Scholarlink

<p>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.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Analyzing Climate Variability Using Non-Parametric and Machine Learning Approaches: Empirical Evidence from Dimapur District of Nagaland, India

  • 期刊论文

  • 发表日期：

    2026年2月20日

查看全文
  • Geeta Kumari,
  • Haroon Sajjad,
  • Tamal Kanti Saha,
  • Aastha Sharma,
  • Rayees Ali,
  • Md Masroor,
  • Pankaj Kumar,
  • Ankita Gupta,
  • Tarun Yadav,
  • Ram Avtar

摘要

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.