We analyse time series of velocity fluctuations of the low-speed stream of
the solar wind measured by the Helios spacecraft in the inner heliosphere.
We use a nonlinear filter to give faithful representation of nonlinear beha
viour of the flow. We demonstrate that the influence of noise in the data r
ecords can be much more efficiently reduced by a nonlinear filter than with
conventional filters. We argue that due to this nonlinear noise reduction
we get with much reliability estimates of the largest Lyapunov exponent and
the Kolmogorov entropy. The Lyapunov exponent and the entropy are plausibl
y positive, which would exhibit sensitivity to initial conditions. These re
sults show that the solar wind in the inner heliosphere is likely a determi
nistic chaotic system. We hope that these studies could shed light on the p
hysical mechanism of coronal structure. (C) 2001 Elsevier Science Ltd. All
rights reserved.