Total column ozone is measured by recording the differential absorption of solar light intensity in the UV band. The Microtops II Ozonometer (a sun targeted photometer) measure the direct solar radiation in five channels using a collimator with 2.5˚ field of view. For each channel a photodiode and narrow band interference filters for the three UV wavelengths 305.5 nm, 312.5 nm and 320.0 nm are dedicated for the total column ozone measurements. The near infrared bands centered at 940 nm and 1020 nm are used to retrieve the water vapor and aerosol optical depth. The Microtops II measures the optical block temperature and it also incorporates a solid state pressure sensor to provide the surface pressure for each measurements. The physical and operational methods of Microtops II instruments were found elsewhere [27] [28] . The performance of Microtops II is validated using the Brewer spectrometer, which is considered as the standard instrument for measuring total column ozone. Brewer Spectrometer points towards sun and measure radiance at six wavelengths viz 303.2, 306.3, 310.1, 313.5, 316.8 and 320.1 nm. The FOV is approximately 3˚ and spectral resolution is 0.6 nm. Measurements are carried out manually by pointing the instrument against the sun with the help of light indicator which reflects the sun position in the sun targeted window.

The ozone measurement using Microtops II ozonometer is based on the assumption that intensity of radiation measured at each channel obeys Lambert-Beer Law, which when applied to ozone absorption and Rayleigh scattering by the atmosphere gives intensity as

I_o is the intensity of light of a particular wavelength before it passes through the atmosphere, I, the inten- sity remaining after all processes attenuating the incident radiation have occurred, α is the ozone absorption coefficient at that specific wavelength, Ω is the amount of ozone, μ the ratio of the actual and vertical path lengths of the radiation through the ozone layer, P is the pressure of the atmosphere in hPa, m is a quantity known as the air mass, P_o is the standard pressure (1013.25 mbar). For m < 2 the parameters μ and m are virtually identical. The Microtops II uses the combination of three wavelengths for calculation of the TCO. The channel I (Ch_12TCO―combination of wavelengths 305.5 nm and 312.5 nm), Channel II (Ch_23TCO―combination of 312.5 nm and 320.0 nm) and Chanel III (Ch_123TCO―combination of 305.5 nm, 312.5 nm and 320.0 nm) calculate TCO based on the Equation (2), where α₁₂ and α₂₃ are difference between the ozone absorption coefficients for the corresponding filters.

The measurements of TCO, surface temperature, and pressure taken at 11.30 hrs and 14.30 hrs were averaged to get the daily values. Monthly means are calculated using the daily means to study the seasonal variability. The data gaps were filled using the linear interpolation technique. The tropical monsoon rainfall and the monsoon regions are undergoing a lot of intra-seasonal variability with different time scales which can alter the weather and climate, which in turn produce extreme weather events frequently. Considering the above facts, efforts are taken in this work to understand the different time scales of ISOs which play significant role in the total intra- seasonal variability. The daily total column ozone, surface temperature and pressure were taken for the period 01 June-30th September, 2015 to study the ISV.

Wavelet Transform converts a signal into a series of wavelets, which provide a way for analyzing waveforms, bounded in both frequency and time domain. It allows signals to be stored more efficiently than Fourier transform and is very well-suited for approximating data with sharp discontinuities. Wavelet transform can be used effectively to extract the hidden frequency based information from the raw signal which is normally present in the time-domain. Basically the frequency components (spectral components) of the signal show what frequencies exist in the signal. Fourier Transform can tell how much of each frequency exists in a signal and it cannot provide what frequency exists at what time intervals. Wavelet transform can overcome this issue and hence it is widely used for climate data analysis. In the wavelet analysis raw signal (time-domain) decomposes into different frequency components. A Discrete Wavelet Transform (DWT) provides sufficient information both for analysis and synthesis to reduce the computation time sufficiently. It is easier to implement and analyze the signal at different frequency bands with different resolutions. Discrete wavelet transform decompose the signal into a coarse approximation but with detailed information. There are different families of wavelets with different properties suitable for different applications. In this study Mayer’s discrete wavelet analysis is used which is not a crude method and have many advantages over other wavelet transform methods. To study the low and high frequency oscillations (ISOs) during the monsoon season, an advance statistical method called Discreet Mayer’s Wavelet (DMW) is used. Mathematical steps, algorithms and software packages are available for the easy use of this technique for the climate data analysis [29] .

To study the seasonal variation of TCO, surface temperature and pressure, daily in situ observations were taken using the hand held Microtop II Ozonometer. Many extreme weather events (increase in frequency of tropical cyclones with hurricane speeds, drought and flood situations) and climate variability has happened across globe in the year 2015. India witnessed the extremes in the weather due to tele-connections of strong El Niño events at the pacific and associated weather changes in the rainfall amount during 2015.

The present work mainly concentrated on the seasonal and intraseasonal study of total column ozone (TCO) measured over Cochin with other weather parameters such as daily surface temperature and pressure at 11.30 hrs and 14.30 hrs. Figures 1(a)-(c), represent the daily mean of monthly values subtracted from the annual mean

(279.7 DU) of TCO, surface temperature (28.6˚C) and pressure (1009.5 hPa) respectively. Significant seasonal variations are noted in the above parameters for the year 2015. Maximum variations in the negative TCO anomaly are noted in the months November to February with a peak (−17 DU) in January 2015. From March to September there observed a steady increase in the Total column ozone anomaly and reached a peak value of +15 DU in September. From October onwards there was sudden decline in TCO anomaly to −15 DU within a gap of two months (October-November) from the annual mean. From earlier studies it is observed that total ozone variability is high over tropical Indian monsoon region especially the stations close to the proximity of equator [30] . The above results are true when compared with the 35 year climatology of ECMWF reanalysis of TCO [36] .

Interannual variability of TCO is different due to intra-seasonal variability (ISV) of the respective years. During monsoon season (June-September) TCO shows a little higher value from the annual average (279.7 DU) with an increase of ~9 DU, which is quite relevant when the tropical ozone variability is concerned. The pre monsoon season (March-May) shows a little lower value (283.7 DU) than the annual average and post monsoon season (October-December) shows still lower value of 274.4 DU (difference of ~14 DU from the monsoonal average and ~5 DU from the annual average). During March-April months, ozone anomaly show very less values (>5 DU) which indicates the thinning of ozone layer and this may lead to more UVB radiation reaching the earth surface over the study location. In the last few years, a number of sunburn cases have been reported over this region and hence this sort of study is quite important when we consider the regional importance of Cochin. From the Figure 1 it is quite evident that TCO varies about 30 DU from January to December period of the year 2015. Even though this value of TCO variation is less compared to that in the high and polar latitudes, this change in value is quite relevant considering the tropical weather and climate.

Figure 1(b) & Figure 1(c), represents the seasonal variations of surface temperature and pressure over Cochin for the year 2015. The annual and seasonal average of surface temperature and pressure are calculated using the daily values measured by Microtops II. Two observations were made daily at 11.30 hrs and 14.30 hrs and the average value of these two observations are taken as daily value of TCO, Temperature and pressure. The monthly surface temperature fluctuate between −1˚C and +1˚C from the annual average (28.6˚C) with difference of 1.8˚C from maximum to minimum during 2015. That is temperature change is about 2˚C only throughout the year. In monsoon and post monsoon seasons surface temperature varies in a similar fashion which is almost equal to the annual average (28.6˚C). Pre monsoon shows a little lesser value of surface temperature (27.8˚C) with a difference of ~0.8˚C from the annual average. During monsoon period surface temperature fluctuates between −0.3˚C and +0.3˚C and for other seasons, it is greater than 0.3˚C from the annual mean. Another interesting observation is that temperature fluctuates between negative and positive anomaly in the alternate months during summer monsoon period June to September.

Figure 1(c) shows the monthly surface pressure anomaly over Cochin. Surface pressure oscillates between −2 hPa and +2 hPa from the annual mean (1009.5 hPa) for the entire seasons of 2015. The highest surface pressure anomaly is recorded in December (~+2 hPa) and the lowest in June (~−2 hPa). The surface pressure over Cochin during January to March shows a positive anomaly, but from April to June it shows a negative anomaly with peak value in the month of June. This pressure deficient brings the south west monsoon current over Cochin. Hence the intra seasonal studies surface weather parameters would help in predicting the onset of monsoon over Kerala (normal onset date is June 1), the gate way of Indian summer monsoon. During the monsoon the pressure anomaly changes from −2 hPa (June) to +1 hPa (September). This pressure deficit during monsoon season shows the strength of monsoon circulation, ISV of monsoon current and the associated rainfall pattern over the Indian subcontinent. The surface pressure anomaly during monsoon clearly shows the intra-seasonal variability (ISV), which is connected to the more rainy days (active cycles) and less rainy days (break cycles) of Indian summer monsoon season and associated amount of rainfall variability. There are studies available for defining the active and break cycles of Indian summer monsoon, which is of prime importance and concern in the short term prediction of monsoon rainfall [31] .

The ISV of the tropical TCO are large (~±10 DU) and comparable to the total TCO variations in annual and interannual time scales associated with the annual cycle, quasi-biennial oscillation, El Niño-Southern Oscillation, and solar cycle. Insufficient knowledge in the ISV is a major challenge faced by the scientific community in the short term weather prediction and forecast. Indian economy is based on agriculture and hence the prediction and timely forecast is very much essential for the farmers to have safer way of doing agricultural activity, which can reduce the economic loss of the country. The study of ISV is very important in the prediction of active and breaks cycles of Indian summer monsoon which is linked to the extreme weather events. The long durations of active and break days can lead to the weather extremes such as floods and drought events. Hence the accurate short term prediction and forecast of weather systems are prime concern for the country to reduce the economic loss.

In this work we have made an attempt to understand the ISV of TCO, surface temperature and pressure during monsoon season and the results were presented in the Figures 2(a)-(d). Figure 2(a) represents the ISV of Indian summer monsoon rainfall, which is adopted from the research report [32] and Figures 2(b)-(d) represent the ISV in TCO, Surface Temperature and Pressure respectively. The ISV in the weather parameters TCO, surface temperature and pressure is clearly understood from the Figure 2. In the figure x-axis denotes the number of days from 1 June to 30th September 5 (June: 1 - 30, July: 31 - 60, August: 61 - 90, and September: 91 - 122). Figure 2(b) represent the ISV in the TCO during monsoon period, that is daily departure from the monsoonal mean (288.8 DU). The intra seasonal variability in the TCO values is noted and in the month of June-July, there is a negative ozone anomaly noted over this region with very short time scale variations. In many of the days in the beginning of this season, the TCO anomaly deviate −10 DU from the seasonal mean value. But in August, the anomaly varies from −5 DU to +5 DU. A positive TCO anomaly variation is noted during almost all the days in the month of September (see Figure 1(a)) and in the last few days the TCO anomaly crossed +10 DU from the seasonal average. Figure 2(c) and Figure 2(d) shows the ISV in the surface temperature and pressure over Cochin. The ISV in the surface temperate anomaly varies between −1.5˚C and +1.5˚C from the seasonal average (28.6˚C) and the surface pressure varies between −2.5 hPa and +2 hPa as respectively (Figure 2(d)). Table 1 shows the statistics of ISV in TCO, surface temperature and pressure over Cochin during summer monsoon 2015. There is no correlation of TCO with surface temperature and a positive correlation of 0.35 is noted with surface pressure.

The MJO consists of two phases, one is the enhanced phase and the other is the suppressed phase. These two phases produce opposite changes in clouds, rainfall, surface pressure, temperature, and even in the atmospheric composition [16] . The persistence of 40 - 50 day oscillations in stratospheric ozone and temperature was noted in earlier studies [33] . There is also evident that the MJO propagates even higher into the ionosphere [34] . The characteristics of intraseasonal variability are mainly determined by the low and high frequency Intra-seasonal oscillations (ISOs) present during the season. The dynamics and chemistry of the tropical atmosphere is mainly modulated by the ISOs. Hence an understanding of the low and high frequency of ISOs during the season gives a better picture of intra-seasonal variability. The ISOs with different periodicities in the season which modulate the weather and climate over a region implies that ISV is a collective action of ISOs. In order to study the ISV in TCO, surface temperature and pressure over Cochin during the monsoon period, we used Discrete Mayer’s Wavelet, techniques which decomposes all possible and significant ISOs present in the above parameters. Hence ISV can be considered as the sum and collective actions of different modes of ISOs present in the particular season. For the wavelet analysis daily values of TCO, surface Temperature and pressure from 29th May to 3rd October 2015 (128 data points) were used and the significant ISOs are shown in Figure 3 to Figure 5 respectively. The x-axis represents the days and y-axis represents the wavelet coefficients for the different modes of ISOs.

Figure 3(a) represents the TCO in Dobson Unit and Figures 3(b)-(f) shows the different possible low and high frequency ISOs during summer monsoon period. The prominent modes of ISOs are MJO mode (30 - 60 days), quasi triweekly mode (18 - 22 days) and quasi bi-weekly (10 - 16 days) mode and weekly mode (6 - 8 days). The other modes with lower periodicities are insignificant in contributing to the ISV. The dominant mode of ISO present in the TCO ISV is MJO mode having 2.5 waves with a periodicity of approximately 48 days during the monsoon season (Figure 3(b)). The phases of MJO modulate the ISV of TCO during the period. It can be seen from the figure that during the positive and negative phases of MJO a positive and negative anomaly of TCO is observed with minor exceptions (see Figure 2(b)).

The second major ISO is having a periodicity greater than quasi-biweekly oscillations and we call it as triweekly oscillations (18 - 22 days) with almost 6 waves during the season (Figure 3(c)). The positive and negative phases of this triweekly ISO contribute to the ISV with a positive TCO anomaly during positive phase and

negative TCO anomaly during negative phases (see Figure 2(b)). Figure 3(d) represent the ISO mode with 10 - 16 days periodicity called quasi biweekly oscillations. This mode also contributes to the ISV of TCO. The amplitude of tri-weekly and quasi bi-weekly ISOs are not symmetric. Some waves are more amplified at certain days during the period of study which can determine the amount of ISV variability of ozone (Figure 3(c) and Figure 3(d)). It is also observed that if the amplitude is high for particular wave at particular time of an ISO, corresponding increase (positive phase) and decrease (negative phase) of TCO anomalies are noted during that time with minor exceptions. Figures 3(d)-(f) contribute least to the ISV and it can be considered as simple perturbations or noise. In the ISV of TCO, there are three (MJO, triweekly and biweekly ISOs) modes really contributing to the intraseasonal variability (ISV) of TCO during the monsoon period.

Figure 4(a) represent the surface temperature (˚C) measured over Cochin and Figures 4(b)-(f) shows the different modes of ISOs present in the surface temperature. Figure 4(b), represent the dominant ISO mode in the temperature ISV. This mode also comes under the MJO periodicity (26 - 36 days). This periodicity is less than the MJO periodicity observed in the TCO. Here total three waves observed but the amplitude of oscillations are not symmetric and the amplitude varies with different period of time during the season.

The phases of MJO modulates the surface temperature over Cochin during monsoon season with +ve phase showing a positive surface temperature anomalies and negative phase with negative temperature anomalies (Figure 4(b) and Figure 3(c)). The two (MJO, and quasi-biweekly modes) ISO periodicities are significant and no tri-weekly modes were present in the surface temperature ISV. The other modes of ISO shown in Figures 4(d)-(f) do not contribute much to the ISV of temperature. In the studies of intra-seasonal oscillations of Indian summer monsoon rainfall, these two modes were already reported [12] [35] .

Figure 5(a) represent the variation of surface pressure during the monsoon period and Figures 5(b)-(f), represent the different ISO modes present in the surface pressure. These ISOs modulate the pressure anomalies over the region. Figure 5(b) represents the negative and positive phases of ISOs which correlate with negative and positive pressure anomalies (see Figure 2(d)). The two prominent modes of ISOs in surface pressure are MJO mode (30 - 36 days) and quasi biweekly modes (10 - 16 days). The ISO modes keep an out of phase

relation with Indian summer rainfall during this season (Figure 2(a)). The positive phase of the ISO fall in the break cycles of the Indian summer monsoon rainfall and opposite is true for the active cycles. Thus more case studies are needed to understand the intraseasonal oscillation for the proper predictions of ISV, which in turn affect forecast of extreme weather events. Hence more case studies are required for the better conclusion.