Chaos theory discusses unstable and non-periodic behavior of non-linear fluctuating dynamic
systems. Since evaporation level variations of lakes have a dynamic identity, chaos theory can have a unique
role in collecting data of these variations. Therefore it is an important and challenging problem in the field of
planning and qualitative and quantitative management of Urmia Lake to verify whether the evaporation time
series are stochastic or chaotic that is discussed in the present study. In addition to introduction of different
utilities of chaos theory, the monthly evaporation amounts of Urmia Lake in the past 40 years have been
studied and predicted in the present research. So after calculating the delay time (T=7.5) by using average
mutual information method (AMI) and embedding dimension (d=3) by using false nearest neighbor algorithm
(FNN), the slope of correlation dimension diagram has been computed. The non-integer amount of the slope
(2.47) represents that the system is chaotic. Lyapunov exponent and broad band in Fourier power spectrum
are other indexes reported in the present study and their provided results ensures that the system is chaotic.
Thus the amount of Urmia Lake evaporation is predictable. Therefore the amount of evaporation in the recent
10 years (1997-2007) have been predicted by means of false nearest neighbor algorithm and verified with the
observed data. The results agree with the high accuracy of chaos theory predictions so the amount of evaporation
of the Lake is predicted for 10 following years (2007-2017).