Scaling Analysis of the Flare Index Data from Kandilli Observatory

The daily time series Flare Index (FI) data of Northern Hemisphere, Southern Hemisphere and Total Disk for Solar Cycle 21- 23 and 24 up to Dec. 2014 has been pre-processed using a 2nd order exponential smoothing algorithm to remove orthogonal noise. The smoothed data in each case is processed for scaling analysis using Rescaled-Range Analysis as well as Finite Variance Scaling Method in order to search for the Hurst exponent. As the value of H obtained from our analysis lies in between 0 and 1, so it can be said that the signal may behave like Fractional Brownian Motion. Also, it is observed that H is less than 0.5 which indicates the data is anti-persistent in nature and it has a strong negative correlation within the signal. The value of H also indicates the oscillating features of the signal which might have some fundamental periodicities in the Suns atmosphere.