combining six consecutive months into one state vector x.In other words, x contains the monthly and zonally averaged ozone number densities from all grid cells for six months.Sudden steps due to gaps in individual observation records are expected to be smoothed and the sparse information is exploited more fully.Based on the correlation tables in Fioletov and Shepherd (2003), the 6 month seasons are defined as May to October and November to April.Correlations between the seasons are low so that this procedure is not expected to introduce large steps at the transitions of the seasons.The zonal mean ozone fields comprise 36 latitudes and 30 pressure levels (0.01, 0.05, 0.1, 0.3, 0.5, 1, 1.6, 2.5, 4, 6.3, 10, 15.9, 25.1, 39.8, 63.1, 100, 150, 200, 250, 350, 450, 550, 650, 750, 800, 850, 900, 950, 980, 1000 hPa), of which only 25 levels are used, as ozone is set to zero in the top layer and the four lowermost layers.After combining the data into six month blocks, the final state vector x has a length of 5400 and the covariance matrix P b has around 3 × 10 7 elements.

Observation operator
The function H(x) transforms the model space into the observation space: mon(x) selects the month, i.e. mon equals 1 (else 0) if the month of x is the month of the observations.The layer thickness is calculated from geopotential height (GPH) from the ensemble mean.For simplicity (and after finding only a very small dependence on interannual variability), the GPH profile from 1960 was used for all years.
As H expresses a linear combination of the elements of x, the Jacobian matrix of H, termed H, consists of the coefficients: In our implementation, H is a vector (because observations are assimilated serially) and its scalar product with x gives the integrated and interpolated ozone column for the latitude and month of the observation.
3 Data sets

Model simulations
In this study we use an initial condition ensemble of nine simulations from 1900 to 1999 that was performed with the chemistry-climate model SOCOL Version 2 (Schraner et al., 2008).The simulations are described in Fischer et al. (2008b).SOCOL is a combination of the middle atmosphere version of the ECHAM4 spectral AGCM (Manzini and McFarlane, 1998) and the chemistry-transport model MEZON (Egorova et al., 2003).It was run with a spectral truncation of T30 and 39 levels on a hybrid sigmapressure coordinate system, with the model top at 0.01 hPa.A hybrid numerical advection scheme is used for transporting chemical species (Zubov et al., 1999), with the Prather advection scheme in the vertical (Prather, 1986) and the Semi-Lagrangian scheme in the horizontal direction (Williamson and Rasch, 1989).The model participated in the CCMval validation experiments (SPARC CCMVal, 2010) and the C20C intercomparisons for AGCMs (e.g.Scaife et al., 2008).Introduction

Conclusions References
Tables Figures

Back Close
Full The model was constrained at the boundaries with monthly SSTs and sea ice from the HadISST1 data set (Rayner et al., 2003) and land-surface properties based on Hagemann (2002).Solar variability was prescribed using spectral solar irradiance data from Lean (2000).Greenhouse gases and organic chlorine and bromine containing gases were prescribed in the lowermost five layers.In addition, surface emissions of CO and NO x were prescribed (see Fischer et al., 2008b for details).Stratospheric aerosols were taken from Sato et al. (1993).The model set-up included a nudging of the Quasi-Biennial Oscillation (QBO) using a preliminary version of the QBO reconstruction by Br önnimann et al. (2007b).

Ground-based total column ozone observations
Ground-based measurements of total column ozone have been made since the 1920s (for a review see Br önnimann et al., 2003a).In 1957, the start of the International Geophysical Year (IGY), a global standardized network, employing techniques superior to what had previously been available, was established.Hence we distinguish pre-IGY and post-IGY periods in the following analysis.
For the pre-IGY period, we use all series described by Br önnimann et al. (2003a), comprising several long, homogenised, and semi-continuous series such as those from Arosa since 1926 (Staehelin et al., 1998), Oxford, 1924-1975(Vogler et al., 2007), Tromsø since the 1930s (Hansen and Svenøe, 2005), Domb ås and Oslo from the 1940s (Svendby, 2003), and Spitsbergen, 1950-1963(Vogler et al., 2006), as well as several shorter series.The latter were re-evaluated as described in Br önnimann et al. (2003b).Those historical data series qualified as "poor" in Br önnimann et al. (2003b) were excluded.In addition, measurements from Gulmarg and Srinagar were excluded due to their poor seasonal coverage.Despite their rather low quality, total column ozone data derived from measurements by the Smithsonian Institution at Table Mountain, California (Br önnimann, 2005)  important source of information for the 1920s and 1930s.For the post-IGY period, data were obtained from the World Ozone and Ultraviolet Radiation Data Centre (WOUDC).We chose measurement series with at least 5 complete years prior to 1979 and few instrument changes.In all, total column ozone series at 57 locations were used (see Fig. 1, Table S1).All selected series were also used after 1978 for testing and validation purposes.
Monthly means were calculated as the average of the available observations if at least ten daily values were available.Moreover, at least eight monthly means must be available per sequence (i.e. per series obtained with the same instrument type and observation mode, see Table S1), otherwise the series were discarded.This criterion was necessary to retain enough degrees of freedom for the bias adjustment.Figure 2 gives the number of monthly mean values.The first observations became available in 1925.Until about 1950, measurements from around five sites are available in each month.This number then increases to around ten between 1950 and 1957 and to around 20 to 35 from the IGY onwards.

Backscatter Ultraviolet (BUV) observations
Total column ozone was measured from space starting in 1970 with the Backscatter Ultraviolet (BUV) instrument onboard Nimbus-4 (Heath et al., 1973;Stolarski et al., 1997).Due to battery failure, coverage decreased after the first two years, but data were retrieved until 1976.The data quality is inferior to later missions and the data have only rarely been used in publications.Here we use reprocessed BUV data (Stolarski and Frith, 2006).Comparisons with ground-based total column ozone data were performed to check their quality (see Sect. 5.1).Here, we make use of monthly zonal mean total column ozone in 5 • latitude bins.Figure 2 shows the number of available monthly means from BUV.Note that whenever a zonal mean value from BUV was available, no ground-based data from the corresponding 5 • latitude belt were assimilated (but all ground-based data were used for the pre-processing steps described in Sect.4).Introduction

Conclusions References
Tables Figures

Back Close
Full

Auxiliary observation-based data
Additional data sets were used in various pre-processing steps.With respect to ozone, we used the monthly BDBP vertically resolved ozone data set (Bodeker et al., 2012(Bodeker et al., ), 1979(Bodeker et al., -2007, as a reference.In addition, Version 8 TOMS satellite total column ozone data were used for validation and, in some cases (see Table 1), for calibrating the correction function for the zonal mean adjustment (see Sect. 4.3).Note that no TOMS total column ozone data were assimilated.
For estimating zonal mean total column ozone from a station, we used 200 hPa GPH data from ERA-40 from 1957 to 2002 (Uppala et al., 2005) in a regression framework (see Sect. 4.3).ERA-40 was supplemented back in time by statistically reconstructed upper-level fields from Griesser et al. (2010) which are based on historical upper-air and surface data.

Model assessment
As only total column ozone is assimilated, it is vital that the CCM simulates a realistic vertical distribution of ozone and its covariance structure.We therefore assessed the zonal mean ozone variability in the simulations and specifically the effect of external forcings (for further validation studies see SPARC CCMVal, 2010;Fischer et al., 2008b;Scaife et al., 2008).
A multiple linear regression model was applied to the ensemble mean.As our main focus was to reproduce the effects of dynamical and radiative processes in the northern winter polar stratosphere, the regression model was applied to zonally averaged variables over the winter months, i.e.January to March.The predictor variables, which were computed from the model boundary conditions, comprise two quasi-orthogonal QBO time series (zonal mean equatorial winds at 10 hPa and 30 hPa, the latter with Introduction

Conclusions References
Tables Figures

Back Close
Full a lag of 2 months), total solar irradiance, stratospheric aerosol optical depth in the layer between 15 and 35 km altitude, greenhouse gas concentrations, and equivalent effective stratospheric chlorine.From the SST fields, indices for El Ni ño (NINO3.4index with one month lead time), the Pacific Decadal Oscillation (PDO, Mantua et al., 1997), and the Atlantic Multidecadal Oscillation (AMO, Enfield et al., 2001) were computed as additional predictor variables.The regression-based validation was applied to different simulated variables of SOCOL, but for brevity's sake we only discuss the outcome for ozone.The patterns of ozone changes associated with each of the regression model basis functions (predictor variables) are shown in Fig. 3.The well-documented relation of ozone to the QBO (e.g.Baldwin et al., 2001) with a dipole-structure over the equatorial stratosphere and decreases in ozone over the whole Northern Hemisphere is well captured in SOCOL.This is also the case for the response to solar variability that exhibits strong increases in ozone in the tropical upper stratosphere with an increase in solar activity, in line with Hood and Soukharev (2012).Regarding stratospheric aerosol loading, ozone shows a positive response to increased aerosols over much of the northern hemispheric stratosphere.While this may be surprising at first sight, this behaviour can be understood by the fact that the stratospheric aerosol facilitates heterogeneous activation of halogen containing species while suppressing the reactive nitrogen oxides.Therefore, different catalytic chemical cycles are dominant depending on the stratospheric burden of ozone depleting substances (Tie and Brasseur, 1995).This becomes relevant when studying the full 20th century with a number of large volcanic eruptions in the first decades of the century with no anthropogenic halogens in the stratosphere.In this case ozone increase caused by the deactivation of reactive nitrogen can dominate.As expected, ozone shows a strong anti-correlation with effective stratospheric chlorine, over the whole Northern Hemisphere.Finally, the response to ENSO can be best described by a change in the strength of the Brewer-Dobson circulation (BDC) (e.g.Fischer et al., 2008a;Hood and Soukharev, 2012) that leads to a ozone decrease in the tropical lower stratosphere and strengthened transport of equatorial ozone towards to the North Pole.Introduction

Conclusions References
Tables Figures

Back Close
Full These results show that the model captures well the effect of the boundary conditions on ozone.As a result, we expect that some skill over and above that available from the climatology will result from use of the CCM alone, without yet incorporating observations.

Debiasing of model data
Ozone in SOCOL has a seasonally varying bias in its climatology relative to the BDBP validation data set.We removed the bias by means of a regression model that was fitted in the overlapping period 1979-1999 and applied to the preceding period (the bias consists of an ozone-invariant part and one part that is dependent on the modelled ozone amount).The results of this step are shown in Fig. 4 for two levels in three different zonally averaged regions.In general, the debiasing works well and brings the time series closer to the observations.There is, in general, good agreement with respect to the interannual variability (the agreement is solely due to boundary conditions).Yet, in some regions and levels, the trends and/or seasonality do not agree well.

Adjustment of ground-based total ozone to zonal means
The ground based total column ozone data represent ozone at a given location, whereas the assimilation requires zonal averages.Because station data are too sparse to form averages over latitude bands, the data from each station were adjusted to represent the zonal mean total ozone of the corresponding latitude.We assumed that the deviation of total ozone at a given location from its mean seasonal cycle has a largescale, zonal mean contribution which arises, for instance, from changes in the BDC, and a zonally non-uniform change that results from an ozone redistribution in the lowermost stratosphere.To account for the latter, we assumed that the deviation N * of total ozone from its zonal mean month, we fitted the regression model: from which [N] can be estimated as N − c 2 Z * .To test the approach, the algorithm was applied to TOMS total column ozone and ERA-40 200 hPa GPH data for the period 1979-2000.Figure 5 shows the results for two selected months, February 1999 and August 1980.Although the zonally asymmetric features are weakened (in some cases considerably), part of the zonal asymmetry remains.This remaining structure must be accounted for in the calculation of the observation error (see Sect. 5.1).

Debiasing of total column ozone data
Recall that the model data were debiased with respect to BDBP in the overlapping period.A similar debiasing is necessary for the total column ozone observations (both ground-based and BUV satellite data).We applied H to the debiased model data x b and calculated the difference from observations, i.e. y−H(x b ) (BDBP could not be used instead of x b as many observation series do not overlap with BDBP).To this difference we fitted in each sequence (i.e. each part of a series that was obtained with the same instrument type and observation mode, see Table 1) the first harmonic of the seasonal cycle using least squares regression and subtracted it from the observations.All processing steps are summarized, using Rome as an example site, in from 0.6-0.7 to 0.65-0.8,but differences remain.BUV shows an outlier (winter 1975), which is related to an observation error (see Sect. 5.2.).

Observation error covariance
R is a diagonal matrix that contains the observation error variances.We assume that the error variance consists of several sources of errors which can be quantified individually and which are independent from each other: Here σ obs zenith is the error of a zenith observation, which is mainly an instrumental error, σ m is the error that depends on the air mass and hence on the atmosphere, and σ zonal is the error due to insufficient adjustment to zonal means.The error σ obs zenith is estimated from the corresponding literature (Tab.S1).For instance, the quality statements for "precision" in Br önnimann et al. (2003b) were translated in the following way: 4 DU for "excellent", 8 DU for "very good", 12 DU for "good" and 16 DU for "fair" (stations rated as "poor" were not considered at all).The Table Mountain data from the Smithsonian Institution received an error of 20 DU.We proceeded similarly for other series, attributing 4 DU to the homogenized long-term series and 6 DU to all other post-1957 data (see Table 1).This is broadly consistent with the literature (M érgie, 1989; Br önnimann et al., 2003b), although most references do not specifically refer to monthly means.
Comparisons between BUV and ground-based total column ozone data were performed in the late 1970s and early 1980s, but focusing on the accuracy rather than the precision (M érgie, 1989;WMO, 1980WMO, , 1983)).Differences were found to lie in the range of 17 DU or larger than 3 %.weighted within a 4 • × 4 • square around these stations from the same day (R.Bleisch, unpublished report).A regression approach was performed incorporating a linear trend as well as dependencies on total ozone, air mass, and reflectivity (from BUV data) or cloud cover (estimated from ground based stations).Results showed that by far the most important contributor to the variations in differences is air mass.Smaller influences were found for the trend and (in the case of Oxford) for total column ozone.The standard deviation of the (daily) residuals was between 10 and 15 DU.Based on these results we set σ obs zenith for monthly means to 10 DU for BUV data.Given our confidence in the Arosa series (Staehelin et al., 1998), we attribute most of the air mass dependent error to the BUV data and assume a 2 % error per unit air mass for BUV and a 0.5 % error for ground-based data (assuming local noon for all observations).For determining σ zonal we analysed the zonal standard deviation of the adjusted TOMS data (see Sect. 4.3).Smoothing the dependence on latitude and on the seasonal cycle resulted in an estimation of σ zonal (Fig. S2).Note that to this point errors were considered to be uncorrelated.This is, however, not the case for σ zonal , which is similar for two stations close to each other.Rather than letting R become non-diagonal, we accounted for this problem by multiplying σ zonal for each station with a weight that depends on the distances D i to all stations i , including the station itself.These distances were first transformed using an exponential function (see Eq. 9) with a length scale L of 1000 km (representing a typical decorrelation scale for upper-level variables, Griesser et al., 2010) and then summed over all stations in all latitude bands.Hence, R is then a diagonal matrix with r for each observation in the diagonal:

Assessment of errors
The Nashville, Tateno, and corresponding BUV data).Assuming a Gaussian distribution of errors and no bias, a fraction of 5 % is expected.We find fractions of 7 %, 5 %, 19 %, and 8 %, respectively.The high fraction for Tateno could be due to a remaining seasonally varying bias.The assimilation procedure requires that |y − H(x b )| < 3 (σ 2 obs error + σ 2 ensemble ).As a consequence, the erroneous BUV data are not assimilated and the fractions are further reduced.Hence, our error estimates that are based on metadata and independent analyses are broadly consistent with the differences, but further improvements on the side of the observations (debiasing, homogenisation, outlier screening) might be beneficial.

Background error covariance
The background covariance matrix P b was estimated from the ensemble covariance matrix.Unfortunately, our ensemble is small (nine members).Spurious correlations may occur far off the diagonal which may affect the results.A localisation of the matrix (or similar adjustment) is therefore necessary.The covariance matrix was localised in latitude and time, leaving the altitude dimension unaltered.The localisation followed the principle of distance weighting as above (see also Bhend et al., 2012).Ozone is strongly affected by the BDC, which shows a distinct, seasonally dependent, latitudinal structure.Therefore the distance weighting was altered to allow long-range correlations during times and regions when the BDC is active.We accounted for this by weighting the distance with a function a in the following way: and the resulting matrix of scaling factors are shown in Fig. S2.Localisation was also applied in time, setting L to 3 months (consistent with Fioletov and Shepherd, 2003).

The HISTOZ.1.0 data
The assimilation corrects the zonally averaged background ozone field in such a way as to best match all adjusted total column ozone observations.Corrections are usually largest at the locations of observations and at altitudes of high ozone variability.However, large corrections sometimes also occur in polar regions, even though little information is available (see Fig. S3 for four example months).In addition, corrections of opposite sign are sometimes found equatorward of the assimilated information, pointing to the important role of P b .
Figure 6 (bottom) shows results for the zonally adjusted, debiased total column ozone average from Rome, Nashville, Tateno and the corresponding series from x a and x b (and x a and x b ).It is apparent that x a is much closer to the observations than x b .The same holds for x a and x b .Correlations with observations increase from 0.69 to 0.90.The ensemble spread also decreases considerably.The observations very often lie outside the ensemble spread and hence the model might be overconfident.However, the observation error is not accounted for in this figure (assuming uncorrelated errors, the latter is around 2.5 DU).No sudden jumps in the corrections are found during the time of season transition.ditions.In the last period, i.e. from 1979 onward the agreement between HISTOZ.1.0and BDBP is very good for the tropical and subtropical regions, but differences are found over the polar regions, especially Antarctica, where HISTOZ.1.0shows lower minimum values than BDBP.

Validation using quasi-independent data
In the following, the agreement between HISTOZ.1.0and quasi-independent data sets is assessed more quantitatively.We use the reduction of error RE (see Cook et al., 1994) as a measure of skill.RE measures the squared differences between the candidate data set and "truth" (a validation data set) and compares this statistic with a "no knowledge prediction" h 0 .For h 0 , RE is 0, positive values thus indicate improvement over the no knowledge prediction.In our application, we use the debiased model background (x b is used for evaluating x a , x b for x a ) as well as its 1979-1999 climatology as h 0 .The former indicates the skill added by the assimilation while the latter indicates the total skill (model plus assimilation).Note that x b is rather smooth and that RE tends to punish the addition of variance.
As "truth" we use TOMS for total column ozone and BDBP for the vertical ozone distribution.The analysis was performed for the period 1979-1999.During this period, total column ozone observations from around 30 stations were assimilated, which is typical for the post-1957 period.Note that the validation period contains two volcanic eruptions, anthropogenically driven ozone depletion, and strong trends in atmospheric circulation (e.g.positive trends in the North Atlantic Oscillation) that are generally not well depicted by climate models (e.g.Scaife et al., 2008).
The validation against TOMS zonal mean total column ozone shows mostly positive RE values (Fig. 8), indicating that that HISTOZ.1.0 is mostly closer to TOMS than 0.2-0.5 in the extratropics.This skill results entirely from assimilating observations.The skill for the ensemble mean is mostly higher than the skill for the individual ensemble members, but the latitudinal structure is very similar.The validation of the vertical structure using BDBP as "truth" (here we only analyse the ensemble mean) shows much lower RE values (Fig. 9).There is skill relative to climatology, but mostly only on a seasonal mean basis.This suggests that most of the skill comes from the CCM simulations, which show stronger responses to boundary conditions on seasonal-to-interannual than on month-to-month time scales.Relative to x b there is hardly any skill.Although the assimilation correctly increases or decreases the amount of ozone in the column for a given latitude, it may put the difference at a wrong altitude, thus decreasing RE.The fact that the covariance matrix was not localised in the vertical dimension may contribute to this result.Degrading the spatio-temporal resolution (seasonal mean, 10 • latitude bins, 5 altitude levels) before calculating RE leads to higher skill (Fig. 9).In all, the results show that HISTOZ.1.0has considerable skill for total column ozone or for the seasonally averaged vertical distribution on a coarse grid.There is however no skill at the full vertical resolution on a month-to-month scale.As only total column ozone is assimilated, skill in the vertical structure (with respect to x b ) can only come through the background error covariance matrix P b , pointing to the need to further improve the specification of this matrix.Another reason for low skill is that x b (used as h 0 ) was debiased with respect to the BDBP using the same time period as for the validation (unlike the debiasing for y) with the result that its error is small by construction.The skill added by the observations is largest in the lower stratosphere of the midlatitudes where dynamically induced ozone changes are large and observations are available.

Analyses
In this section two analyses of HISTOZ.1.0are presented.They concern the effects of the El Ni ño/Southern Oscillation (Sect.7.1) and of the 11-yr solar cycle (Sect.7.2.) on Introduction

Conclusions References
Tables Figures

Back Close
Full ozone.There is ample literature for both effects, but based primarily on the last 30 yr, which unfortunately also carry other strong external signals (volcanic eruptions, ozone depletion, greenhouse gases).HISTOZ.1.0provides an opportunity to analyse these effects in an independent period that is less disturbed by other forcings, but it does not allow separating the contributions of model and observations to the results.

El Ni ño/Southern Oscillation (ENSO)
It has been suggested that ENSO affects the distribution of ozone in the stratosphere by changing the strength of the BDC (Sassi et al., 2004;Manzini et al., 2004;Br önnimann et al., 2004).The SOCOL model (Fischer et al., 2008a), as well as the simulations used here (Fischer et al., 2008b), have been assessed for the effects of ENSO on ozone (see also Fig. 3).Here we focus on one particular, very strong event, namely the ENSO cycle 1939-1944 with the strong and long-lasting El Ni ño event in 1940-1942(Br önnimann et al., 2004)).We contrast the mean ozone values for El Ni ño event 1940-1942 with the averages of the years 1939, 1943 and 1944, during which La Ni ña was present.The model simulations correctly simulate an increase in ozone in the extratropics in boreal winter and a decrease in the tropics, but the differences vanish when analysing calendar year averages (Fig. 10, upper left).In contrast, the observations show an increase in the annual means (Br önnimann et al., 2004).
In HISTOZ.1.0,after assimilating the observations (Fig. 10b), the ENSO signature is more pronounced.The difference is shown in Fig. 10c, again indicating the locations of stations whose data were assimilated.As expected, ozone is elevated at latitudes where observations are available.Interestingly, an even more pronounced signal is found for the polar regions.Furthermore, ozone decreases in the tropics, although no observations are available from that region.This is consistent with a strengthening of the BDC, showing that the correction resulting from the assimilation is physically meaningful.
Most previous studies on the ozone response to ENSO have focused on the period after 1979.However, two of the El Ni ño events during this period where concurrent 7787 Introduction

Conclusions References
Tables Figures

Back Close
Full with volcanic eruptions, which may obscure a clean attribution to ENSO, and not many other strong events have been observed.Using HISTOZ.1.0we analysed the average ozone fields in January-March and composited all strong El Ni ño winters after 1934 (to ensure a minimum number of assimilated observations) minus strong La Ni ña winters using the list given in Br önnimann et al. (2007a) (see Table 1).The difference (Fig. 10d) shows a very prominent ozone signature with an increase in the northern extratropics and a decrease in the tropics (relative deviations are shown in Fig. S4).This pattern, reflecting an increase in strength of the BDC, is consistent with the literature (see Sassi et al., 2004;Manzini et al., 2004;Br önnimann et al., 2004;Fischer et al., 2008b).The composite of three El Ni ño and three La Ni ña events after 1979 from BDBP (Fig. 10e) shows a signal that is much less clear and even shows a decrease in northern high latitudes.

The 11-yr sunspot cycle
Analyses have also been performed with respect to the effect of the 11 yr solar cycle.Rather than reconstructed total solar irradiance from Lean (2000), which is only annually resolved, we used the monthly International Sunspot Number record, available from the Royal Observatory of Belgium's website (http://sidc.oma.be/sunspot-data/SIDCpub.php) as the proxy for solar activity.After smoothing the series with a 24 month moving average, we chose periods of 24 consecutive months of maximum and minimum activity in the 11 yr sunspot cycle (Table 2), subdivided into pre-1979 (4 cycles, analysed in HISTOZ.1.0)and post-1979 periods (3 cycles, BDBP).Note that using Lean (2000) total solar irradiance, the same maxima and minima would be found within the uncertainty of the temporal resolution.
The post-1979 period shows an ozone increase throughout the stratosphere with increasing solar activity, strongest in the lower stratosphere (in number density) or upper stratosphere (in mixing ratio, see also Fig. 3).The average difference in x b in the pre-1979 period (Fig. 11) shows a similar pattern, although the increase in the Northern Hemisphere is no longer uniform.Compared to x b , the signature in HISTOZ.1.0 is 7788 Introduction

Conclusions References
Tables Figures

Back Close
Full slightly weaker (relative deviations are shown in Fig. S4).In all, the new results confirm the findings from previous studies based on the post-1979 period.Note that the role of the background fields and the observations cannot be distinguished in this approach.

Conclusions and outlook
In this paper we have produced, validated, and analysed a two-dimensional ozone data set from 1900-2008 termed HISTOZ.1.0that can be used as a boundary condition for climate model simulations.The data set is based on chemistry-climate model simulations up to 1925 and an off-line Ensemble Kalman Filter approach, combining the simulations with historical total column ozone data afterwards.Independent validation suggests that the data set has high skill relative to a climatology.Relative to the model background, additional skill is only found for total column ozone or for the seasonal mean ozone distribution on a coarse resolution, while no additional skill is found at the full temporal and spatial resolution.HISTOZ.1.0supports analyses of the effects of ENSO and the 11 yr solar cycle on ozone in periods that have not yet been previously studied.Importantly, the 1930s to 1970s period studied here, unlike the satellite period that is studied in most other publications, is much less strongly affected by volcanic eruptions, ozone depletion, and greenhouse gas emissions.Our results largely confirm previous studies.In the case of ENSO, the signature is more robust.A clear strengthening of the BDC during El Ni ño winters relative to La Ni ña winters is found.High solar irradiance elevates ozone throughout the stratosphere, again consistent with previous studies.
Several shortcomings were identified and improvements could be explored for future versions of the data set, comprising the number of ensemble members, and the quality of the simulations themselves, as well as details in the set-up of the assimilation such as the debiasing of observations, the localisation of the background error covariance matrix, or the incorporation of a non-diagonal R matrix.Using the same general approach (outlined by Bhend et al., 2012), it might also be possible to assimi-Introduction

Conclusions References
Tables Figures

Back Close
Full  Full affects the radiation budget of the atmosphere and therefore needs to be incorporated in climate model simulations.Models usually require a twodimensional (latitude, pressure) ozone distribution as a boundary condition.As the Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | ) where ϕ(x) is the latitude of element x of x, ∆ϕ is the grid spacing, and ϕ(y) is the latitude of the observation y (i.e. the term max(0, 1 − |ϕ(x) − ϕ(y)|/∆ϕ) provides the weights of a linear interpolation), ∆z(x) is the layer thickness and the constant c = 2.6868 × 10 16 molec cm −2 converts the result to Dobson Units (DU).The function Introduction Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | were used because they constitute an Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | [N] is related to the deviation Z * of 200 hPa GPH from its zonal mean [Z].Previous work has shown the strong relation between midlatitude total ozone and 200 hPa GPH (Br önnimann et al., 2000) in all seasons.For each calendar Figures Back Close Full Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Fig. 6 (top, note that November-April averages are shown here).The zonal correction (dashed) removes some of the variance of the original series (solid).The debiasing (dotted) does not change the variance but brings the curve closer to H(x b ) (blue).The middle panel shows debiased observations (solid) as well as H(x b ) (dashed, November-April averages) for three sites (Rome, Nashville, Tateno) and one BUV series (35 • -40 • N).All series represent similar latitudes.As a consequence, H(x b ) is quite similar for all sites.The zonal correction should bring the three observation-based series (which are at very different longitudes) closer together.Indeed, the mutual correlations increase Introduction Discussion Paper | Discussion Paper | Discussion Paper | For estimating σ obs zenith and σ m for BUV we compared daily ground-based total column ozone data from Arosa and Oxford with BUV total column ozone data distance-Discussion Paper | Discussion Paper | Discussion Paper | consistency of errors was assessed monthly by analysing the fraction of differences |y − Hx b | larger than 2 (σ 2 obs error + σ 2 ensemble ) for the series in Fig. 6 (Rome, Discussion Paper | Discussion Paper | Discussion Paper | i ,j represent scaling factors applied to the covariances between elements i and j of x b .D in this case is the latitudinal distance (in degrees; L 1 is set to 20 • , L 2 is set to 40 • ).The function a depends on the season, switching the sign of the latitude axis between boreal winter (November-April) and austral winter (May-October).Function a Discussion Paper | Discussion Paper | Discussion Paper | The HISTOZ.1.0ensemble mean at 25.1 hPa is compared to the data sets by Cionni et al. (2011) (interpolated to 25.1 hPa) and BDBP (from 1979 onward) (Fig. 7) in the form of a latitude time cross-sections.As expected, HISTOZ.1.0shows clear interannual variability which is lacking in the Cionni et al. (2011) data by construction.The interannual variability in HISTOZ.1.0 is already seen in the first subperiod plotted (1901-1926, i.e. before observations are assimilated), generated by the model boundary con-Discussion Paper | Discussion Paper | Discussion Paper | x b .With respect to the 1979-1999 climatology of x b , RE reaches values between 0.3 and 0.6, indicating the total skill of the product.The high skill in the equatorial zone is arguably due to the well-modelled effect of the QBO on ozone.Using x b as h 0 , values are close to 0 in the tropics (where almost no station data are assimilated) and around 7785 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | late upper-air information to better constrain ozone or to generate 3-dimensional ozone distributions.Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper |

Fig. 1 .
Fig. 1.Map showing the locations of the ground-based total column ozone stations used in this study.

Fig. 2 .
Fig. 2. Number of monthly mean values assimilated in the assimilation period (1925-1978) and validation period (1979-1999).Ground-based and BUV data are shown in light and dark grey shading, respectively.

Figure 3 .
Figure 3. Coefficients of a multiple regression model with zonal mean ozone mixing ratios of the ensemble mean as predictand and various predictor variables.Shaded areas, together with black contour line, mark statistically significant areas (p-value < 0.05).For comparability reasons the regression-coefficients are displayed as ppbv change with respect to one standard deviation of the predictor.Due to the long-term trend in EESC, however, this coefficient is plotted with respect to 100 pptv.

Fig. 3 .
Fig. 3. Coefficients of a multiple regression model with zonal mean ozone mixing ratios of the ensemble mean as predictand and various predictor variables.Shaded areas, together with black contour line, mark statistically significant areas (p-value < 0.05).For comparability reasons the regression-coefficients are displayed as ppbv change with respect to one standard deviation of the predictor.Due to the long-term trend in EESC, however, this coefficient is plotted with respect to 100 pptv.

Fig. 5 .Fig. 6 .
Fig. 5. Comparison of raw and zonally adjusted TOMS total column ozone data for February 1999 and August 1980.

Table 2 .
Definition of periods of solar maxima and minima.Dec 1936 to Nov 1938 May 1979 to Apr 1981 Apr 1943 to Mar 1945 Apr 1985 to Mar 1987  Apr 1947 to Mar 1949 Dez 1988 to Nov 1990 Feb 1953 to Jan 1955 Aug 1995 to Jul 1997  Apr 1957 to Mar 1959 Feb 2000 to Jan 2002 Dec 1963 to Nov 1965 Sep 2007 to Aug 2009    * Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper |