226
distributaries and turning them into dried bed locally known as Kana Nadi, e.g.
Ghia, Behula, Kantool and Kana Damodar (Bagchi 1977; Bhattacharyya 2011;
Ghosh 2011). According to Lahiri-Dutt (2012), the sufferings of the people living in
the lower part of this river valley never really diminished. Therefore, the present
flood protection measures have failed to protect people from the devastating effects
of floods; there is an urgent need for new solutions for the flood management programs (Mukhopadhyay 2010). Thus, more resource and skilled manpower are
essential to handle the situation (Ali 2007).
There are so many types of flood, but riverine flood is the most common and
dominant one in this region. It has become an annual phenomenon in the eastern
parts of India, especially in West Bengal, where 55.43% area is flood prone
(Mukhopadhyay 2010; Roy 2012). Floods in the deltaic region have the capacity to
destroy the environmental setting of the region (Jha and Bairagya 2011). Among the
global trends of natural disasters, Asia is becoming increasingly vulnerable (Shaw
and Krishnamurthy 2009). According to Central Water Commission, Government
of India, on an annual average, 7.21 million hectares of land is inundated, and nearly
32 million people are affected by floods (Kale 2003; Bandyopadhyay et al. 2016).
Formulation of an effective flood management strategy in developing countries like
India is very essential, because flood happens very frequently over large parts of the
country (Sanyal and Lu 2004; Nath et al. 2008; Pandey et al. 2010).
Remote sensing and GIS has become an important tool to prepare the maps for
the inundated areas and to assess the damages moreover to suggest the most suitable
development plan to reduce the intensity of flood (Anselmo et al. 1996; Islam et al.
2001; Jain and Sinha 2003; Jain et al. 2005; Bera et al. 2012; Masood and Takeuchi
2012). Updated floodplain maps play an important role to avoid several social and
economic losses during the floods (Sanyal and Lu 2003). Identification of floodprone properties for timely warning is essential to reduce the damages (Schumann
2011). Government organizations can take remedial actions before the disaster if
proper information is available to them (Godschalk 1991; Ali 2007).
Escalante-Sanboval and Raynal-Villasenor (1998) used the trivariate gumbel distribution for estimating the frequency of flood. In this purpose, the logistic model
has been taken into consideration for trivariate analysis. The parameters of this distribution have been derived numerically for the complexity of likelihood functions,
which has been in the form of estimators. They concluded that the trivariate distribution has been established as a reliable tool for flood frequency analysis. Rahman
et al. (2013) used the site-specific flood frequency, which is based on the probability
distribution of long period stream flow data. In this research, the choice of appropriate probability distribution in terms of the nature of the data sets is more vital for
finding out the actual scenario. Apart from this, region-specific difference has been
found in terms of the appropriate probability distribution. Bezak et al. (2014) used
annual maximum (AM) and peaks over threshold (POT) to compare better accuracy
for estimating the frequencies of the flood. They have identified that the POT method
is more realistic than the AM method. In this case, the Poisson distribution is more
realistic than the Binomial distribution, considering the annual threshold limit.
Arnaud et al. (2016) used the stochastic simulation which is based on hourly rainfall
S. C. Pal et al.
distributaries and turning them into dried bed locally known as Kana Nadi, e.g.
Ghia, Behula, Kantool and Kana Damodar (Bagchi 1977; Bhattacharyya 2011;
Ghosh 2011). According to Lahiri-Dutt (2012), the sufferings of the people living in
the lower part of this river valley never really diminished. Therefore, the present
flood protection measures have failed to protect people from the devastating effects
of floods; there is an urgent need for new solutions for the flood management programs (Mukhopadhyay 2010). Thus, more resource and skilled manpower are
essential to handle the situation (Ali 2007).
There are so many types of flood, but riverine flood is the most common and
dominant one in this region. It has become an annual phenomenon in the eastern
parts of India, especially in West Bengal, where 55.43% area is flood prone
(Mukhopadhyay 2010; Roy 2012). Floods in the deltaic region have the capacity to
destroy the environmental setting of the region (Jha and Bairagya 2011). Among the
global trends of natural disasters, Asia is becoming increasingly vulnerable (Shaw
and Krishnamurthy 2009). According to Central Water Commission, Government
of India, on an annual average, 7.21 million hectares of land is inundated, and nearly
32 million people are affected by floods (Kale 2003; Bandyopadhyay et al. 2016).
Formulation of an effective flood management strategy in developing countries like
India is very essential, because flood happens very frequently over large parts of the
country (Sanyal and Lu 2004; Nath et al. 2008; Pandey et al. 2010).
Remote sensing and GIS has become an important tool to prepare the maps for
the inundated areas and to assess the damages moreover to suggest the most suitable
development plan to reduce the intensity of flood (Anselmo et al. 1996; Islam et al.
2001; Jain and Sinha 2003; Jain et al. 2005; Bera et al. 2012; Masood and Takeuchi
2012). Updated floodplain maps play an important role to avoid several social and
economic losses during the floods (Sanyal and Lu 2003). Identification of floodprone properties for timely warning is essential to reduce the damages (Schumann
2011). Government organizations can take remedial actions before the disaster if
proper information is available to them (Godschalk 1991; Ali 2007).
Escalante-Sanboval and Raynal-Villasenor (1998) used the trivariate gumbel distribution for estimating the frequency of flood. In this purpose, the logistic model
has been taken into consideration for trivariate analysis. The parameters of this distribution have been derived numerically for the complexity of likelihood functions,
which has been in the form of estimators. They concluded that the trivariate distribution has been established as a reliable tool for flood frequency analysis. Rahman
et al. (2013) used the site-specific flood frequency, which is based on the probability
distribution of long period stream flow data. In this research, the choice of appropriate probability distribution in terms of the nature of the data sets is more vital for
finding out the actual scenario. Apart from this, region-specific difference has been
found in terms of the appropriate probability distribution. Bezak et al. (2014) used
annual maximum (AM) and peaks over threshold (POT) to compare better accuracy
for estimating the frequencies of the flood. They have identified that the POT method
is more realistic than the AM method. In this case, the Poisson distribution is more
realistic than the Binomial distribution, considering the annual threshold limit.
Arnaud et al. (2016) used the stochastic simulation which is based on hourly rainfall
S. C. Pal et al.