A better understanding of the spatio-temporal variations of ozone in the Arctic is crucial for climate change studies. When the vortex breaks down, the perturbations associated with such a yearly and abrupt event have a very strong impact on the spatio-temporal variability of ozone in the Arctic region. Our goal is to determine how the breakdown of the vortex influences the statistical spatio-temporal characteristics of ozone in this region. More precisely, a statistical analysis of the total column ozone from the TOMS Version 7 in the Arctic region over 15 years is undertaken to assess ozone changes. In addition to decompose the ozone field into several parts, a statistical model is developed to take into the vortex breakdown. Our methodology is based on a wavelet multi-resolution method, and it has the advantage of characterizing breakpoint distributions, including not only their timings but also their strengths. Valid covariogram construction for nonstationary spatio-temporal processes will be also used.