- English
Climate variability poses a serious challenge to the
sustainability of fragile mountain ecosystems where local livelihoods
are closely dependent on natural resources. However, limited
studies have examined long-term trends and local perceptions of
climate variability in the north-eastern region of Dimapur district of
Nagaland state in India. Thus, this study aims to analyze the
temporal trends and forecast changes in rainfall, temperature and
relative humidity in Dimapur district during 1998–2020. Mann–
Kendall test was utilized for examining trend in the meteorological
variables. Magnitude of these variables was assessed using Sen’s
slope estimator. Multilayer Perceptron (MLP) and random forest
(RF) machine learning algorithms were used for forecasting
meteorological variables in the study area. A household-level
survey was conducted to record perception on climate variability
and its impact using structured questionnaire. The results revealed
significant increasing trend in maximum temperature during winter
and pre-monsoon seasons whereas decreasing trend in minimum
temperature during pre-monsoon and monsoon seasons. Though
variation in rainfall pattern was noticed but no significant trend was
observed. The MLP model was found effective than RF for forecasting
of meteorological variables based on the performance
assessors. Forecasting of variables has also shown decreasing trend
in rainfall and increasing trend in temperature and relative
humidity. These findings highlight considerable climate variability
in the district which may have serious implications for the fragile
and sensitive mountain ecosystem and the livelihood of local
communities. The integration of climate variability analysis with
community perceptions may help in devising suitable adaptation
strategies in other mountainous regions.
- English
