Development of a wavelet based tool for variability statistical analysis

It can sometimes be problematic to distinguish between variations in data on a light curve due to interesting occurrences and statistical or background noise. By using a simple Haar wavelet transform on the data with error propagation through the coefficients calculation, it becomes possible to establish the confidence level with which variability at specific time scales is observed. Advantages and disadvantages of the Haar wavelets over other basis are discussed and example applications are presented.