Nonlinear regression technique to assess the landslide susceptibility of the Kalapahar hill, Guwahati, Assam State (India)

In the past few decades, the increasing frequency of landslides has
become a concern for Guwahati city (capital of the State of Assam),
especially around the low lying hills. In particular, the Kalapahar hill is
one of the major landslide-prone areas of Guwahati, due to its peculiar
geomorphological aspects.. The area is characterized by steep slopes
underlying loose unconsolidated soil, which lead to frequent slope
failure, especially during the monsoon season. Moreover, the intense
urbanization of the hills has led to slope instability, making these areas
more vulnerable. This paper provides a preliminary study on landslide
susceptibility assessment, representing a first step towards landslide
risk reduction. In particular, a semi-parametric nonlinear regression
method, namely the GAM (Generalized Additive Model), was applied
for landslide susceptibility mapping.