Institute of Geophysics, University of TehranJournal of the Earth and Space Physics2538-371XArticles in Press20220419Temporal variability analysis of measured surface ozone at the Geophysics Institute Station of the Tehran UniversityTemporal variability analysis of measured surface ozone at the Geophysics Institute Station of the Tehran University8690610.22059/jesphys.2022.329346.1007355FANajmehKaffashzadehSpace Physics, Geophysics Institute, University of TehranAbbasAliAliakbaribidokhtiUniversity of TehranJournal Article20210829Near surface ozone (O$_{3} ^{surf}$), or tropospheric ozone at the ground level, is a secondary air pollutant that deteriorates human health and plants via damaging respiratory systems. This species is one of the main greenhouse gases associated with global warming and climate change. Despite many efforts to study and to make policy control program, this gas is still increasing and is a recent serious threat for human. So, a comprehensive understating of its variation and controlling factors is necessary for having a precise plan for its regulation.<br />
Here, a measured time series of O$_{3} ^{surf}$ at one of the air quality monitoring sites in Iran, i.e. Geophysics Institute of the University of Tehran, was selected to assess the O$_{3} ^{surf}$ variation in more detail. Although this time series has been measured since 2007, there are many gaps in the data and a few years without data. Nevertheless, the data possess a high quality which has been discussed in this paper. The series was prepared for the period of four years, i.e. 2007-2008 and 2019-2020. <br />
The data series was decomposed to five spectral components, i.e. intraday (ID), diurnal (DU), synoptic (SY), seasonal (SE), and baseline (BL), by applying Kolmogorov-Zurbenko (KZ) filter. This filter introduced by Kolmogorov and later formalized by Zurbenko in 1997. Theoretically, the KZ filter is a technique consists of iterative running moving average (MA), in which a simple MA of m points is computed by:<br />
S(t) = $frac{1}{m}$ $sum_{j=-(m-1)/2}^{(m-1)/2} ORG(t)_j$<br />
where ORG and t represent the original time series and its time steps, respectively, and S is the input for each iteration. Therefore, the filter can be express as:<br />
KZ$_{m,k}$= R$_{i=1} ^{k}$ {J$_{p=1} ^{w_i}$ [S(t${_i}$)$_{p}$]}<br />
Here m and k are window length and number of iterations, respectively. R and J represent iteration and running window, respectively, and w$_{i}$ is defined as: <br />
W$_{i}$ = L$_{i}$ - m + 1<br />
where L$_{i}$ is the length of S(t$_{i}$). KZ$_{m,k}$ is a low pass filter in which high frequency (short time period) variation are removed from the time series. The band of frequency and the level of suppression in this filter are controlled by m and k, respectively. Here, the ozone time series was decomposed to five spectral components as:<br />
ORG(t) = ID(t$_{<12h}$) + DU(t$_{12h-2.5d}$) + SY(t$_{2.5d-21d}$) + SE(t$_{21d-365d}$) + BL(t$_{>365d}$)<br />
The results indicate that the contribution of each component to the O$_{3} ^{surf}$ variability is different as such the DU component constitutes more than 50% of the ozone variability. In fact, this component makes most of the ozone variability which attributes to light variation (daytime-nighttime). The SE component has the second largest contribution to the O$_{3} ^{surf}$ variability. The contribution of the SY component is different and that depends on the year. As an example, the relative contribution of this component in 2007 is 8.93% and in 2019 is 4.84%. Only 5% of the total O$_{3} ^{surf}$ variability makes by the variation of the ID component. This implies that the contribution of each component to the total O$_{3} ^{surf}$ variability is different and this information should be considered in ozone control strategies.Near surface ozone (O$_{3} ^{surf}$), or tropospheric ozone at the ground level, is a secondary air pollutant that deteriorates human health and plants via damaging respiratory systems. This species is one of the main greenhouse gases associated with global warming and climate change. Despite many efforts to study and to make policy control program, this gas is still increasing and is a recent serious threat for human. So, a comprehensive understating of its variation and controlling factors is necessary for having a precise plan for its regulation.<br />
Here, a measured time series of O$_{3} ^{surf}$ at one of the air quality monitoring sites in Iran, i.e. Geophysics Institute of the University of Tehran, was selected to assess the O$_{3} ^{surf}$ variation in more detail. Although this time series has been measured since 2007, there are many gaps in the data and a few years without data. Nevertheless, the data possess a high quality which has been discussed in this paper. The series was prepared for the period of four years, i.e. 2007-2008 and 2019-2020. <br />
The data series was decomposed to five spectral components, i.e. intraday (ID), diurnal (DU), synoptic (SY), seasonal (SE), and baseline (BL), by applying Kolmogorov-Zurbenko (KZ) filter. This filter introduced by Kolmogorov and later formalized by Zurbenko in 1997. Theoretically, the KZ filter is a technique consists of iterative running moving average (MA), in which a simple MA of m points is computed by:<br />
S(t) = $frac{1}{m}$ $sum_{j=-(m-1)/2}^{(m-1)/2} ORG(t)_j$<br />
where ORG and t represent the original time series and its time steps, respectively, and S is the input for each iteration. Therefore, the filter can be express as:<br />
KZ$_{m,k}$= R$_{i=1} ^{k}$ {J$_{p=1} ^{w_i}$ [S(t${_i}$)$_{p}$]}<br />
Here m and k are window length and number of iterations, respectively. R and J represent iteration and running window, respectively, and w$_{i}$ is defined as: <br />
W$_{i}$ = L$_{i}$ - m + 1<br />
where L$_{i}$ is the length of S(t$_{i}$). KZ$_{m,k}$ is a low pass filter in which high frequency (short time period) variation are removed from the time series. The band of frequency and the level of suppression in this filter are controlled by m and k, respectively. Here, the ozone time series was decomposed to five spectral components as:<br />
ORG(t) = ID(t$_{<12h}$) + DU(t$_{12h-2.5d}$) + SY(t$_{2.5d-21d}$) + SE(t$_{21d-365d}$) + BL(t$_{>365d}$)<br />
The results indicate that the contribution of each component to the O$_{3} ^{surf}$ variability is different as such the DU component constitutes more than 50% of the ozone variability. In fact, this component makes most of the ozone variability which attributes to light variation (daytime-nighttime). The SE component has the second largest contribution to the O$_{3} ^{surf}$ variability. The contribution of the SY component is different and that depends on the year. As an example, the relative contribution of this component in 2007 is 8.93% and in 2019 is 4.84%. Only 5% of the total O$_{3} ^{surf}$ variability makes by the variation of the ID component. This implies that the contribution of each component to the total O$_{3} ^{surf}$ variability is different and this information should be considered in ozone control strategies.