- Email: [email protected]
Long-term trends of total ozone column over the Iberian Peninsula for the period 1979–2008
Atmospheric Environment 45 (2011) 6283e6290
Contents lists available at SciVerse ScienceDirect
Atmospheric Environment journal homepage: www.elsevier.com/locate/atmosenv
Long-term trends of total ozone column over the Iberian Peninsula for the period 1979e2008 M. Antóna, b, c, *, D. Bortolic, d, P.S. Kulkarnic, M.J. Costac, A.F. Dominguesc, D. Loyolae, A.M. Silvac, L. Alados-Arboledasa, b a
Departamento de Física Aplicada, Universidad de Granada, Granada, Spain Centro Andaluz de Medio Ambiente (CEAMA), Granada, Spain Geophysics Centre of Évora, University of Évora, Évora, Portugal d Institute for Atmospheric Sciences and Climate e National Council for the Researches (ISAC-CNR), Bologna, Italy e Deutsches Zentrum für Luft- und Raumfahrt (DLR), Institut für Methodik der Fernerkundung (IMF), 82234 Oberpfaffenhofen, Germany b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 26 April 2011 Received in revised form 18 August 2011 Accepted 19 August 2011
The objective of this work is to analyze the total ozone column (TOC) trends over the Iberian Peninsula during the last 30 years (1979e2008). This study is carried out using TOC data derived from the Multi Sensor Reanalysis (MSR), Total Ozone Mapping Spectrometer (TOMS) and Global Ozone Monitoring Experiment (GOME). The analysis of the long-term ozone trends is focused on two sub-periods (1979e1994 and 1995e2008) in order to detect changes in the ozone trend pattern. The results show that the ozone depletion was statistically significant at the 95% confidence level during the first sub-period (1979e1994) in the entire region of study (except in the Southerner locations), with linear trends from 4.5%/decade to 2.5%/decade. These linear trends present a clear dependence on latitude, being higher for the Northerner locations than for the Southerner. By contrast, the analysis of the second sub-period of study (1995e2008) shows positive ozone trends over the Iberian Peninsula, with the highest values (þ2.5%/decade) in the Northeast of this region. This result indicates that the ozone layer may be responding as expected to the controls on ozone-depleting substances imposed by the Montreal Protocol. Additionally, a seasonal trend analysis is performed using the average of the deseasonalized monthly values for each season of the year. The seasonal analysis showed that the negative ozone trends during the first sub-period of study were statistically significant in the spring and winter, while that the seasonal ozone trends obtained during the second sub-period are positive but in general not significant at 95%. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Total ozone column Trend TOMS GOME Iberian Peninsula
1. Introduction It is well known that the ozone layer has an outstanding role in protecting all living organisms from harmful ultraviolet (UV) radiation. Ozone is continually produced, destroyed and circulated in the Earth’s atmosphere by a great variety of natural processes (Grewe, 2006). The ozone layer has additionally been clearly affected by human activities. In 1974, chlorofluorocarbons (CFCs, manufactured as refrigerants and aerosol propellants) were identified as the major source of stratospheric chlorine compounds (Molina and Rowland, 1974). Several studies showed that this compound could destroy ozone in the Earth’s stratosphere
* Corresponding author. Departamento de Física Aplicada, Universidad de Granada, 18006 Granada, Spain. Tel.: þ34 924 289536; fax: þ34 924 289651. E-mail address: [email protected] (M. Antón). 1352-2310/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2011.08.058
(e.g., Stolarski and Cicerone, 1974). This hypothesis was confirmed in 1985 by a measured ozone loss over Antarctica (Farman et al., 1985), and in 1988 by a significant decrease of total ozone during winter over northern midlatitudes, later confirmed by satellite observations (Stolarski et al., 1991, 1992). Several studies using ground-based and satellite measurements have demonstrated that stratospheric ozone levels significantly declined until the mid1990s in the middle and high-latitude regions of the two hemispheres (e.g., Herman et al., 1993; Callis et al., 1997; Harris et al., 1997; Staehelin et al., 2001). As consequence of the Montreal Protocol from 1987 and its amendments (UNEP, 2006) the production and release of CFCs and other ozone depleting substances (ODSs) containing chlorine and bromine decreased. This agreement prevented a further thinning of the ozone layer and lead to a gradual recovery in the next decades (WMO, 2011). The work of Newman et al. (2009) simulated a future world (“world avoided”) where the ODSs were never regulated, and
6284
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290
its production grew at an annual rate of 3%. They found by means of this “world avoided” simulation that 17% of the globally averaged column ozone would be destroyed by 2020 and 67% by 2065 in comparison to 1980. Close monitoring of the changes in the ozone layer has become a subject of major concern both by the scientific community and the general public. In this sense, sustainable global ozone observations with space-borne instruments play an essential role in explaining and understanding the global changes of stratospheric ozone and its impact on climate and human health (Loyola et al., 2009). The main objective of this work is to quantify the decline and subsequent recovery of the TOC values over the Iberian Peninsula using satellite measurements. The long-term analysis is divided into two phases in order to detect changes in the ozone trend pattern: from 1979 to 1994 (period of increase in the ODSs concentration in the stratosphere) and from 1995 to 2008 (slow decrease in the ODSs). A similar approach has been followed by other authors (e.g., Weatherhead and Andersen, 2006). The ozone trends have been analyzed using three datasets: the NASA Total Ozone Mapping Spectrometer (TOMS) for the first sub-period (1979e1994), the ESA Global Ozone Monitoring Experiment (GOME) for the second sub-period (1995e2008), and the Multi Sensor Reanalysis (MSR) TOC data (van der A et al., 2010) for the complete period (1979-e2008). Therefore, this paper is expected to contribute to improve the understanding of long-term TOC evolution in the Southwestern Europe. This paper is divided into four sections. The TOC data used in this work are presented in Section 2. In Section 3, the methodology used to characterize the long-term TOC variability is explained. The results and its discussion are given in Section 4. The conclusions of the paper are summarized in Section 5. 2. Total ozone data The TOMS instruments installed on three successive satellites has provided daily images of the global ozone distribution with good spatial resolution from November 1978 until December 2005 (McPeters et al., 1996, 1998). In this work, the TOMS TOC data have been inferred using the last available version (V8) of the TOMS retrieval algorithm (Bhartia and Wellemeyer, 2002; Wellemeyer et al., 2004). Extensive validation exercises on these TOMS TOC data using ground-based Brewer and Dobson spectrophotometers have shown an excellent agreement (at the 1% level) (Labow et al., 2004; Vanicek, 2006; Balis et al., 2007; Antón et al., 2010). The GOME instrument (Burrows et al., 1999) recorded daily TOC measurements from July 1995 to June 2011. The current operational algorithm for the retrieval of TOC data using this satellite instrument is the GOME Data Processor Version (GDP) Version 4.x (van Roozendael et al., 2006). The accuracy of the total ozone columns retrieved by this GOME retrieval algorithm is, in average, very high (within a few percent) (Balis et al., 2007; Antón et al., 2008, 2009; Loyola et al., 2011). The GOME data is freely available product information and imagery is available at http://atmos.caf.dlr.de/ gome. MSR is a continuous and consistent ozone column dataset of 30 years, which has been created from the assimilation of fourteen total ozone satellite retrieval datasets from the following instruments: TOMS (on the satellites Nimbus-7 and Earth Probe), SBUV (Nimbus-7, NOAA-9, NOAA-11 and NOAA-16), GOME (ERS-2), SCIAMACHY (Envisat), OMI (EOS-Aura), and GOME-2 (Metop-A). The MSR dataset is available on a grid of 1 1.5 with a sample frequency of 6 h for the complete time period between November 1978 and December 2008. Details of appropriate correction function used and assimilation of multiple satellite datasets are discussed by van der A et al. (2010). MSR TOC data are distributed via
internet through the Tropospheric Emission Monitoring Internet Service (TEMIS) which can be found at http://www.temis.nl. 3. Methodology In the present analysis, the MSR dataset for spatial coverage 35 Ne45 N/10.5 We4.5 E latitude/longitude (121 grid points) over the Iberian Peninsula was used. Additionally, we select the TOC data derived from MSR, TOMS and GOME over nine representative locations whose spatial distribution is shown in Fig. 1. From North to South and from West to East, the nine locations are: Coruña (43.3 N, 8.4 W), Santander (43.5 N, 3.8 W), Barcelona (41.4 N, 2.2 W), Figueira da Foz (40.1 N, 8.9 W), Madrid (40.4 N, 3.7 W), Valencia (39.5 N, 0.4 W), Faro (37.2 N, 7.9 W), Tarifa (36.0 N., 5.6 W), and Almería (36.8 N, 2.4 W). In this work we are interested in long-term TOC trends and not day to day variations, therefore, monthly mean total ozone column time-series datasets, M(t), were used. These datasets were derived from the monthly averages of the 6-hourly instantaneous MSR data, and from the daily TOMS and GOME data. The fist step in the analysis of the long-term ozone trends is to deseasonalize the monthly TOC series for each individual location. Deseasonalization process of a given series is based on obtaining a good model of the seasonal component, and then subtracting it from the measured (monthly) series. In this work, the annual cycle of TOC data (seasonal pattern) was estimated from the best fit of monthly TOC values by least squares method using the following function (Fioletov et al., 2008):
SðtÞ ¼ a þ b$sin½w$t þ c$cos½w$t
(1)
where t is the time in months, w ¼ 2p/12, a is the central TOC value, and the term b$sin(w$t) þ c$cos(w$t) represents the seasonal component of the TOC variability. Theffi amplitude of this seasonal pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi component was calculated as b2 þ c2 . Thus, the deseasonalized monthly time series is calculated as:
DðtÞ ¼ MðtÞ SðtÞ
(2)
The annual mean time series ðDÞ is derived from the average of these deseasonalized monthly values for each year. The linear ozone trend over each MSR grid point and the selected TOMS and GOME locations is obtained from the linear regression analysis
Fig. 1. Spatial distribution of the nine locations selected to analyze the long-term TOC variability over the Iberian Peninsula.
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290
applied on this annual time series. Thus, the linear trend in functional form is equal to Dwa þ b$t; where a is the intercept, b the slope, and t is the time in the units of years. Estimated value of b represents the long-term ozone trend of the annual time series. The entire period (January 1979eDecember 2008) was divided into two sub-periods (1979e1994 and 1995e2008) to analyze and compare the TOC trends over the Iberian Peninsula. Additionally, we work with four time series for the four seasons of the year. From the climate point of view, Iberian Peninsula has four seasons, winter (December, January, and February), spring (March, April, and May), summer (June, July, and August), and autumn (September, October, and November) (Rodrigo and Trigo, 2007). These four seasonal time series are inferred from the average of the deseasonalized values for each season of the year. Thus, the longterm ozone trends are also calculated separately for each season by means of a linear regression analysis applied on each time series. The deseasonalized TOC values retain the inter-annual variations which could produce persistence in this deseasonalized TOC series (Weatherhead et al., 2000; Vyushin et al., 2007; Tandon and Attri, 2010). The possible existence of persistence was analyzed by means of the temporal autocorrelation coefficients, which were calculated for a certain temporal lag of n months by the following expression (Schmalwieser et al., 2003):
PTn TOCi TOC TOCiþn TOC Cn ¼ P i ¼ 1 2 PTn 2 Tn i ¼ 1 TOCi TOC i ¼ 1 TOCiþn TOC
(3)
where TOCi and TOCiþn denote the deseasonalized TOC values for the months i and i þ n, and T the whole numbers of months. The mean value of the deseasonalized TOC series for the whole period is calculated as:
TOC ¼
T 1X TOCi T i¼1
(4)
6285
The uncertainty of the ozone trends is given by their standard error (se). In addition, the statistical significance of these trends for Gaussian distribution are derived from the confidence limits of the long-term ozone trend, thus, these limits at the 95% confidence level are calculated as b (1.96$se). Presence of the first order autocorrelation (C1) in the deseasonalized TOC series reduces the effective sample size, affecting the standard error estimates (Weatherhead et al., 1998, 2000; Tandon and Atrri, 2010):
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð1 þ C1 Þ se ¼ se$ ð1 C1 Þ 0
(5)
Therefore, the estimation of standard error without accounting the presence of first order autocorrelation would underestimate the error, and in turn, this would affect the statistical significance of the long-term trends of total ozone column. 4. Results and discussion It is well known that the major component of long-term TOC variation at midlatitudes is its annual cycle, which is controlled by a balance between transport associated with the diabatic mean circulation of the stratosphere or BrewereDobson circulation and photochemical loss (Tung and Yang, 1988). This component is clearly appreciated in the evolution of the monthly total ozone values derived from the MSR dataset over Madrid for the period 1979e2008 (Fig. 2, top). A buildup of TOC values during winter and early spring can be seen, when the transport is dominant, while a decline is observed through late spring and summer, when transport decreases and photochemical loss dominates with the increase of solar radiation. To characterize the annual cycle, the monthly TOC series of each selected location were fitted to a periodic function (Eq. (1)). The amplitude (error) of this seasonal component for each individual location is shown in Table 1. The
Fig. 2. Top: Time series of the monthly TOC values over Madrid during the period 1979e2008 using the MSR data. Bottom: Evolution of the deseasonalized monthly time series over Madrid during the period 1979e2008 using the MSR data.
6286
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290 Table 1 Amplitude of the seasonal component (error) for the nine locations at the Iberian Peninsula for the period 1979e2008 using the MSR data. Amplitude (DU) Coruña Santander Barcelona Figueira Madrid Valencia Faro Tarifa Almería
35.9 35.5 33.5 32.0 31.6 30.5 28.2 26.0 27.8
2.1 2.1 2.0 2.3 2.3 2.2 2.0 2.1 2.0
amplitude varies from (35.9 2.1) DU at Coruña to (27.8 2.1) DU at Almería, indicating that the amplitude of the annual TOC cycle increases with latitude over the Iberian Peninsula. This clear dependence of seasonal amplitude with respect to latitude is associated with a greater wintertime transport in the higher latitudes due to larger planetary-wave amplitudes which are modulated by the tropical zonal winds (Holton and Tan, 1980; Tung and Yang, 1988). This amplitude dependence on latitude produces higher values of TOC during spring for the northern latitudes than for the southern latitudes in the region of study. By contrast, the TOC values during autumn are quite similar among latitudes in this region. In order to analyze the long-term ozone trend, the major component of natural ozone variations (seasonal component) must be subtracted from the monthly total ozone dataset (Eq. (2)). Fig. 2 (bottom) shows the evolution of this deseasonalized monthly time series derived from the MSR dataset over Madrid for the period 1979e2008. In addition, following the approach of Hadjinicolaou et al. (2005), this time series was smoothed with a 24-month running averages in order to make the long-term ozone changes more prominent. A distinct ozone decline can be clearly appreciated for the sub-period 1979e1994. By contrast, it can be seen a more stable behavior during the second sub-period 1995e2008, showing even a slight increase in TOC values. Similar evolutions of the deseasonalized monthly time series are obtained for the rest of individual locations at the Iberian Peninsula (not shown). The next paragraphs deal with the quantification and explanation of these ozone long-term changes observed in each sub-period of study. The annual mean time series for each MSR pixel over the Iberian Peninsula is derived from the deseasonalized monthly values as was explained in Section 3. Fig. 3 shows the spatial distributions of the trends (expressed in percentage per decade) over the Iberian Peninsula for the periods 1979e1994 (top), 1995e2008 (middle) and 1979e2008 (bottom). In addition, Table 2 shows the trend results of the mean annual time series derived from MSR, TOMS and GOME datasets expressed in percentage per decade for each selected location. Those trends statistically significant at the 95% confidence level are marked with asterisk. For the first sub-period (1979e1994), Fig. 3 (top) shows negative trends derived from MSR data over the entire region of study which vary from (4.5 1.8)%/ decade (Santander) to (2.6 1.9)%/decade (Tarifa), being all of them statistically significant except for the Southerner locations. Table 2 also shows a good agreement between the ozone trends derived from MSR and TOMS data. This is expected because the MSR data for this time period is based on TOMS and SBUV data that are already inter-calibrated. These clear negative trends are mainly due to the increase of the atmosphere contamination by ODSs during this sub-period. WMO reports on the state of ozone layer (WMO, 2003, 2007, 2011) have underlined that this ODSs increase was mostly responsible for the ozone depletion in the extratropics
Fig. 3. Spatial distributions of the trends (expressed in percentage) over the Iberian Peninsula for the periods 1979e1994 (top), 1995e2008 (middle) and 1979e2008 (bottom) using the MSR data.
in the 1980s and early 1990s. Additionally, several studies have confirmed that the long-term TOC variability over South Europe is partially controlled by climate variability described mainly by the North Atlantic Oscillation (NAO) (e.g., Appenzeller et al., 2000; Orsolini and Limpasuvan, 2001). The NAO induces a clear signature on the tropospheric storm tracks and hence, on synoptic-scale TOC fluctuations, which significantly impinge in monthly and
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290 Table 2 Linear trends in percent per decade (error) for the nine locations at the Iberian Peninsula using the MSR data (upper rows). Results for TOMS data (period 1979e1994) and GOME data (period 1995e2008) are shown in the lower rows. Trends statistically significant at the 95% confidence level are marked with asterisk. Trend 1979e1994 (%/decade) Coruña Santander Barcelona Figueira Madrid Valencia Faro Tarifa Almería
(4.3 (4.2 (4.5 (4.5 (3.9 (4.1 (3.4 (3.6 (3.6 (3.0 (3.6 (3.6 (2.8 (2.9 (2.6 (3.0 (2.7 (3.2
1.6)* 1.5)* 1.8)* 1.7)* 1.8)* 1.7)* 1.6)* 1.5)* 1.8)* 1.5)* 1.8)* 1.7)* 1.8) 1.6) 1.9) 2.1) 1.9) 2.0)
Trend 1995e2008 (%/decade) (þ1.4 (þ0.9 (þ1.9 (þ1.5 (þ2.4 (þ1.8 (þ1.8 (þ0.6 (þ2.3 (þ2.9 (þ2.1 (þ1.3 (þ1.5 (þ0.7 (þ1.1 (þ1.1 (þ1.4 (þ0.7
1.6) 1.5) 1.7) 1.0) 1.5) 1.2) 1.3) 1.0) 1.2)* 1.4)* 1.3) 1.0) 1.2) 0.6) 1.0) 0.7) 1.1) 0.8)
Trend 1979e2008 (%/decade) (1.4 0.7)* e (1.5 0.7)* e (1.2 0.8) e (1.1 0.6) e (1.1 0.7) e (1.1 0.7) e (1.1 0.6) e (1.0 0.7) e (1.0 0.7) e
seasonal TOC levels. Therefore, the NAO’s tendency to remain in its high phase in late eighties and nineties may also account for part of the decreasing trends over the Iberian Peninsula (Orsolini and Doblas-Reyes, 2003). Fig. 3 (top) also shows that the negative ozone trend during this first sub-period present a clear latitudinal dependence, being the trends larger for Northerner latitudes over the region of study. This result is in agreement with previously published works on regions located at higher latitudes than the Iberian Peninsula (e.g., Chen and Nunez, 1998; Fioletov et al., 2002; Svendby and Dahlback, 2004; Hadjinicolaou et al., 2005; Smedley et al., 2010) which showed ozone trends from 1979 to middle 1990s larger than the trends shown in this work. Fig. 3 (middle) shows that the ozone trends derived from MSR data over the Iberian Peninsula for the second sub-period (1995e2008) are all positive with values between (þ1.1 1.0)%/ decade (Tarifa) and (þ2.4 1.5)%/decade (Barcelona). It can be seen that there is no significant dependence on latitude. Nevertheless, it is appreciated the northeaster region of the Iberian Peninsula presents the highest trend values. From Table 2, it is noted that the ozone trends obtained from GOME data are also positive, but
6287
smaller than the trends derived from MSR data (except at Madrid). The GOME instrument is very stable during the complete time period of this study as demonstrated by the comparisons with ground-based instruments (Balis et al., 2007; Antón et al., 2008; Loyola et al., 2009). The difference on the magnitude in the trends computed with MSR and GOME can be explained by the fact that MSR dataset uses not only GOME, but also SCIAMACHY (since 2002), OMI (since 2004) and GOME-2 (since 2007) and those instruments are not as stable as GOME. Unfortunately long-term ozone trends derived from ground-based data can not been obtained at the Iberian Peninsula because ground-based measurements in this region are only available since the late 1990s. All these trend results are in accordance with previous studies. Vyushin et al. (2007) showed positive ozone trends but not statistically significant (in monthly mean satellite data for 5 latitudinal zones) for the period 1996e2005. Yang et al. (2009) found statistically significant positive ozone trend of about 1%/decade in the period 1996e2007 for the averaged 50 Se50 N satellite data. Krzyscin (2010) showed statistically significant positive trends (about 0.5e1%/decade) in the period 1996e2008 for the ozone data averaged over the globe (boreal spring), and in 50 Se50 N zone (boreal summer and whole year). The ozone recovery is partially explained by the control of the production (and hence their emissions into the atmosphere) of the ODSs as result of the implementation of a series of international agreementsdthe Montreal Protocol and its amendmentsdwhich began in 1989 (WMO, 2003). Because of the long lifetimes of ODSs, their atmospheric abundances continued to increase in the early 1990s even as their emissions were decreasing. This fact explains that the change of ozone trend is not observed until middle 1990s. Nevertheless, several studies have shown that the recent positive ozone trends in Northern middle latitudes are larger than those expected from the decline in ODSs (e.g., Wohltmann et al., 2007; and Harris et al., 2008). Thus, in addition to the depletion in ODSs, another reason that could partially explain the positive trend found over the Iberian Peninsula during the second sub-period may be the significant increase in tropospheric ozone in this region since 1996 up to now. This fact has been reported by Kulkarni et al. (2011), who quantify an increase of tropospheric ozone almost 5%e24% over the Iberian Peninsula between the period 1979e1993 and 1996e2005. These authors also reported that the highest values of this tropospheric ozone increase were obtained in the northeaster region of the Iberian Peninsula in accordance with our results. Therefore, this increase of the tropospheric ozone could add to the stratospheric
Fig. 4. Evolution of the annual mean time series from the deseasonalized monthly values for Madrid during the period 1979e2008 using the MSR data. The linear regression trends for the whole period 1979e2008 (black continuous line), and for the sub-periods 1979e1994 and 1995e2008 (dashed-lines) have also been added to the plot.
6288
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290
Fig. 5. Autocorrelation coefficients of the deseasonalized monthly MSR data for Madrid during the period 1979e2008 with a time lag of up to 60 months. The ‘decorrelation’ levels of 0.36 (w1/e) are shown as horizontal black lines.
ozone recovery related to the decline in ODSs over the study region, producing the positive TOC trend found in this work during 1995e2008. Fig. 3 (bottom) shows the ozone trends derived from MSR data over the Iberian Peninsula for the whole period (1979e2008). It can be seen that all trends are negative with similar values between (1.0 0.7)%/decade (Tarifa and Almería) and (1.5 0.7)%/decade (Santander). The trends obtained in this period are statistically significant for the Northerner locations. In addition, there is a slight dependence on latitude in agreement with the spatial distributions of the ozone trends for the first sub-period 1979e1994. These results show that the ozone trend during the whole period is affected larger by the ozone depletion during this first sub-period than by the ozone recovery during the second sub-period. Annually averaged total ozone column is projected to return to 1980 values between 2015 and 2030 over northern midlatitudes, while for southern this return is projected to occur between 2030 and 2040 (WMO, 2011).Around these dates, it is expected a change in the sign of the long-term TOC trend. The evolution of the annual mean time series from the deseasonalized monthly values for Madrid is shown in Fig. 4. The linear regression trends have also been added to the plot which illustrates the apparent change in trend that has occurred over the Iberian Peninsula. The TOC trends are clearly negative during the subperiod 1979e1994, and slightly positive during 1995e2008 as has been explained above. The notable TOC reduction obtained in 1993 is related to the effects of the Mount Pinatubo eruption in June 1991 (e.g., Chandra, 1993; Herman and Larko, 1994). Fig. 4 also shows large inter-annual variability of deseasonalized TOC values, which may make the trend analysis notably sensitive to the length of the data set (Weatherhead et al., 1998). Inter-annual fluctuations of TOC over northern midlatitudes are caused by several factors, such as the regional dynamical structure of the atmosphere, variability of the stratospheric circulation patterns related to anomalously fluxes
of planetary-wave activity from the troposphere, volcanic eruptions, solar flux and the equatorial QBO (Fusco and Salby, 1999; Soukharev, 1999; Appenzeller et al., 2000; Orsolini and Limpasuvan, 2001; Staehelin et al., 2001). This inter-annual variation may produce persistence in the deseasonalized TOC series (Weatherhead et al., 2000; Vyushin et al., 2007; Tandon and Attri, 2010). In this sense, the temporal correlation coefficients (Eq. (3)) were calculated with a time lag of up to 60 months in order to analyze the persistence of the deseasonalized TOC data over the Iberian Peninsula during the period 1979e2008. Fig. 5 shows these autocorrelation coefficients for Madrid. The ‘decorrelation’ levels of 0.36 (w1/e) are also shown as horizontal black lines. It can be seen that the autocorrelation coefficients are below this level for all time lags, indicating that the deseasonalized TOC data over the Iberian Peninsula present no significant autocorrelation. The correlation coefficients for the other individual locations have a very similar behavior (not shown). These results are in agreement with previous works (e.g., Weatherhead et al., 2000; Vyushin et al., 2007) which shown the spatial distributions of the month-tomonth autocorrelation parameter, reporting maximum values at the equator and a notable decreases toward the poles accompanied by some longitudinal variations. Therefore, the influence of the inter-annual fluctuations over the ozone trends obtained in this work is negligible. Nevertheless, the first order autocorrelation coefficient varies from 0.34 (Figueira) to 0.46 (Almeria) which produce an increase of the error in calculated long-term trend (Eq. (5)) by factor of around 1.5. Finally, we obtain the long-term ozone trend separately for each season and location from the deseasonalized monthly values. The results obtained for Santander, Madrid and Tarifa are shown in Table 3. It can be seen that all seasonal ozone trends obtained during the first sub-period are negative in agreement with the strong decline derived from the annual mean time series. These trends are significant at 95% during spring and winter except for the
Table 3 Linear trends in percent per decade (error) for each season at Santander, Madrid and tarifa using the MSR data. Trends statistically significant at the 95% confidence level are marked with asterisk. Trend 1979e1994 (%/decade)
Winter Spring Summer Autumn
Trend 1995e2008 (%/decade)
Santander
Madrid
6.2 7.8 3.2 0.5
5.5 5.8 2.2 0.5
3.1* 2.8* 1.8 1.5
Tarifa 2.6* 2.9* 1.6 1.5
4.0 4.0 1.1 1.4
2.5 4.2 1.8 1.8
Santander
Madrid
þ0.4 þ4.0 þ1.0 þ1.8
þ0.7 þ4.5 þ0.8 þ3.1
3.4 3.9 1.4 1.9
Tarifa 4.6 3.8 1.0 1.6*
þ0.9 þ1.9 0.3 þ1.8
3.9 3.8 1.5 0.9*
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290
Southerner locations. The TOC declines during spring show large values between (4.0 4.2)%/decade for Tarifa and (7.8 2.8)%/ decade for Santander which is consistent with other findings for Northern Midlatitudes (Fioletov et al., 2002; Wohltmann et al., 2007; Antón et al., 2011). In addition, the negative summer trend could justify the observed increase of UV radiation level in the midlatitudes of the Northern Hemisphere in summer months during the past decades (Fioletov et al., 2001; Den Outer et al., 2005). It should be noted that the negative ozone trends derived from spring, summer and winter confirm the latitude dependence (Northerner trends larger than Southerner trends) observed in the annual ozone trends shown above. By contrast, most seasonal ozone trends obtained during the second sub-period are positive but not significant at 95% in accordance with the results inferred from the annual trend analysis. Nevertheless, it can be seen that the spring ozone trends present the largest positive values, opposing to the great negative trends found for this season during the first subperiod. 5. Conclusions Some important conclusions may be drawn from the analysis of the ozone trends using MSR, TOMS and GOME data over the Iberian Peninsula for the period 1979e2008, and two sub-periods (1979e1994 and 1995e2008). The ozone depletion was statistically significant during the first sub-period in this entire region (except in the Southerner locations), with linear trends from 4.5%/ decade to 2.5%/decade. These linear trends presented a clear dependence on latitude, being higher for the Northerner locations than for the Southerner. In addition, the ozone depletion had a distinct seasonal signature, with statistically significant rates observed during winter and spring. By contrast, the analysis of the second sub-period of study (1995e2008) showed positive ozone trends over the Iberian Peninsula, finding the highest values (þ2.5%/decade) in the Northeast of this region. These positive ozone trends were higher than the values reported in the literature for other regions. This behavior was related to the significant increase of the tropospheric ozone over the study region, adding this increase to the stratospheric ozone recovery. Overall we would like to emphasize the remarkable success of the Montreal Protocol for controlling production of ozone depleting substances. Nevertheless, even though the abundance of these substances has significantly diminished in the atmosphere, other anthropogenic effects could complicate the ozone recovery process and may result in considerably altered ozone levels in the future. Therefore, the sustainable measurements of accurate total ozone data with ground-based and satellite instruments are essential to monitor the recovery of ozone layer in the next decades. Acknowledgments The authors would like to thank to the Royal Netherlands Meteorological Institute (KNMI) for providing the Multi Sensor Reanalysis (MSR) data, the German Aerospace Center (DLR) for providing the GOME/ERS-2 ozone data on behalf of ESA, and the NASA TOMS Science Team for providing the TOMS/NIMBUS-7 and TOMS/ METEOR-3 ozone data. Manuel Antón thanks Ministerio de Ciencia e Innovación and Fondo Social Europeo for the award of a postdoctoral grant (Juan de la Cierva). Pavan S Kulkarn is thankful to the Geophysics Centre of the University of Evora (CGE-UE) for the fellowship in the project ‘SPATRAM-MIGE Polar Project’, funded by the Portuguese Science Foundation d FCT. This work was partially supported by the Andalusian Regional Governments through project P10-RNM-6299 and P08-RNM-3568, the Spanish Ministry of Science and Technology through projects CGL2010-18782 and CSD2007-
6289
00067, the European Union through ACTRIS project (EU INFRA2010-1.1.16-262254)., and by Fundação para a Ciência e a Tecnologia though projects: PTDC/AAC-CLI/114031/2009 (FCOMP-01-0124FEDER-014024) and PTDC/CTE-ATM/102142/2008 (FCOMP-010124-FEDER-009303). References Antón, M., Loyola, D., Navascúes, B., Valks, P., 2008. Comparison of GOME total ozone data with ground data from the Spanish Brewer spectroradiometers. Annales Geophysicae 26, 401e412. doi:10.5194/angeo-26-401-2008. Antón, M., Loyola, D., López, M., Vilaplana, J.M., Bañón, M., Zimmer, W., Serrano, A., 2009. Comparison of GOME-2/MetOp total ozone data with Brewer spectroradiometer data over the Iberian Peninsula. Annales Geophysicae 27, 1377e1386. doi:10.5194/angeo-27-1377-2009. Antón, M., Koukouli, M.E., Kroon, M., McPeters, R.D., Labow, G.J., Balis, D., Serrano, A., 2010. Global validation of empirically corrected EP-Total Ozone Mapping Spectrometer (TOMS) total ozone columns using Brewer and Dobson ground-based measurements. Journal of Geophysical Research 115 (D19305). doi:10.1029/2010JD014178. Antón, M., Bortoli, D., Costa, M.J., Kulkarni, P.S., Domingues, A.F., Barriopedro, D., Serrano, A., Silva, A.M., 2011. Temporal and spatial variabilities of total ozone column over Portugal. Remote Sensing of Environment 115, 855e863. Appenzeller, C., Weiss, A.K., Staehelin, J., 2000. North Atlantic Oscillation modulates total ozone winter trends. Geophysical Research Letters 27, 1131e1134. Balis, D., Lambert, J.C., Van Roozendael, M., Spurr, R., Loyola, D., Livschitz, Y., Valks, P., Amiridis, V., Gerard, P., Granville, J., Zehner, C., 2007. Ten years of GOME/ERS2 total ozone data: the new GOME Data Processor (GDP) Version 4: II. Ground-based validation and comparisons with TOMS V7/V8. Journal of Geophysical Research 112 (D07307). doi:10.1029/2005JD006376. Bhartia, P.K., Wellemeyer, C., 2002. TOMS-V8 total O3 algorithm. In: Bhartia, P.K. (Ed.), OMI Algorithm Theoretical Basis Document. OMI Ozone Products, vol. II. NASA Goddard Space Flight Cent., Greenbelt, Md, pp. 15e31. Available at: http:// eospso.gsfc.nasa.gov/eos_homepage/for_scientists/atbd/index.php. Burrows, J.P., Weber, M., Buchwitz, M., Razonov, V., Ladstatter, A., Richter, A., De Beerk, R., Hoogen, R., Bramsdted, D., Eichmann, K.U., Eisenger, M., Perner, D., 1999. The Global Ozone Monitoring Experiment (GOME): mission concept and first scientific results. Journal of Atmospheric Science 56, 151e175. Callis, L.B., Natarajan, M., Lambeth, J.D., Boughner, R.E., 1997. On the origin of midlatitude ozone changes: data analysis and simulations for 1979e1993. Journal of Geophysical Research 102, 1215e1228. Chandra, S., 1993. Changes in stratospheric ozone and temperature due to the eruptions of Mt. Pinatubo. Geophysical Research Letters 20, 33e36. Chen, D., Nunez, M., 1998. Temporal and spatial variability of total ozone in southwest Sweden revealed by two ground-based instruments. International Journal of Climatology 18, 1237e1246. Den Outer, P.N., Slaper, H., Tax, R.B., 2005. UV radiation in the Netherlands: assessing long-term variability and trends in relation to ozone and clouds. Journal of Geophysical Research 110 (D02203). doi:10.1029/2004JD004824. European Spatial Agency: (ESA), 1995. GOME Global Ozone Monitoring Experiment Users Manual. ESA. SP-1182. Farman, J.C., Gardiner, B.G., Shanklin, J.D., 1985. Large losses of total ozone in Antarctica reveal seasonal ClOx/NOx interaction. Nature 315, 207e210. Fioletov, V.E., McArthur, L., Kerr, J.E., Wardle, D.I., 2001. Long-term variations of UV-B irradiance over Canada estimated from Brewer observations and derived from ozone and pyranometer measurements. Journal of Geophysical Research 106, 2307e2309. Fioletov, V.E., Bodeker, G.E., Miller, A.J., McPeters, R.D., Stolarski, R., 2002. Global and zonal total ozone variations estimated from ground-based and satellite measurements: 1964e2000. Journal of Geophysical Research 107 (4647). doi:10.1029/2001JD001350. Fioletov, V.E., Labow, G., Evans, R., et al., 2008. Performance of the ground-based total ozone network assessed using satellite data. Journal of Geophysical Research 113 (D14313). doi:10.1029/2008JD009809. Fusco, A.C., Salby, M.L., 1999. Interannual variations of total ozone and their relationship to variations of planetary wave activity. Journal of Climatology 12, 1619e1629. Grewe, V., 2006. The origin of ozone. Atmospheric Chemistry and Physics 6, 1495e1511. Hadjinicolaou, P., Pyle, J.A., Harrise, N.R.P., 2005. The recent turnaround in stratospheric ozone over northern middle latitudes: a dynamical modelling perspective. Geophysical Research Letters 32 (L12821). doi:10.1029/ 2005GL022476. Harris, N.R.P., Ancellet, J., Bishop, L., Hofmann, D.J., Kerr, J.B., McPeters, R.D., et al., 1997. Trends in stratospheric and free tropospheric ozone. Journal of Geophysical Research 102, 1571e1590. Harris, N.R.P., Kyrö, E., Staehelin, J., Brunner, D., Andersen, S.-B., GodinBeekmann, S., Dhomse, S., Hadjinicolaou, P., Hansen, G., Isaksen, I., Jrrar, A., Karpetchko, A., Kivi, R., Knudsen, B., Krizan, P., Lastovicka, J., Maeder, J., Orsolini, Y., Pyle, J.A., Rex, M., Vanicek, K., Weber, M., Wohltmann, I., Zanis, P., Zerefos, C., 2008. Ozone trends at northern mid- and high latitudes e a European perspective. Annales Geophysicae 26, 1207e1220.
6290
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290
Herman, J.R., McPeters, R.D., Larko, D., 1993. Ozone depletion at Northern and Southern latitudes derived from January 1979 to December 1991 Total Ozone Mapping Spectrometer data. Journal of Geophysical Research 98, 12,783e12,793. doi:10.1029/93JD00601. Herman, J.R., Larko, D., 1994. Low ozone amounts during 1992e1993 from Nimbus 7 and Meteor 3 total ozone mapping spectrometers. Journal of Geophysical Research 99, 3483e3496. doi:10.1029/93JD02594. Holton, J.R., Tan, H.-C., 1980. The influence of the equatorial quasi-biennial oscillation on the global circulation at 50 mb. Journal of Atmospheric Sciences 37, 2200e2208. Krzyscin, J.W., 2010. Onset of the total ozone increase based on statistical analyses of global ground-based data for the period 1964e2008. International Journal of Climatology, doi:10.1002/joc.2264. Kulkarni, P.S., Bortoli, D., Salgado, R., Antón, M., Costa, M.J., Silva, A.M., 2011. Tropospheric ozone variability over Iberian Peninsula. Atmospheric Environment 45, 174e182. Labow, G.J., McPeters, R.D., Barthia, P.K., 2004. A Comparison of TOMS & SBUV Version 8 Total Column Ozone Data with Data from Ground Stations, Paper Presented at Quadrennial Ozone Symposium. Eur. Comm., Kos, Greece. Loyola, D., Coldewey-Egbers, M., Dameris, M., Garny, H., Stenke, A., Van Roozendael, M., Lerot, C., Balis, D., Koukouli, M., 2009. Global long-term monitoring of the ozone layer e a prerequisite for predictions. International Journal of Remote Sensing 30, 4295e4318. Loyola, D., Koukouli, M.E., Valks, P., Balis, D.S., Hao, N., Van Roozendael, M., Spurr, R.J.D., Zimmer, W., Kiemle, S., Lerot, C., Lambert, J.-C., 2011. The GOME-2 total column ozone product: retrieval algorithm and ground-based validation. Journal of Geophysical Research 116 (D07302). doi:10.1029/2010JD014675. McPeters, R.D., Bhartia, P.K., Krueger, A.J., 1996. Nimbus-7 Total Ozone Mapping Spectrometer (TOMS) Data Products User’s Guide. NASA Reference Publication, NASA, Greenbelt, Md. McPeters, R.D., et al., 1998. Earth Probe Total Ozone Mapping Spectrometer (TOMS) Data Products User’s Guide. NASA Reference Publication, NASA, Greenbelt, Md. Molina, M.J., Rowland, F.S., 1974. Stratospheric sink for chlorofluoromethanes: chlorine atom catalysed destruction of ozone. Nature 249, 810e812. Newman, P.A., et al., 2009. What would have happened to the ozone layer if chlorofluorocarbons (CFCs) had not been regulated? Atmospheric Chemistry and Physics 9, 2113e2128. Orsolini, Y.J., Limpasuvan, V., 2001. The North Atlantic Oscillation and the occurrences of ozone miniholes. Geophysical Research Letters 28, 4099e4102. Orsolini, Y.J., Doblas-Reyes, F.J., 2003. Ozone signatures of climate patterns over the Euro-Atlantic sector in the spring. Quarterly Journal Royal Meteorological Society 129, 3251e3263. Rodrigo, F.S., Trigo, R.M., 2007. Trends in daily rainfall in the Iberian Peninsula from 1951 to 2002. International Journal of Climatology 27, 513e529. Schmalwieser, A.W., Schauberger, G., Janouchu, M., 2003. Temporal and spatial variability of total ozone content over Central Europe: analysis in respect to the biological effect on plants. Agricultural and Forest Meteorology 120, 9e26. Smedley, A.R.D., Rimmer, J.S., Moore, D., Toumic, R., Webb, A.R., 2010. Total ozone and surface UV trends in the United Kingdom: 1979e2008. International Journal of Climatology. doi:10.1002/joc.2275. Soukharev, B., 1999. On the solar/QBO effect on the interannual variability of total ozone and the stratospheric circulation over Northern Europe. Journal of Atmospheric and Solar-Terrestrial Physics 61, 1093e11091. Staehelin, J., Harris, N.R.P., Appenzeller, C., Eberhard, J., Piechowski, M., 2001. Observations of ozone trends. Reviews of Geophysics 39, 231e290. Stolarski, R.S., Cicerone, R.J., 1974. Stratospheric chlorine: a possible sink for ozone. Canadian Journal of Chemistry 52, 1610e1615.
Stolarski, R.S., Bloomfield, P., McPeters, R.D., Herman, J.R., 1991. Total ozone trends deduced from Nimbus 7 TOMS data. Geophysical Research Letters 18, 1015e1018. Stolarski, R., Bojkov, R., Bishop, L., Zerefos, C., Staehelin, J., Zawodny, J., 1992. Measured trends in stratospheric ozone. Science 256, 342e349. Svendby, T.M., Dahlback, A., 2004. Statistical analysis of total ozone measurements in Oslo, Norway, 1978e1998. Journal of Geophysical Research 109 (D16107). doi:10.1029/2004JD004679. Tandon, A., Attri, A.K., 2010. Trends in total ozone column over India: 1979e2008. Atmospheric Environment 45, 1648e1654. Tung, K.K., Yang, H., 1988. Dynamic variability of column ozone. Journal of Geophysical Research 93, 11123e11128. UNEP, 2006. Handbook for the Montreal Protocol on Substances that Deplete the Ozone Layer. Ozone Secretariat e United Nations Environment Programme. Available online at: http://ozone.unep.org/Publications/MP_Handbook/index. shtml. van der A, R.J., Allaart, M.A.F., Eskes, H.J., 2010. Multi sensor reanalysis of total ozone. Atmospheric Chemistry and Physics 10, 11277e11294. Vanicek, K., 2006. Differences between ground Dobson, Brewer and satellite TOMS8, GOME-WFDOAS total ozone observations at Hradec Kralove, Czech. Atmospheric Chemistry and Physics 6, 5163e5171. van Roozendael, M., Loyola, D., Spurr, R., Balis, D., Lambert, J.C., Livschitz, Y., et al., 2006. Ten years of GOME/ERS-2 total ozone dataethe new GOME data processor (GDP) version 4: 1. Algorithm description. Journal of Geophysical Research 111, D14311. doi:10.1029/2005JD006375. Vyushin, D., Fioletov, V.E., Shepherd, T.G., 2007. Impact of long-range correlations on trend detection in total ozone. Journal of Geophysical Research 112 (D14307). doi:10.1029/2006JD008168. Weatherhead, E.C., Reinsel, G.C., Tiao, G.C., et al., 1998. Factors affecting the detection of trends: statistical considerations and applications to environmental data. Journal of Geophysical Research 103 (D14), 17149e17161. Weatherhead, E.C., Reinsel, G.C., Tiao, G.C., Jackman, C.H., Bishop, L., Frith, S.M.H., DeLuisi, J., Keller, T., Oltmans, S.J., Fleming, E.L., Wuebbles, D.J., Kerr, J.B., Miller, A.J., Herman, J., McPeters, R., Nagatani, R.M., Frederick, J.E., 2000. Detecting the recovery of total column ozone. Journal of Geophysical Research 105 (D17), 22201e22210. Weatherhead, E.C., Andersen, S.B., 2006. The search for signs of recovery of the ozone layer. Nature 441, 39e45. doi:10.1038/nature04746. Wellemeyer, C.G., Bhartia, P.K., Taylor, S.L., Qin, W., Ahn, C., 2004. Version 8 Total Ozone Mapping Spectrometer (TOMS) Algorithm, Paper Presented at Quadrennial Ozone Symposium. Eur. Comm., Kos, Greece. Wohltmann, I., Lehmann, R., Rex, M., Brunner, D., Mäder, J.A., 2007. A processoriented regression model for column ozone. Journal of Geophysical Research 112 (D12304). doi:10.1029/2006JD007573. World Meteorological Organization (WMO), 2003. Scientific Assessment of Ozone Depletion: 2002. Global Ozone Research and Monitoring Project Technical Report 47, Geneva, Switzerland. World Meteorological Organization (WMO), 2007. Scientific Assessment of Ozone Depletion: 2006. Global Ozone Research and Monitoring Project Technical Report 50, Geneva, Switzerland. World Meteorological Organization (WMO), 2011. Scientific Assessment of Ozone Depletion: 2010 Global Ozone Research and Monitoring Project Technical Report 52, Geneva, Switzerland. Yang, S.-K., Long, C.S., Miller, A.J., He, X., Yang, Y., Wuebbles, D.J., Tiao, G., 2009. Modulation of natural variability on a trend analysis of updated cohesive SBUV(/2) total ozone. International Journal of Remote Sensing 30, 3975e3986. doi:10.1080/01431160902821924.
Contents lists available at SciVerse ScienceDirect
Atmospheric Environment journal homepage: www.elsevier.com/locate/atmosenv
Long-term trends of total ozone column over the Iberian Peninsula for the period 1979e2008 M. Antóna, b, c, *, D. Bortolic, d, P.S. Kulkarnic, M.J. Costac, A.F. Dominguesc, D. Loyolae, A.M. Silvac, L. Alados-Arboledasa, b a
Departamento de Física Aplicada, Universidad de Granada, Granada, Spain Centro Andaluz de Medio Ambiente (CEAMA), Granada, Spain Geophysics Centre of Évora, University of Évora, Évora, Portugal d Institute for Atmospheric Sciences and Climate e National Council for the Researches (ISAC-CNR), Bologna, Italy e Deutsches Zentrum für Luft- und Raumfahrt (DLR), Institut für Methodik der Fernerkundung (IMF), 82234 Oberpfaffenhofen, Germany b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 26 April 2011 Received in revised form 18 August 2011 Accepted 19 August 2011
The objective of this work is to analyze the total ozone column (TOC) trends over the Iberian Peninsula during the last 30 years (1979e2008). This study is carried out using TOC data derived from the Multi Sensor Reanalysis (MSR), Total Ozone Mapping Spectrometer (TOMS) and Global Ozone Monitoring Experiment (GOME). The analysis of the long-term ozone trends is focused on two sub-periods (1979e1994 and 1995e2008) in order to detect changes in the ozone trend pattern. The results show that the ozone depletion was statistically significant at the 95% confidence level during the first sub-period (1979e1994) in the entire region of study (except in the Southerner locations), with linear trends from 4.5%/decade to 2.5%/decade. These linear trends present a clear dependence on latitude, being higher for the Northerner locations than for the Southerner. By contrast, the analysis of the second sub-period of study (1995e2008) shows positive ozone trends over the Iberian Peninsula, with the highest values (þ2.5%/decade) in the Northeast of this region. This result indicates that the ozone layer may be responding as expected to the controls on ozone-depleting substances imposed by the Montreal Protocol. Additionally, a seasonal trend analysis is performed using the average of the deseasonalized monthly values for each season of the year. The seasonal analysis showed that the negative ozone trends during the first sub-period of study were statistically significant in the spring and winter, while that the seasonal ozone trends obtained during the second sub-period are positive but in general not significant at 95%. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Total ozone column Trend TOMS GOME Iberian Peninsula
1. Introduction It is well known that the ozone layer has an outstanding role in protecting all living organisms from harmful ultraviolet (UV) radiation. Ozone is continually produced, destroyed and circulated in the Earth’s atmosphere by a great variety of natural processes (Grewe, 2006). The ozone layer has additionally been clearly affected by human activities. In 1974, chlorofluorocarbons (CFCs, manufactured as refrigerants and aerosol propellants) were identified as the major source of stratospheric chlorine compounds (Molina and Rowland, 1974). Several studies showed that this compound could destroy ozone in the Earth’s stratosphere
* Corresponding author. Departamento de Física Aplicada, Universidad de Granada, 18006 Granada, Spain. Tel.: þ34 924 289536; fax: þ34 924 289651. E-mail address: [email protected] (M. Antón). 1352-2310/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2011.08.058
(e.g., Stolarski and Cicerone, 1974). This hypothesis was confirmed in 1985 by a measured ozone loss over Antarctica (Farman et al., 1985), and in 1988 by a significant decrease of total ozone during winter over northern midlatitudes, later confirmed by satellite observations (Stolarski et al., 1991, 1992). Several studies using ground-based and satellite measurements have demonstrated that stratospheric ozone levels significantly declined until the mid1990s in the middle and high-latitude regions of the two hemispheres (e.g., Herman et al., 1993; Callis et al., 1997; Harris et al., 1997; Staehelin et al., 2001). As consequence of the Montreal Protocol from 1987 and its amendments (UNEP, 2006) the production and release of CFCs and other ozone depleting substances (ODSs) containing chlorine and bromine decreased. This agreement prevented a further thinning of the ozone layer and lead to a gradual recovery in the next decades (WMO, 2011). The work of Newman et al. (2009) simulated a future world (“world avoided”) where the ODSs were never regulated, and
6284
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290
its production grew at an annual rate of 3%. They found by means of this “world avoided” simulation that 17% of the globally averaged column ozone would be destroyed by 2020 and 67% by 2065 in comparison to 1980. Close monitoring of the changes in the ozone layer has become a subject of major concern both by the scientific community and the general public. In this sense, sustainable global ozone observations with space-borne instruments play an essential role in explaining and understanding the global changes of stratospheric ozone and its impact on climate and human health (Loyola et al., 2009). The main objective of this work is to quantify the decline and subsequent recovery of the TOC values over the Iberian Peninsula using satellite measurements. The long-term analysis is divided into two phases in order to detect changes in the ozone trend pattern: from 1979 to 1994 (period of increase in the ODSs concentration in the stratosphere) and from 1995 to 2008 (slow decrease in the ODSs). A similar approach has been followed by other authors (e.g., Weatherhead and Andersen, 2006). The ozone trends have been analyzed using three datasets: the NASA Total Ozone Mapping Spectrometer (TOMS) for the first sub-period (1979e1994), the ESA Global Ozone Monitoring Experiment (GOME) for the second sub-period (1995e2008), and the Multi Sensor Reanalysis (MSR) TOC data (van der A et al., 2010) for the complete period (1979-e2008). Therefore, this paper is expected to contribute to improve the understanding of long-term TOC evolution in the Southwestern Europe. This paper is divided into four sections. The TOC data used in this work are presented in Section 2. In Section 3, the methodology used to characterize the long-term TOC variability is explained. The results and its discussion are given in Section 4. The conclusions of the paper are summarized in Section 5. 2. Total ozone data The TOMS instruments installed on three successive satellites has provided daily images of the global ozone distribution with good spatial resolution from November 1978 until December 2005 (McPeters et al., 1996, 1998). In this work, the TOMS TOC data have been inferred using the last available version (V8) of the TOMS retrieval algorithm (Bhartia and Wellemeyer, 2002; Wellemeyer et al., 2004). Extensive validation exercises on these TOMS TOC data using ground-based Brewer and Dobson spectrophotometers have shown an excellent agreement (at the 1% level) (Labow et al., 2004; Vanicek, 2006; Balis et al., 2007; Antón et al., 2010). The GOME instrument (Burrows et al., 1999) recorded daily TOC measurements from July 1995 to June 2011. The current operational algorithm for the retrieval of TOC data using this satellite instrument is the GOME Data Processor Version (GDP) Version 4.x (van Roozendael et al., 2006). The accuracy of the total ozone columns retrieved by this GOME retrieval algorithm is, in average, very high (within a few percent) (Balis et al., 2007; Antón et al., 2008, 2009; Loyola et al., 2011). The GOME data is freely available product information and imagery is available at http://atmos.caf.dlr.de/ gome. MSR is a continuous and consistent ozone column dataset of 30 years, which has been created from the assimilation of fourteen total ozone satellite retrieval datasets from the following instruments: TOMS (on the satellites Nimbus-7 and Earth Probe), SBUV (Nimbus-7, NOAA-9, NOAA-11 and NOAA-16), GOME (ERS-2), SCIAMACHY (Envisat), OMI (EOS-Aura), and GOME-2 (Metop-A). The MSR dataset is available on a grid of 1 1.5 with a sample frequency of 6 h for the complete time period between November 1978 and December 2008. Details of appropriate correction function used and assimilation of multiple satellite datasets are discussed by van der A et al. (2010). MSR TOC data are distributed via
internet through the Tropospheric Emission Monitoring Internet Service (TEMIS) which can be found at http://www.temis.nl. 3. Methodology In the present analysis, the MSR dataset for spatial coverage 35 Ne45 N/10.5 We4.5 E latitude/longitude (121 grid points) over the Iberian Peninsula was used. Additionally, we select the TOC data derived from MSR, TOMS and GOME over nine representative locations whose spatial distribution is shown in Fig. 1. From North to South and from West to East, the nine locations are: Coruña (43.3 N, 8.4 W), Santander (43.5 N, 3.8 W), Barcelona (41.4 N, 2.2 W), Figueira da Foz (40.1 N, 8.9 W), Madrid (40.4 N, 3.7 W), Valencia (39.5 N, 0.4 W), Faro (37.2 N, 7.9 W), Tarifa (36.0 N., 5.6 W), and Almería (36.8 N, 2.4 W). In this work we are interested in long-term TOC trends and not day to day variations, therefore, monthly mean total ozone column time-series datasets, M(t), were used. These datasets were derived from the monthly averages of the 6-hourly instantaneous MSR data, and from the daily TOMS and GOME data. The fist step in the analysis of the long-term ozone trends is to deseasonalize the monthly TOC series for each individual location. Deseasonalization process of a given series is based on obtaining a good model of the seasonal component, and then subtracting it from the measured (monthly) series. In this work, the annual cycle of TOC data (seasonal pattern) was estimated from the best fit of monthly TOC values by least squares method using the following function (Fioletov et al., 2008):
SðtÞ ¼ a þ b$sin½w$t þ c$cos½w$t
(1)
where t is the time in months, w ¼ 2p/12, a is the central TOC value, and the term b$sin(w$t) þ c$cos(w$t) represents the seasonal component of the TOC variability. Theffi amplitude of this seasonal pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi component was calculated as b2 þ c2 . Thus, the deseasonalized monthly time series is calculated as:
DðtÞ ¼ MðtÞ SðtÞ
(2)
The annual mean time series ðDÞ is derived from the average of these deseasonalized monthly values for each year. The linear ozone trend over each MSR grid point and the selected TOMS and GOME locations is obtained from the linear regression analysis
Fig. 1. Spatial distribution of the nine locations selected to analyze the long-term TOC variability over the Iberian Peninsula.
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290
applied on this annual time series. Thus, the linear trend in functional form is equal to Dwa þ b$t; where a is the intercept, b the slope, and t is the time in the units of years. Estimated value of b represents the long-term ozone trend of the annual time series. The entire period (January 1979eDecember 2008) was divided into two sub-periods (1979e1994 and 1995e2008) to analyze and compare the TOC trends over the Iberian Peninsula. Additionally, we work with four time series for the four seasons of the year. From the climate point of view, Iberian Peninsula has four seasons, winter (December, January, and February), spring (March, April, and May), summer (June, July, and August), and autumn (September, October, and November) (Rodrigo and Trigo, 2007). These four seasonal time series are inferred from the average of the deseasonalized values for each season of the year. Thus, the longterm ozone trends are also calculated separately for each season by means of a linear regression analysis applied on each time series. The deseasonalized TOC values retain the inter-annual variations which could produce persistence in this deseasonalized TOC series (Weatherhead et al., 2000; Vyushin et al., 2007; Tandon and Attri, 2010). The possible existence of persistence was analyzed by means of the temporal autocorrelation coefficients, which were calculated for a certain temporal lag of n months by the following expression (Schmalwieser et al., 2003):
PTn TOCi TOC TOCiþn TOC Cn ¼ P i ¼ 1 2 PTn 2 Tn i ¼ 1 TOCi TOC i ¼ 1 TOCiþn TOC
(3)
where TOCi and TOCiþn denote the deseasonalized TOC values for the months i and i þ n, and T the whole numbers of months. The mean value of the deseasonalized TOC series for the whole period is calculated as:
TOC ¼
T 1X TOCi T i¼1
(4)
6285
The uncertainty of the ozone trends is given by their standard error (se). In addition, the statistical significance of these trends for Gaussian distribution are derived from the confidence limits of the long-term ozone trend, thus, these limits at the 95% confidence level are calculated as b (1.96$se). Presence of the first order autocorrelation (C1) in the deseasonalized TOC series reduces the effective sample size, affecting the standard error estimates (Weatherhead et al., 1998, 2000; Tandon and Atrri, 2010):
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð1 þ C1 Þ se ¼ se$ ð1 C1 Þ 0
(5)
Therefore, the estimation of standard error without accounting the presence of first order autocorrelation would underestimate the error, and in turn, this would affect the statistical significance of the long-term trends of total ozone column. 4. Results and discussion It is well known that the major component of long-term TOC variation at midlatitudes is its annual cycle, which is controlled by a balance between transport associated with the diabatic mean circulation of the stratosphere or BrewereDobson circulation and photochemical loss (Tung and Yang, 1988). This component is clearly appreciated in the evolution of the monthly total ozone values derived from the MSR dataset over Madrid for the period 1979e2008 (Fig. 2, top). A buildup of TOC values during winter and early spring can be seen, when the transport is dominant, while a decline is observed through late spring and summer, when transport decreases and photochemical loss dominates with the increase of solar radiation. To characterize the annual cycle, the monthly TOC series of each selected location were fitted to a periodic function (Eq. (1)). The amplitude (error) of this seasonal component for each individual location is shown in Table 1. The
Fig. 2. Top: Time series of the monthly TOC values over Madrid during the period 1979e2008 using the MSR data. Bottom: Evolution of the deseasonalized monthly time series over Madrid during the period 1979e2008 using the MSR data.
6286
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290 Table 1 Amplitude of the seasonal component (error) for the nine locations at the Iberian Peninsula for the period 1979e2008 using the MSR data. Amplitude (DU) Coruña Santander Barcelona Figueira Madrid Valencia Faro Tarifa Almería
35.9 35.5 33.5 32.0 31.6 30.5 28.2 26.0 27.8
2.1 2.1 2.0 2.3 2.3 2.2 2.0 2.1 2.0
amplitude varies from (35.9 2.1) DU at Coruña to (27.8 2.1) DU at Almería, indicating that the amplitude of the annual TOC cycle increases with latitude over the Iberian Peninsula. This clear dependence of seasonal amplitude with respect to latitude is associated with a greater wintertime transport in the higher latitudes due to larger planetary-wave amplitudes which are modulated by the tropical zonal winds (Holton and Tan, 1980; Tung and Yang, 1988). This amplitude dependence on latitude produces higher values of TOC during spring for the northern latitudes than for the southern latitudes in the region of study. By contrast, the TOC values during autumn are quite similar among latitudes in this region. In order to analyze the long-term ozone trend, the major component of natural ozone variations (seasonal component) must be subtracted from the monthly total ozone dataset (Eq. (2)). Fig. 2 (bottom) shows the evolution of this deseasonalized monthly time series derived from the MSR dataset over Madrid for the period 1979e2008. In addition, following the approach of Hadjinicolaou et al. (2005), this time series was smoothed with a 24-month running averages in order to make the long-term ozone changes more prominent. A distinct ozone decline can be clearly appreciated for the sub-period 1979e1994. By contrast, it can be seen a more stable behavior during the second sub-period 1995e2008, showing even a slight increase in TOC values. Similar evolutions of the deseasonalized monthly time series are obtained for the rest of individual locations at the Iberian Peninsula (not shown). The next paragraphs deal with the quantification and explanation of these ozone long-term changes observed in each sub-period of study. The annual mean time series for each MSR pixel over the Iberian Peninsula is derived from the deseasonalized monthly values as was explained in Section 3. Fig. 3 shows the spatial distributions of the trends (expressed in percentage per decade) over the Iberian Peninsula for the periods 1979e1994 (top), 1995e2008 (middle) and 1979e2008 (bottom). In addition, Table 2 shows the trend results of the mean annual time series derived from MSR, TOMS and GOME datasets expressed in percentage per decade for each selected location. Those trends statistically significant at the 95% confidence level are marked with asterisk. For the first sub-period (1979e1994), Fig. 3 (top) shows negative trends derived from MSR data over the entire region of study which vary from (4.5 1.8)%/ decade (Santander) to (2.6 1.9)%/decade (Tarifa), being all of them statistically significant except for the Southerner locations. Table 2 also shows a good agreement between the ozone trends derived from MSR and TOMS data. This is expected because the MSR data for this time period is based on TOMS and SBUV data that are already inter-calibrated. These clear negative trends are mainly due to the increase of the atmosphere contamination by ODSs during this sub-period. WMO reports on the state of ozone layer (WMO, 2003, 2007, 2011) have underlined that this ODSs increase was mostly responsible for the ozone depletion in the extratropics
Fig. 3. Spatial distributions of the trends (expressed in percentage) over the Iberian Peninsula for the periods 1979e1994 (top), 1995e2008 (middle) and 1979e2008 (bottom) using the MSR data.
in the 1980s and early 1990s. Additionally, several studies have confirmed that the long-term TOC variability over South Europe is partially controlled by climate variability described mainly by the North Atlantic Oscillation (NAO) (e.g., Appenzeller et al., 2000; Orsolini and Limpasuvan, 2001). The NAO induces a clear signature on the tropospheric storm tracks and hence, on synoptic-scale TOC fluctuations, which significantly impinge in monthly and
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290 Table 2 Linear trends in percent per decade (error) for the nine locations at the Iberian Peninsula using the MSR data (upper rows). Results for TOMS data (period 1979e1994) and GOME data (period 1995e2008) are shown in the lower rows. Trends statistically significant at the 95% confidence level are marked with asterisk. Trend 1979e1994 (%/decade) Coruña Santander Barcelona Figueira Madrid Valencia Faro Tarifa Almería
(4.3 (4.2 (4.5 (4.5 (3.9 (4.1 (3.4 (3.6 (3.6 (3.0 (3.6 (3.6 (2.8 (2.9 (2.6 (3.0 (2.7 (3.2
1.6)* 1.5)* 1.8)* 1.7)* 1.8)* 1.7)* 1.6)* 1.5)* 1.8)* 1.5)* 1.8)* 1.7)* 1.8) 1.6) 1.9) 2.1) 1.9) 2.0)
Trend 1995e2008 (%/decade) (þ1.4 (þ0.9 (þ1.9 (þ1.5 (þ2.4 (þ1.8 (þ1.8 (þ0.6 (þ2.3 (þ2.9 (þ2.1 (þ1.3 (þ1.5 (þ0.7 (þ1.1 (þ1.1 (þ1.4 (þ0.7
1.6) 1.5) 1.7) 1.0) 1.5) 1.2) 1.3) 1.0) 1.2)* 1.4)* 1.3) 1.0) 1.2) 0.6) 1.0) 0.7) 1.1) 0.8)
Trend 1979e2008 (%/decade) (1.4 0.7)* e (1.5 0.7)* e (1.2 0.8) e (1.1 0.6) e (1.1 0.7) e (1.1 0.7) e (1.1 0.6) e (1.0 0.7) e (1.0 0.7) e
seasonal TOC levels. Therefore, the NAO’s tendency to remain in its high phase in late eighties and nineties may also account for part of the decreasing trends over the Iberian Peninsula (Orsolini and Doblas-Reyes, 2003). Fig. 3 (top) also shows that the negative ozone trend during this first sub-period present a clear latitudinal dependence, being the trends larger for Northerner latitudes over the region of study. This result is in agreement with previously published works on regions located at higher latitudes than the Iberian Peninsula (e.g., Chen and Nunez, 1998; Fioletov et al., 2002; Svendby and Dahlback, 2004; Hadjinicolaou et al., 2005; Smedley et al., 2010) which showed ozone trends from 1979 to middle 1990s larger than the trends shown in this work. Fig. 3 (middle) shows that the ozone trends derived from MSR data over the Iberian Peninsula for the second sub-period (1995e2008) are all positive with values between (þ1.1 1.0)%/ decade (Tarifa) and (þ2.4 1.5)%/decade (Barcelona). It can be seen that there is no significant dependence on latitude. Nevertheless, it is appreciated the northeaster region of the Iberian Peninsula presents the highest trend values. From Table 2, it is noted that the ozone trends obtained from GOME data are also positive, but
6287
smaller than the trends derived from MSR data (except at Madrid). The GOME instrument is very stable during the complete time period of this study as demonstrated by the comparisons with ground-based instruments (Balis et al., 2007; Antón et al., 2008; Loyola et al., 2009). The difference on the magnitude in the trends computed with MSR and GOME can be explained by the fact that MSR dataset uses not only GOME, but also SCIAMACHY (since 2002), OMI (since 2004) and GOME-2 (since 2007) and those instruments are not as stable as GOME. Unfortunately long-term ozone trends derived from ground-based data can not been obtained at the Iberian Peninsula because ground-based measurements in this region are only available since the late 1990s. All these trend results are in accordance with previous studies. Vyushin et al. (2007) showed positive ozone trends but not statistically significant (in monthly mean satellite data for 5 latitudinal zones) for the period 1996e2005. Yang et al. (2009) found statistically significant positive ozone trend of about 1%/decade in the period 1996e2007 for the averaged 50 Se50 N satellite data. Krzyscin (2010) showed statistically significant positive trends (about 0.5e1%/decade) in the period 1996e2008 for the ozone data averaged over the globe (boreal spring), and in 50 Se50 N zone (boreal summer and whole year). The ozone recovery is partially explained by the control of the production (and hence their emissions into the atmosphere) of the ODSs as result of the implementation of a series of international agreementsdthe Montreal Protocol and its amendmentsdwhich began in 1989 (WMO, 2003). Because of the long lifetimes of ODSs, their atmospheric abundances continued to increase in the early 1990s even as their emissions were decreasing. This fact explains that the change of ozone trend is not observed until middle 1990s. Nevertheless, several studies have shown that the recent positive ozone trends in Northern middle latitudes are larger than those expected from the decline in ODSs (e.g., Wohltmann et al., 2007; and Harris et al., 2008). Thus, in addition to the depletion in ODSs, another reason that could partially explain the positive trend found over the Iberian Peninsula during the second sub-period may be the significant increase in tropospheric ozone in this region since 1996 up to now. This fact has been reported by Kulkarni et al. (2011), who quantify an increase of tropospheric ozone almost 5%e24% over the Iberian Peninsula between the period 1979e1993 and 1996e2005. These authors also reported that the highest values of this tropospheric ozone increase were obtained in the northeaster region of the Iberian Peninsula in accordance with our results. Therefore, this increase of the tropospheric ozone could add to the stratospheric
Fig. 4. Evolution of the annual mean time series from the deseasonalized monthly values for Madrid during the period 1979e2008 using the MSR data. The linear regression trends for the whole period 1979e2008 (black continuous line), and for the sub-periods 1979e1994 and 1995e2008 (dashed-lines) have also been added to the plot.
6288
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290
Fig. 5. Autocorrelation coefficients of the deseasonalized monthly MSR data for Madrid during the period 1979e2008 with a time lag of up to 60 months. The ‘decorrelation’ levels of 0.36 (w1/e) are shown as horizontal black lines.
ozone recovery related to the decline in ODSs over the study region, producing the positive TOC trend found in this work during 1995e2008. Fig. 3 (bottom) shows the ozone trends derived from MSR data over the Iberian Peninsula for the whole period (1979e2008). It can be seen that all trends are negative with similar values between (1.0 0.7)%/decade (Tarifa and Almería) and (1.5 0.7)%/decade (Santander). The trends obtained in this period are statistically significant for the Northerner locations. In addition, there is a slight dependence on latitude in agreement with the spatial distributions of the ozone trends for the first sub-period 1979e1994. These results show that the ozone trend during the whole period is affected larger by the ozone depletion during this first sub-period than by the ozone recovery during the second sub-period. Annually averaged total ozone column is projected to return to 1980 values between 2015 and 2030 over northern midlatitudes, while for southern this return is projected to occur between 2030 and 2040 (WMO, 2011).Around these dates, it is expected a change in the sign of the long-term TOC trend. The evolution of the annual mean time series from the deseasonalized monthly values for Madrid is shown in Fig. 4. The linear regression trends have also been added to the plot which illustrates the apparent change in trend that has occurred over the Iberian Peninsula. The TOC trends are clearly negative during the subperiod 1979e1994, and slightly positive during 1995e2008 as has been explained above. The notable TOC reduction obtained in 1993 is related to the effects of the Mount Pinatubo eruption in June 1991 (e.g., Chandra, 1993; Herman and Larko, 1994). Fig. 4 also shows large inter-annual variability of deseasonalized TOC values, which may make the trend analysis notably sensitive to the length of the data set (Weatherhead et al., 1998). Inter-annual fluctuations of TOC over northern midlatitudes are caused by several factors, such as the regional dynamical structure of the atmosphere, variability of the stratospheric circulation patterns related to anomalously fluxes
of planetary-wave activity from the troposphere, volcanic eruptions, solar flux and the equatorial QBO (Fusco and Salby, 1999; Soukharev, 1999; Appenzeller et al., 2000; Orsolini and Limpasuvan, 2001; Staehelin et al., 2001). This inter-annual variation may produce persistence in the deseasonalized TOC series (Weatherhead et al., 2000; Vyushin et al., 2007; Tandon and Attri, 2010). In this sense, the temporal correlation coefficients (Eq. (3)) were calculated with a time lag of up to 60 months in order to analyze the persistence of the deseasonalized TOC data over the Iberian Peninsula during the period 1979e2008. Fig. 5 shows these autocorrelation coefficients for Madrid. The ‘decorrelation’ levels of 0.36 (w1/e) are also shown as horizontal black lines. It can be seen that the autocorrelation coefficients are below this level for all time lags, indicating that the deseasonalized TOC data over the Iberian Peninsula present no significant autocorrelation. The correlation coefficients for the other individual locations have a very similar behavior (not shown). These results are in agreement with previous works (e.g., Weatherhead et al., 2000; Vyushin et al., 2007) which shown the spatial distributions of the month-tomonth autocorrelation parameter, reporting maximum values at the equator and a notable decreases toward the poles accompanied by some longitudinal variations. Therefore, the influence of the inter-annual fluctuations over the ozone trends obtained in this work is negligible. Nevertheless, the first order autocorrelation coefficient varies from 0.34 (Figueira) to 0.46 (Almeria) which produce an increase of the error in calculated long-term trend (Eq. (5)) by factor of around 1.5. Finally, we obtain the long-term ozone trend separately for each season and location from the deseasonalized monthly values. The results obtained for Santander, Madrid and Tarifa are shown in Table 3. It can be seen that all seasonal ozone trends obtained during the first sub-period are negative in agreement with the strong decline derived from the annual mean time series. These trends are significant at 95% during spring and winter except for the
Table 3 Linear trends in percent per decade (error) for each season at Santander, Madrid and tarifa using the MSR data. Trends statistically significant at the 95% confidence level are marked with asterisk. Trend 1979e1994 (%/decade)
Winter Spring Summer Autumn
Trend 1995e2008 (%/decade)
Santander
Madrid
6.2 7.8 3.2 0.5
5.5 5.8 2.2 0.5
3.1* 2.8* 1.8 1.5
Tarifa 2.6* 2.9* 1.6 1.5
4.0 4.0 1.1 1.4
2.5 4.2 1.8 1.8
Santander
Madrid
þ0.4 þ4.0 þ1.0 þ1.8
þ0.7 þ4.5 þ0.8 þ3.1
3.4 3.9 1.4 1.9
Tarifa 4.6 3.8 1.0 1.6*
þ0.9 þ1.9 0.3 þ1.8
3.9 3.8 1.5 0.9*
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290
Southerner locations. The TOC declines during spring show large values between (4.0 4.2)%/decade for Tarifa and (7.8 2.8)%/ decade for Santander which is consistent with other findings for Northern Midlatitudes (Fioletov et al., 2002; Wohltmann et al., 2007; Antón et al., 2011). In addition, the negative summer trend could justify the observed increase of UV radiation level in the midlatitudes of the Northern Hemisphere in summer months during the past decades (Fioletov et al., 2001; Den Outer et al., 2005). It should be noted that the negative ozone trends derived from spring, summer and winter confirm the latitude dependence (Northerner trends larger than Southerner trends) observed in the annual ozone trends shown above. By contrast, most seasonal ozone trends obtained during the second sub-period are positive but not significant at 95% in accordance with the results inferred from the annual trend analysis. Nevertheless, it can be seen that the spring ozone trends present the largest positive values, opposing to the great negative trends found for this season during the first subperiod. 5. Conclusions Some important conclusions may be drawn from the analysis of the ozone trends using MSR, TOMS and GOME data over the Iberian Peninsula for the period 1979e2008, and two sub-periods (1979e1994 and 1995e2008). The ozone depletion was statistically significant during the first sub-period in this entire region (except in the Southerner locations), with linear trends from 4.5%/ decade to 2.5%/decade. These linear trends presented a clear dependence on latitude, being higher for the Northerner locations than for the Southerner. In addition, the ozone depletion had a distinct seasonal signature, with statistically significant rates observed during winter and spring. By contrast, the analysis of the second sub-period of study (1995e2008) showed positive ozone trends over the Iberian Peninsula, finding the highest values (þ2.5%/decade) in the Northeast of this region. These positive ozone trends were higher than the values reported in the literature for other regions. This behavior was related to the significant increase of the tropospheric ozone over the study region, adding this increase to the stratospheric ozone recovery. Overall we would like to emphasize the remarkable success of the Montreal Protocol for controlling production of ozone depleting substances. Nevertheless, even though the abundance of these substances has significantly diminished in the atmosphere, other anthropogenic effects could complicate the ozone recovery process and may result in considerably altered ozone levels in the future. Therefore, the sustainable measurements of accurate total ozone data with ground-based and satellite instruments are essential to monitor the recovery of ozone layer in the next decades. Acknowledgments The authors would like to thank to the Royal Netherlands Meteorological Institute (KNMI) for providing the Multi Sensor Reanalysis (MSR) data, the German Aerospace Center (DLR) for providing the GOME/ERS-2 ozone data on behalf of ESA, and the NASA TOMS Science Team for providing the TOMS/NIMBUS-7 and TOMS/ METEOR-3 ozone data. Manuel Antón thanks Ministerio de Ciencia e Innovación and Fondo Social Europeo for the award of a postdoctoral grant (Juan de la Cierva). Pavan S Kulkarn is thankful to the Geophysics Centre of the University of Evora (CGE-UE) for the fellowship in the project ‘SPATRAM-MIGE Polar Project’, funded by the Portuguese Science Foundation d FCT. This work was partially supported by the Andalusian Regional Governments through project P10-RNM-6299 and P08-RNM-3568, the Spanish Ministry of Science and Technology through projects CGL2010-18782 and CSD2007-
6289
00067, the European Union through ACTRIS project (EU INFRA2010-1.1.16-262254)., and by Fundação para a Ciência e a Tecnologia though projects: PTDC/AAC-CLI/114031/2009 (FCOMP-01-0124FEDER-014024) and PTDC/CTE-ATM/102142/2008 (FCOMP-010124-FEDER-009303). References Antón, M., Loyola, D., Navascúes, B., Valks, P., 2008. Comparison of GOME total ozone data with ground data from the Spanish Brewer spectroradiometers. Annales Geophysicae 26, 401e412. doi:10.5194/angeo-26-401-2008. Antón, M., Loyola, D., López, M., Vilaplana, J.M., Bañón, M., Zimmer, W., Serrano, A., 2009. Comparison of GOME-2/MetOp total ozone data with Brewer spectroradiometer data over the Iberian Peninsula. Annales Geophysicae 27, 1377e1386. doi:10.5194/angeo-27-1377-2009. Antón, M., Koukouli, M.E., Kroon, M., McPeters, R.D., Labow, G.J., Balis, D., Serrano, A., 2010. Global validation of empirically corrected EP-Total Ozone Mapping Spectrometer (TOMS) total ozone columns using Brewer and Dobson ground-based measurements. Journal of Geophysical Research 115 (D19305). doi:10.1029/2010JD014178. Antón, M., Bortoli, D., Costa, M.J., Kulkarni, P.S., Domingues, A.F., Barriopedro, D., Serrano, A., Silva, A.M., 2011. Temporal and spatial variabilities of total ozone column over Portugal. Remote Sensing of Environment 115, 855e863. Appenzeller, C., Weiss, A.K., Staehelin, J., 2000. North Atlantic Oscillation modulates total ozone winter trends. Geophysical Research Letters 27, 1131e1134. Balis, D., Lambert, J.C., Van Roozendael, M., Spurr, R., Loyola, D., Livschitz, Y., Valks, P., Amiridis, V., Gerard, P., Granville, J., Zehner, C., 2007. Ten years of GOME/ERS2 total ozone data: the new GOME Data Processor (GDP) Version 4: II. Ground-based validation and comparisons with TOMS V7/V8. Journal of Geophysical Research 112 (D07307). doi:10.1029/2005JD006376. Bhartia, P.K., Wellemeyer, C., 2002. TOMS-V8 total O3 algorithm. In: Bhartia, P.K. (Ed.), OMI Algorithm Theoretical Basis Document. OMI Ozone Products, vol. II. NASA Goddard Space Flight Cent., Greenbelt, Md, pp. 15e31. Available at: http:// eospso.gsfc.nasa.gov/eos_homepage/for_scientists/atbd/index.php. Burrows, J.P., Weber, M., Buchwitz, M., Razonov, V., Ladstatter, A., Richter, A., De Beerk, R., Hoogen, R., Bramsdted, D., Eichmann, K.U., Eisenger, M., Perner, D., 1999. The Global Ozone Monitoring Experiment (GOME): mission concept and first scientific results. Journal of Atmospheric Science 56, 151e175. Callis, L.B., Natarajan, M., Lambeth, J.D., Boughner, R.E., 1997. On the origin of midlatitude ozone changes: data analysis and simulations for 1979e1993. Journal of Geophysical Research 102, 1215e1228. Chandra, S., 1993. Changes in stratospheric ozone and temperature due to the eruptions of Mt. Pinatubo. Geophysical Research Letters 20, 33e36. Chen, D., Nunez, M., 1998. Temporal and spatial variability of total ozone in southwest Sweden revealed by two ground-based instruments. International Journal of Climatology 18, 1237e1246. Den Outer, P.N., Slaper, H., Tax, R.B., 2005. UV radiation in the Netherlands: assessing long-term variability and trends in relation to ozone and clouds. Journal of Geophysical Research 110 (D02203). doi:10.1029/2004JD004824. European Spatial Agency: (ESA), 1995. GOME Global Ozone Monitoring Experiment Users Manual. ESA. SP-1182. Farman, J.C., Gardiner, B.G., Shanklin, J.D., 1985. Large losses of total ozone in Antarctica reveal seasonal ClOx/NOx interaction. Nature 315, 207e210. Fioletov, V.E., McArthur, L., Kerr, J.E., Wardle, D.I., 2001. Long-term variations of UV-B irradiance over Canada estimated from Brewer observations and derived from ozone and pyranometer measurements. Journal of Geophysical Research 106, 2307e2309. Fioletov, V.E., Bodeker, G.E., Miller, A.J., McPeters, R.D., Stolarski, R., 2002. Global and zonal total ozone variations estimated from ground-based and satellite measurements: 1964e2000. Journal of Geophysical Research 107 (4647). doi:10.1029/2001JD001350. Fioletov, V.E., Labow, G., Evans, R., et al., 2008. Performance of the ground-based total ozone network assessed using satellite data. Journal of Geophysical Research 113 (D14313). doi:10.1029/2008JD009809. Fusco, A.C., Salby, M.L., 1999. Interannual variations of total ozone and their relationship to variations of planetary wave activity. Journal of Climatology 12, 1619e1629. Grewe, V., 2006. The origin of ozone. Atmospheric Chemistry and Physics 6, 1495e1511. Hadjinicolaou, P., Pyle, J.A., Harrise, N.R.P., 2005. The recent turnaround in stratospheric ozone over northern middle latitudes: a dynamical modelling perspective. Geophysical Research Letters 32 (L12821). doi:10.1029/ 2005GL022476. Harris, N.R.P., Ancellet, J., Bishop, L., Hofmann, D.J., Kerr, J.B., McPeters, R.D., et al., 1997. Trends in stratospheric and free tropospheric ozone. Journal of Geophysical Research 102, 1571e1590. Harris, N.R.P., Kyrö, E., Staehelin, J., Brunner, D., Andersen, S.-B., GodinBeekmann, S., Dhomse, S., Hadjinicolaou, P., Hansen, G., Isaksen, I., Jrrar, A., Karpetchko, A., Kivi, R., Knudsen, B., Krizan, P., Lastovicka, J., Maeder, J., Orsolini, Y., Pyle, J.A., Rex, M., Vanicek, K., Weber, M., Wohltmann, I., Zanis, P., Zerefos, C., 2008. Ozone trends at northern mid- and high latitudes e a European perspective. Annales Geophysicae 26, 1207e1220.
6290
M. Antón et al. / Atmospheric Environment 45 (2011) 6283e6290
Herman, J.R., McPeters, R.D., Larko, D., 1993. Ozone depletion at Northern and Southern latitudes derived from January 1979 to December 1991 Total Ozone Mapping Spectrometer data. Journal of Geophysical Research 98, 12,783e12,793. doi:10.1029/93JD00601. Herman, J.R., Larko, D., 1994. Low ozone amounts during 1992e1993 from Nimbus 7 and Meteor 3 total ozone mapping spectrometers. Journal of Geophysical Research 99, 3483e3496. doi:10.1029/93JD02594. Holton, J.R., Tan, H.-C., 1980. The influence of the equatorial quasi-biennial oscillation on the global circulation at 50 mb. Journal of Atmospheric Sciences 37, 2200e2208. Krzyscin, J.W., 2010. Onset of the total ozone increase based on statistical analyses of global ground-based data for the period 1964e2008. International Journal of Climatology, doi:10.1002/joc.2264. Kulkarni, P.S., Bortoli, D., Salgado, R., Antón, M., Costa, M.J., Silva, A.M., 2011. Tropospheric ozone variability over Iberian Peninsula. Atmospheric Environment 45, 174e182. Labow, G.J., McPeters, R.D., Barthia, P.K., 2004. A Comparison of TOMS & SBUV Version 8 Total Column Ozone Data with Data from Ground Stations, Paper Presented at Quadrennial Ozone Symposium. Eur. Comm., Kos, Greece. Loyola, D., Coldewey-Egbers, M., Dameris, M., Garny, H., Stenke, A., Van Roozendael, M., Lerot, C., Balis, D., Koukouli, M., 2009. Global long-term monitoring of the ozone layer e a prerequisite for predictions. International Journal of Remote Sensing 30, 4295e4318. Loyola, D., Koukouli, M.E., Valks, P., Balis, D.S., Hao, N., Van Roozendael, M., Spurr, R.J.D., Zimmer, W., Kiemle, S., Lerot, C., Lambert, J.-C., 2011. The GOME-2 total column ozone product: retrieval algorithm and ground-based validation. Journal of Geophysical Research 116 (D07302). doi:10.1029/2010JD014675. McPeters, R.D., Bhartia, P.K., Krueger, A.J., 1996. Nimbus-7 Total Ozone Mapping Spectrometer (TOMS) Data Products User’s Guide. NASA Reference Publication, NASA, Greenbelt, Md. McPeters, R.D., et al., 1998. Earth Probe Total Ozone Mapping Spectrometer (TOMS) Data Products User’s Guide. NASA Reference Publication, NASA, Greenbelt, Md. Molina, M.J., Rowland, F.S., 1974. Stratospheric sink for chlorofluoromethanes: chlorine atom catalysed destruction of ozone. Nature 249, 810e812. Newman, P.A., et al., 2009. What would have happened to the ozone layer if chlorofluorocarbons (CFCs) had not been regulated? Atmospheric Chemistry and Physics 9, 2113e2128. Orsolini, Y.J., Limpasuvan, V., 2001. The North Atlantic Oscillation and the occurrences of ozone miniholes. Geophysical Research Letters 28, 4099e4102. Orsolini, Y.J., Doblas-Reyes, F.J., 2003. Ozone signatures of climate patterns over the Euro-Atlantic sector in the spring. Quarterly Journal Royal Meteorological Society 129, 3251e3263. Rodrigo, F.S., Trigo, R.M., 2007. Trends in daily rainfall in the Iberian Peninsula from 1951 to 2002. International Journal of Climatology 27, 513e529. Schmalwieser, A.W., Schauberger, G., Janouchu, M., 2003. Temporal and spatial variability of total ozone content over Central Europe: analysis in respect to the biological effect on plants. Agricultural and Forest Meteorology 120, 9e26. Smedley, A.R.D., Rimmer, J.S., Moore, D., Toumic, R., Webb, A.R., 2010. Total ozone and surface UV trends in the United Kingdom: 1979e2008. International Journal of Climatology. doi:10.1002/joc.2275. Soukharev, B., 1999. On the solar/QBO effect on the interannual variability of total ozone and the stratospheric circulation over Northern Europe. Journal of Atmospheric and Solar-Terrestrial Physics 61, 1093e11091. Staehelin, J., Harris, N.R.P., Appenzeller, C., Eberhard, J., Piechowski, M., 2001. Observations of ozone trends. Reviews of Geophysics 39, 231e290. Stolarski, R.S., Cicerone, R.J., 1974. Stratospheric chlorine: a possible sink for ozone. Canadian Journal of Chemistry 52, 1610e1615.
Stolarski, R.S., Bloomfield, P., McPeters, R.D., Herman, J.R., 1991. Total ozone trends deduced from Nimbus 7 TOMS data. Geophysical Research Letters 18, 1015e1018. Stolarski, R., Bojkov, R., Bishop, L., Zerefos, C., Staehelin, J., Zawodny, J., 1992. Measured trends in stratospheric ozone. Science 256, 342e349. Svendby, T.M., Dahlback, A., 2004. Statistical analysis of total ozone measurements in Oslo, Norway, 1978e1998. Journal of Geophysical Research 109 (D16107). doi:10.1029/2004JD004679. Tandon, A., Attri, A.K., 2010. Trends in total ozone column over India: 1979e2008. Atmospheric Environment 45, 1648e1654. Tung, K.K., Yang, H., 1988. Dynamic variability of column ozone. Journal of Geophysical Research 93, 11123e11128. UNEP, 2006. Handbook for the Montreal Protocol on Substances that Deplete the Ozone Layer. Ozone Secretariat e United Nations Environment Programme. Available online at: http://ozone.unep.org/Publications/MP_Handbook/index. shtml. van der A, R.J., Allaart, M.A.F., Eskes, H.J., 2010. Multi sensor reanalysis of total ozone. Atmospheric Chemistry and Physics 10, 11277e11294. Vanicek, K., 2006. Differences between ground Dobson, Brewer and satellite TOMS8, GOME-WFDOAS total ozone observations at Hradec Kralove, Czech. Atmospheric Chemistry and Physics 6, 5163e5171. van Roozendael, M., Loyola, D., Spurr, R., Balis, D., Lambert, J.C., Livschitz, Y., et al., 2006. Ten years of GOME/ERS-2 total ozone dataethe new GOME data processor (GDP) version 4: 1. Algorithm description. Journal of Geophysical Research 111, D14311. doi:10.1029/2005JD006375. Vyushin, D., Fioletov, V.E., Shepherd, T.G., 2007. Impact of long-range correlations on trend detection in total ozone. Journal of Geophysical Research 112 (D14307). doi:10.1029/2006JD008168. Weatherhead, E.C., Reinsel, G.C., Tiao, G.C., et al., 1998. Factors affecting the detection of trends: statistical considerations and applications to environmental data. Journal of Geophysical Research 103 (D14), 17149e17161. Weatherhead, E.C., Reinsel, G.C., Tiao, G.C., Jackman, C.H., Bishop, L., Frith, S.M.H., DeLuisi, J., Keller, T., Oltmans, S.J., Fleming, E.L., Wuebbles, D.J., Kerr, J.B., Miller, A.J., Herman, J., McPeters, R., Nagatani, R.M., Frederick, J.E., 2000. Detecting the recovery of total column ozone. Journal of Geophysical Research 105 (D17), 22201e22210. Weatherhead, E.C., Andersen, S.B., 2006. The search for signs of recovery of the ozone layer. Nature 441, 39e45. doi:10.1038/nature04746. Wellemeyer, C.G., Bhartia, P.K., Taylor, S.L., Qin, W., Ahn, C., 2004. Version 8 Total Ozone Mapping Spectrometer (TOMS) Algorithm, Paper Presented at Quadrennial Ozone Symposium. Eur. Comm., Kos, Greece. Wohltmann, I., Lehmann, R., Rex, M., Brunner, D., Mäder, J.A., 2007. A processoriented regression model for column ozone. Journal of Geophysical Research 112 (D12304). doi:10.1029/2006JD007573. World Meteorological Organization (WMO), 2003. Scientific Assessment of Ozone Depletion: 2002. Global Ozone Research and Monitoring Project Technical Report 47, Geneva, Switzerland. World Meteorological Organization (WMO), 2007. Scientific Assessment of Ozone Depletion: 2006. Global Ozone Research and Monitoring Project Technical Report 50, Geneva, Switzerland. World Meteorological Organization (WMO), 2011. Scientific Assessment of Ozone Depletion: 2010 Global Ozone Research and Monitoring Project Technical Report 52, Geneva, Switzerland. Yang, S.-K., Long, C.S., Miller, A.J., He, X., Yang, Y., Wuebbles, D.J., Tiao, G., 2009. Modulation of natural variability on a trend analysis of updated cohesive SBUV(/2) total ozone. International Journal of Remote Sensing 30, 3975e3986. doi:10.1080/01431160902821924.