RNA free energy landscapes are analysed by means of ''time-series'' th
at are obtained from random walks restricted to excursion sets. The po
wer spectra, the scaling of the jump size distribution, and the scalin
g of the curve length measured with different yard stick lengths are u
sed to describe the structure of these ''time series''. Although they
are stationary by construction, we find that their local behavior is c
onsistent with both AR(I) and self-affine processes. Random walks conf
ined to excursion sets (i.e., with the restriction that the fitness va
lue exceeds a certain threshold at each step) exhibit essentially the
same statistics as free random walks. We find that an AR(I) time serie
s is in general approximately self-affine on timescales up to approxim
ately the correlation length. We present an empirical relation between
the correlation parameter rho of the AR(I) model and the exponents ch
aracterizing self-affinity. (C) 1996 Academic Press Limited