Scholarlink

<p>This study presents a comprehensive analysis of the spatiotemporal variability in rainfall over Kerala, India, for the period 1951–2023. The structure, variability and drivers of spatial patterns in rainfall over the state were analysed using multi-stage assessments combining unsupervised clustering, change-point detection, wavelet analysis, monsoon onset and teleconnection assessment. Pettitt’s test of spatially averaged annual rainfall detected a statistically significant change point around the year 1982. The total time period was split around this change point to understand the pre- and post-shift characteristics. Spatial clustering techniques were employed to identify spatial homogeneity in the monthly rainfall over the state, and six distinct spatial clusters were identified. Discrete Wavelet Transform (DWT) revealed distinct spatial heterogeneity in rainfall, with increasing rainfall along the southern coast and Western Ghats and declining trends in central and northern areas. Extreme rainfall frequency and intensity have risen in comparatively drier zones, indicating growing climate vulnerability. Wavelet coherence analysis showed that El Niño-Southern Oscillation (ENSO) has a limited influence on monthly rainfall, while the Indian Ocean Dipole (IOD) and North Atlantic Oscillation (NAO) exhibit stronger, more consistent impacts, with contrasting effects during the southwest and northeast monsoon seasons. The Indian Summer Monsoon onset analysis showed that onset timing has only a weak effect on seasonal rainfall totals and extremes across most regions except the higher rainfall clusters, with responses strongly mediated by the underlying spatial rainfall climatology. Overall, the findings deepen understanding of Kerala’s rainfall dynamics and support improved climate adaptation and resource management in rainfall-sensitive sectors.</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

Spatiotemporal characterisation of rainfall variability in Kerala, India, using clustering and wavelet-based techniques

  • 期刊论文

  • 发表日期：

    2025年12月29日

查看全文
  • T. Anand,
  • Jose George,
  • R. Gowri,
  • K. P. Indulekha

摘要

This study presents a comprehensive analysis of the spatiotemporal variability in rainfall over Kerala, India, for the period 1951–2023. The structure, variability and drivers of spatial patterns in rainfall over the state were analysed using multi-stage assessments combining unsupervised clustering, change-point detection, wavelet analysis, monsoon onset and teleconnection assessment. Pettitt’s test of spatially averaged annual rainfall detected a statistically significant change point around the year 1982. The total time period was split around this change point to understand the pre- and post-shift characteristics. Spatial clustering techniques were employed to identify spatial homogeneity in the monthly rainfall over the state, and six distinct spatial clusters were identified. Discrete Wavelet Transform (DWT) revealed distinct spatial heterogeneity in rainfall, with increasing rainfall along the southern coast and Western Ghats and declining trends in central and northern areas. Extreme rainfall frequency and intensity have risen in comparatively drier zones, indicating growing climate vulnerability. Wavelet coherence analysis showed that El Niño-Southern Oscillation (ENSO) has a limited influence on monthly rainfall, while the Indian Ocean Dipole (IOD) and North Atlantic Oscillation (NAO) exhibit stronger, more consistent impacts, with contrasting effects during the southwest and northeast monsoon seasons. The Indian Summer Monsoon onset analysis showed that onset timing has only a weak effect on seasonal rainfall totals and extremes across most regions except the higher rainfall clusters, with responses strongly mediated by the underlying spatial rainfall climatology. Overall, the findings deepen understanding of Kerala’s rainfall dynamics and support improved climate adaptation and resource management in rainfall-sensitive sectors.