The expected time of impact, also known as the mean first passage time
(MFPT) to reach failure, is a critical metric in the management of
natural disasters. The complexity of the dynamics governing natural
disasters lead to stochastic behaviour. This book shows that state
transitions of many such systems translate into random walks on their
respective state spaces, biased and shaped by environmental
inhomogeneity. Thus the probabilistic treatment of those random walks
gives valuable insights of expected behaviour. A comprehensive case
study of predicting cyclone induced flood is followed by a discussion of
generic methods that predict MFPT addressing directional bias. This is
followed by discussing MFPT prediction methods in systems showing
network inhomogeneity. All presented methods are illustrated using real
datasets of natural disasters. The book ends with a short discussion of
possible future research areas introducing the problem of predicting
MFPT for bush-fire propagation.