Scholarlink

<p>Heatwaves in India have intensified in recent decades, with previously less-affected regions such as Upper Assam becoming increasingly vulnerable. This study investigates the heatwave characteristics over Upper Assam during 1991–2024 and assesses how soil moisture (SM) and other key hydro-meteorological variables jointly modulate heatwaves. Daily maximum temperature (T<sub>max</sub>) at the 90<sup>th</sup> percentile threshold from ERA5-Land is used to derive long-term heatwave metrics, while SM, solar radiation (SR), relative humidity (RH), rainfall (RF), sensible heat flux (SHF), and land surface temperature (LST) are analysed from ERA5-Land, IMDAA, and SMAP. For the most intense events between 2015 and 2024, 16-day windows centring the longest heatwave duration each year are used to examine short-term land–atmosphere feedbacks via spatio-temporal correlations. Interpretable machine learning (ML) models (Random Forest, LightGBM, XGBoost) are then employed to rank the relative importance of these drivers for T<sub>max</sub> during heatwave periods. Results show a significant rise in both the highest and threshold heatwave temperatures, alongside an increase in total heatwave days over the last decade. During core heatwave days, SM commonly declines by up to about 30%, coinciding with enhanced SHF, suppressed RH, and elevated LST. Notably, ML models consistently ranked SM as the fourth most influential factor after RH, SR, and LST in determining T<sub>max</sub> during heatwaves, with XGBoost showing the highest predictive skill (R² = 0.821). These findings emphasise the crucial, though often underestimated, role of SM in heatwave dynamics and highlight the value of incorporating SM into predictive climate risk assessments for vulnerable 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

Understanding the interactions between soil moisture and other hydro-meteorological parameters during heatwaves over upper Assam in North-East India

  • 期刊论文

  • 发表日期：

    2026年7月24日

查看全文
  • Aniket Chakraborty,
  • Binita Pathak,
  • Partha Pratim Gogoi,
  • U. P. V. Sreelakshmi,
  • Partha Jyoti Sahu,
  • Pradip Kumar Bhuyan

摘要

Heatwaves in India have intensified in recent decades, with previously less-affected regions such as Upper Assam becoming increasingly vulnerable. This study investigates the heatwave characteristics over Upper Assam during 1991–2024 and assesses how soil moisture (SM) and other key hydro-meteorological variables jointly modulate heatwaves. Daily maximum temperature (Tmax) at the 90th percentile threshold from ERA5-Land is used to derive long-term heatwave metrics, while SM, solar radiation (SR), relative humidity (RH), rainfall (RF), sensible heat flux (SHF), and land surface temperature (LST) are analysed from ERA5-Land, IMDAA, and SMAP. For the most intense events between 2015 and 2024, 16-day windows centring the longest heatwave duration each year are used to examine short-term land–atmosphere feedbacks via spatio-temporal correlations. Interpretable machine learning (ML) models (Random Forest, LightGBM, XGBoost) are then employed to rank the relative importance of these drivers for Tmax during heatwave periods. Results show a significant rise in both the highest and threshold heatwave temperatures, alongside an increase in total heatwave days over the last decade. During core heatwave days, SM commonly declines by up to about 30%, coinciding with enhanced SHF, suppressed RH, and elevated LST. Notably, ML models consistently ranked SM as the fourth most influential factor after RH, SR, and LST in determining Tmax during heatwaves, with XGBoost showing the highest predictive skill (R² = 0.821). These findings emphasise the crucial, though often underestimated, role of SM in heatwave dynamics and highlight the value of incorporating SM into predictive climate risk assessments for vulnerable regions.