Total: 1
We present a technique which predicts the energy dependent fractional r.m.s for linear correlated variations of a pair of spectral parameters and apply it to an XMM-Newton observation of Mrk 335. The broadband X-ray spectrum can be interpreted as a patchy absorber partially covering the primary emission, a warm and hot coronal emission or a relativistically blurred reflection along with the primary emission. The fractional r.m.s has a non-monotonic behavior with energy for segments of lengths 3 and 6 ksecs. For each spectral model, we consider every pair of spectral parameters and fit the predicted r.m.s with the observed ones, to get the pair which provides the best fit. We find that a variation in at least two parameters is required for all spectral interpretations. For both time segments, variations in the covering fraction of the absorber and the primary power law index gives the best result for the partial covering model, while a variation in the normalization and spectral index of the warm component gives the best fit in the two corona interpretation. For the reflection model, the best fit parameters are different for the two time segment lengths, and the results suggests that more than two parameters are required to explain the data. This, combined with the extreme values of emissivity index and reflection fraction parameters obtained from the spectral analysis, indicates that the blurred reflection model might not be a suitable explanation for the Mrk 335 spectrum. We discuss the results as well as the potential of the technique to be applied to other data sets of different AGN.