Scholarlink

<p>Kerala frequently experiences both floods and droughts due to highly variable monsoon rainfall, marked by alternating heavy spells and dry breaks. The timing, intensity, and spatial distribution of rainfall are vital to Kerala’s climate and economy. This study proposes a novel hybrid algorithm that integrates Differential Evolution (DE) with the Apriori association rule mining technique to detect spatial rainfall co-occurrence patterns. Conventional algorithms like Apriori and FP-Growth (Frequent Pattern Growth) face challenges in identifying quantitative associations among variables. Furthermore, statistical techniques like correlation methods and Analysis of Variance (ANOVA) fail to capture the nonlinear and evolutionary nature of rainfall dynamics. The proposed study analyzes Automatic Weather Station (AWS) rainfall data from June (2008–2020) to decode spatial variability and inter-site associations during the early monsoon season. Data from three key AWS sites i.e., Vengad Kannur (North), Okkal Ernakulam (Central), and Pullad Pathanamthitta (South) were examined and identifies a strong co-occurrence of heavy rainfall events (&gt; 100&#xa0;mm/day) between Vengad Kannur and Okkal Ernakulam, largely independent of rainfall at Pullad Pathanamthitta. Understanding of spatial rainfall variability helps in water resource management and monsoon risk mitigation in Kerala.</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

Pattern discovery in early monsoon rainfall across Kerala using data mining techniques

  • 期刊论文

  • 发表日期：

    2026年4月20日

查看全文
  • Anupam Priamvada,
  • Bipasha Paul Shukla

摘要

Kerala frequently experiences both floods and droughts due to highly variable monsoon rainfall, marked by alternating heavy spells and dry breaks. The timing, intensity, and spatial distribution of rainfall are vital to Kerala’s climate and economy. This study proposes a novel hybrid algorithm that integrates Differential Evolution (DE) with the Apriori association rule mining technique to detect spatial rainfall co-occurrence patterns. Conventional algorithms like Apriori and FP-Growth (Frequent Pattern Growth) face challenges in identifying quantitative associations among variables. Furthermore, statistical techniques like correlation methods and Analysis of Variance (ANOVA) fail to capture the nonlinear and evolutionary nature of rainfall dynamics. The proposed study analyzes Automatic Weather Station (AWS) rainfall data from June (2008–2020) to decode spatial variability and inter-site associations during the early monsoon season. Data from three key AWS sites i.e., Vengad Kannur (North), Okkal Ernakulam (Central), and Pullad Pathanamthitta (South) were examined and identifies a strong co-occurrence of heavy rainfall events (> 100 mm/day) between Vengad Kannur and Okkal Ernakulam, largely independent of rainfall at Pullad Pathanamthitta. Understanding of spatial rainfall variability helps in water resource management and monsoon risk mitigation in Kerala.