Table 2 conveys information about the number of possible words that
appeared each year at each station. As can be understood, during the
study period (2002-2017), in 2002, 2007, 2008, and 2015 only one pattern
of words was observed (i.e. 00) which means that the maximum discharge
in each single day throughout those years was less than the threshold.
In contrast, in 2004, 2006, 2010, and 2017, four different words were
observed, telling that the occurred floods were greater than the
threshold value and high enough so that persisted for at least two
consecutive days. Table 3, on the other hand, reports the metric entropy
and the effective measure complexity contents. The results in Table 3
shows that in 2002, 2007, 2008, and 2015, the computed metric entropy
was 0, which means that there was no entropy (randomness), and hence,
there is neither information nor complexity. It can be distinguished
that if three different words were observed, it yields that at least one
day flood event occurred, correspondingly, the other values of
information and complexity contents give additional descriptions about
the frequency and amount of the occurred floods. Generally, during the
years of this study, three different words were reported (i.e. 00, 01,
10), and in some years 4 days indicating that floods in this mountainous
region are an important challenge and hence there are additional actions
must be taken to mitigate and minimize flood implications.
Finally, to elucidate the efficiency of this method, let’s consider the
results recorded at the Ozekiyama station as an example. In one hand, in
2004, it was reported that three flood events were recorded as (11) and
another three events as (01) and similarly three events as (10). The
estimated metric entropy and effective measure of complexity as can be
seen in Table 3 are 0.1 and 0.07, respectively. on the other hand, in
2006, it was documented that two flood events were recorded as (11) and
three flood events as (01) also another three events as (10). In this
case, the metric entropy and the effective measure of complexity as
estimated in Table 3 are 0.09 and 0.03 respectively, which reflect the
merit of this approach to sort out the different information about the
documented floods.