Ozone Is An Important Constituent

Published Date: 02 Nov 2017

ultraviolet radiation from the sun, and protects the biosphere from harmful effects of ultraviolet radiation.

Ozone column amounts in the atmosphere can be obtained from surface measurements and satellite

observations (e.g., Harris et al. 1997, Chipperfield and Fioletov 2007), and can be calculated by

chemistry-climate and chemistry-transport models (e.g., Eyring et al. 2006, Stolarski et al. 2006, Søvde

et al. 2008). The ability of models to reproduce the observed atmosphere comes from the key physical

and chemical processes included in the models. Today models are being improved to include comprehensive

chemistry and physics of both the troposphere and the stratosphere, as has been done for the

Oslo chemical transport model (CTM2). The updated version with improved microphysics and heterogeneous

chemistry, and the extension of vertical layers to 60 has improved the capability to predict the

distribution of ozone and precursors in the UTLS region, in the upper stratospheric region and in the

troposphere (Søvde et al. 2008).

Eleftheratos et al. (2011) provided additional evidence of improved total ozone columns by the updated

Oslo CTM2 model, by comparing monthly mean, seasonal mean and annual mean total ozone for the

period 2001-2007 with respective total ozone averages from satellite retrievals. Here, we extend the period

of comparison by including simulations for the period 1998 to 2009 that includes several QBO cycles

and examine whether ozone variations from improved Oslo CTM2 model simulations reproduce

the well-known perturbation (QBO). Then we compare our results with respective SBUV satellite retrievals

to test the consistency of the modelled ozone variations.

2 Data and Methodology

2.1 Oslo CTM2 model

The Oslo CTM2 is a global off-line chemical transport model, driven by meteorological data from the

European Centre for Medium-Range Weather Forecasts (ECMWF) Integrated Forecast System (IFS)

model. In the IFS forecasts a spectral resolution of T319 is applied (T319 is approximately 0.5 x 0.5

degrees grid resolution, longitude/latitude, with 60 vertical layers). The horizontal resolution of the Oslo

CTM2 can be varied between T21 (resolution of 5.6o x 5.6o, longitude/latitude), T42 (2.8o x 2.8o),

T63 (1.9o x 1.9o) and 1o x 1o, into which the IFS spectral fields are truncated. The IFS data, available as

gridded data, are averaged into the model grid. Here the T42 (2.8o x 2.8o) horizontal resolution of Oslo

CTM2 was used to calculate averages over 10 degree latitude zones.

The Oslo CTM2 has previously been applied in model/model comparisons and tested against observations

(e.g., Isaksen et al. 1990, Gauss et al. 2003, Isaksen et al. 2005, Andresen et al. 2006). The global

chemical transport model Oslo CTM2 has been evaluated against measurements by satellite-based instruments,

ozone sondes and aircraft (Søvde et al. 2008). All reactions and species in the Oslo CTM2

are described in detail in the study by Søvde et al. (2008). Recently, the model was used to investigate

the observed record ozone decline over the Arctic during wither/spring 2011 (Balis et al. 2011, Isaksen

et al. 2012).

2.2 SBUV satellite data

The ozone satellite data used in this study come from the SBUV (Version 8.6) merged total and profile

ozone data sets for the period 1998 to 2009. The SBUV merged ozone data sets are monthly-mean zonal

and gridded average products constructed by merging individual SBUV/SBUV/2 (total and profile

ozone) satellite data sets. The data are available at various altitude layers and the total column in Dobson

Units (DU) at the webpage http://acdb-ext.gsfc.nasa.gov/Data_services/merged/. Details can be

found in the studies of Bhartia et al. (2004), McPeters et al. (2011). Results of SBUV/2 ozone profile

comparisons with other data sources are discussed by Petropavlovskikh et al. (2005), Nazaryan and

McCormick (2005), Fioletov et al. (2006), Terao and Logan (2007).

3 Results

Fig. 1 shows the time series of zonally averaged monthly mean total ozone from Oslo CTM2 for the

period 1998-2009 in comparison to respective satellite-derived ozone retrievals. The average percentage

differences between the two data sets have been calculated from the following formula:

[(model mean – satellite mean) / model mean] x 100 %

From Fig. 1 it is evident that there are specific differences between the model and satellite-derived

monthly mean total ozone. On average, the model generally underestimates total ozone over northern

and southern mid-latitudes by 5.6% and 4.7%, respectively. On the other hand, total ozone from the

model is overestimated in the tropics by 1.9%. In general it appears that differences in columnar

amounts are small in the tropics and medium over mid-latitudes but correlations are high.

Total ozone data from Oslo CTM2 have been correlated with SBUV satellite data using linear regression

analysis. Model and satellite ozone columns were used to compute monthly zonal means per 10

degree latitude zones. The time series obtained have been prewhitened to remove the annual cycle by

subtracting from each month the long-term monthly mean of the twelve year period of record. The correlation

analysis of model calculations with satellite retrievals was performed using the obtained deseasonalized

time series for the period 1998-2009. Fig. 2 shows the total ozone anomalies (in per cent of

the mean) from Oslo CTM2 model calculations and SBUV satellite retrievals at different latitudes

zones (60–70 oN, 50–60 oN, 40–50 oN, 30–40 oN, 20–30 oN, 10–20 oN, 0–10 oN, the equator, 10–20 oS,

20–30 oS, 30–40 oS, 40–50 oS, 50–60 oS and 60–70 oS). The correlation coefficients between the two

data sets at these latitude zones are summarized in Table 1.

As can be seen from Fig. 2, there is very good agreement between the Oslo CTM2 model and SBUV

satellite ozone anomalies throughout the whole period of record. The correlation coefficients are highly

significant at all latitude zones (Table 1). The dotted line in the middle of Fig. 2 shows the variations in

the zonally averaged winds at 30 hPa taken from over the equator, as index of the QBO. The QBO index

was obtained from the Climate Prediction Centre of NOAA at

http://www.cpc.ncep.noaa.gov/data/indices/. The general features of the QBO in total ozone has been

examined in a several studies dating back to 1964 (e.g. Zerefos, 1983 and references therein). They include

a QBO in total ozone at the equator (between 5 oN and 5 oS) which is nearly in-phase with the

QBO in 50 mb temperature. The out-of-phase relation between the equatorial QBO and the middle latitude

QBO is also easily seen in both hemispheres (Zerefos, 1983). These features are evident in both

Oslo CTM2 model and SBUV satellite data sets.

Next, we have calculated the amplitude of QBO in total ozone, [i.e. (max–min)/2], which is presented

in Fig.3 in Dobson Units (left side) and in per cent of the zonal mean (right side). In the tropics, the differences

in amplitudes between the modelled and satellite derived total ozone are up to 2%. Over the

southern extra tropics the difference is about 0.5%, increasing over northern extra tropics to about 1%

of the mean. Over the northern and southern sub-tropics (10o–20o) the differences are zero.

The highly significant correlations between the modeled and satellite derived ozone variations, as described

above, allowed looking at variations in ozone at lower altitudes as well. In the boundary layer,

ozone is of high importance because it serves as an indicator of air quality. In the troposphere, it affects

the atmospheric environment through radiative and chemical processes. With the Oslo CTM2 we were

able to look at variations in both surface and tropospheric ozone. We have analysed the seasonal variability

of tropospheric ozone columns and surface ozone concentrations as calculated by the model.

Here, we do not compare with satellite observations as we have done in the case of total ozone or with

ground-based measurements. We just present the annual variations of surface and tropospheric ozone

as calculated by the model, and refer to Chapter 7 (Surface Ozone) of the World Data Centre for

Greenhouse Gases (WDCGG) No. 36 report (WMO, 2012) as a reference for discussion of seasonal

variations from ground-truth measurements. A detailed comparison with data from ground-based stations

is planned to be performed in a future study.

The seasonal cycles of monthly mean surface and tropospheric ozone simulated by the model averaged

for each 30o latitudinal zone are shown in Fig. 4. Shown are deviations from the long-term mean for

comparison purposes. It appears that the seasonal variability of surface ozone from the model resembles

the respective one reported in figure 7.1 of the WDCGG report, with the latitudinal mean mole

fractions being elevated in spring in most latitudinal zones. It should be mentioned here that the stations

reporting the mole fraction of surface ozone in WDCGG are few in number and unevenly distributed

around the globe, and that the majority of those stations is located in Europe. Therefore it is not

strange if there are differences between our figure and figure 7.1 of the WDCGG report. However, our

analysis restricts us to provide quantitative results from the comparison of the annual cycles, and only

qualitative estimates can be inferred with caution.

4 Conclusions

This study analysed monthly averaged total ozone amounts from improved Oslo CTM2 model simulations

for the period 1998-2009, and compared them with respective total ozone columns from SBUV

satellite data. Total ozone columns from improved Oslo CTM2 model calculations compared well with

the satellite data and the differences ranged between +2% in the tropics and –6% over middle latitudes.

Comparison of monthly mean total ozone anomalies from the model with satellite retrievals using linear

regression analysis, showed statistically significant correlation coefficients between the two data sets

at all latitude zones (correlations of +0.94 between 10 oN and 10 oS, +0.80 over 10–60 oN, and +0.87

over 10–60 oS). Correlations between modelled ozone and the QBO were found to be the order of +0.8

in the tropics. The impact of QBO was most pronounced at equatorial latitudes with amplitudes of +4%

to –4%.

In summary, model results reproduced global observed ozone column well. Multiyear analysis gave

good agreement between modelled and satellite-derived ozone column variations, and also revealed

large scale impact of the QBO on the ozone column. These findings provide significant level of confidence

when studying inter-annual variations of ozone columns with the Oslo CTM2 model.