We divide the calibration in three steps: Instrumental, wavelength calibration and ETC transfer. Briefly from the know formula of the ozone calculation <br> <math>{O_3} = {{F - ETC} \over {\alpha \,m}}</math>&nbsp; (1,1)

where F is Rayleigh corrected double ratios, α is the ozone absorption coefficient, m is the ozone airmass and ETC is the extraterrestrial constant.

*Instrumental calibration includes all the parameters than affect the measured counts (F): Dead Time, Temperature coefficients and Attenuation of the Filter.

<math>F = \sum\limits_{i = 1}^4 w_i {F_i}\left( {DT,TC,AT} \right)</math>(1.2)

*Wavelength calibration determines the ozone absorption coefficient (<span class=“texhtml”>α</span>).

<math>\alpha = \sum\limits_{i = 1}^4 {{w_i}\alpha ({\lambda _i})}</math>(1.3)

*ETC transfer is determined by comparison with the reference. Changes in instrumental calibration and wavelength calibration affect the final ETC.

For the comparison with the reference we classify the instrument in two types depending on if they where maintenance works or not. The instrument can have modifications during maintenance works, so we divide the period of the campaign in three periods, initial blind days to check the status of the instrument, days for maintenance and finally the final days to set the new calibration. For the instruments without maintenance the time periods are the same (Table 1).

<br>

{| style=“border-spacing: 0pt;”

Table 1&nbsp;: Calibration Periods

<br> For all the instruments we have the initial calibration and the final one, which can be the same. In some instruments you can find also a third calibration, this is a suggested one for the period before the campaign. Most of the data of calibration are on ICF files with some remarkable exceptions like SL reference value.

<br>