Presenter Name: K. Kumar Mohanta and J. Mohanta
Submissionid: 172
Abstract
Linear regression analysis is one of the most common statistical methods used to
investigate trends in rainfall over time. However, conventional regression models
usually require datasets that have been completely and accurately determined for
trend analysis with rainfall, the datasets often include monthly or annual averages
which can result in uncertainty in trend analysis. In order to incorporate this
uncertainty, fuzzy set theory represents the ambiguity from the analyzed datasets. The
fuzzy regression models consider both input and output variables to be Triangular
fuzzy numbers. This study proposes a fuzzy regression approach that describes the
relationships between precipitation and time in a context of excessive variability in
rainfall which creates challenges for the sustainable management of water resources.
Kalahandi district in Odisha has been acknowledged as one of the impoverished
districts in the KBK Region. The utilization of the fuzzy regression model in
addressing trend variability or trends in rainfall in Kalahandi, Odisha is a
comprehensive investigation of historical rainfall records related to trends and
variability of rainfall in Kalahandi, especially the region’s severe effects associated
with climate change; it is of utmost importance to understand how rainfall variability
trends have changed over time for local agrarians or agriculture workers and policymakers to recognize changes in rainfall patterns and develop adapted agriculture
plans and manage water resources in a variable climate. The data collection process
included 100 years of fuzzy rainfall data (1923 – 2023); thus, including the last
century helps frame current rainfall variability trends in Kalahandi. From the
research, it has been shown that Kalahandi is experiencing climate change or other
environmental factors that are leading to a substantial increase in rainfall variability.
The study methodologically establishes fuzzy regression as an alternate approach to
provide a more accurate predictive model for rainfall.
wacth a full video: