Scholarlink

<p>Climate change is intensifying regional hydroclimatic extremes, yet systematic understanding of their multi-scale impacts on agricultural water disasters remains limited, particularly regarding critical crop growth stages in key production zones. This study employed Modified Mann–Kendall (MMK) trend analysis, correlation analysis, and random forest (RF) models to investigate the climate extremes in Hunan Province of China over 1961–2020 at various time scales (seasonal, hydrological, and crop-growing seasonal scales) and their contributions to agricultural water disasters. Results indicate that precipitation decreased in spring (March–May) and autumn seasons (September–November) (significant in 3.7% stations) but increased at 96.6% of stations in summer (June–August) and winter (December-February). Temperature extremes shifted toward warm events, with Warm days (TX90p)/ summer days (SU) increasing and cold nights (TN10p)/ frost days (FD) decreasing at more than 92.6% stations. Hydrologically, heavy rainfall days (R50) surged in 77.8% stations across flood-seasons (April-September) but decreased in 92.6% stations across non-flood (October–March) seasons. The temperature indicators (especially extreme heat events) generally increased. Crop-growing-seasonal analyses showed R50 increased during rice (April-August) and rapeseed (September-May) seasons (77.8% and 88.0%, respectively). The lake districts experienced greater trends of heavy rainfall and warming than non-lake districts. The correlations between climate indicators and the agricultural water disaster-affected areas were most significant during the rice-growing season, especially for precipitation (r = − 0.58 for drought and r = 0.62 for flood), R50 (− 0.55 drought and 0.62 flood) and TX90p (0.39 drought and − 0.37 flood). RF modeling indicated that T<sub>m</sub> (mean temperature) and CDD (consecutive dry days) were dominant cross-temporal climatic factors, and precipitation best explained the variations in autumn/rapeseed-season flood. Mechanistically, droughts were temperature-dependent (T<sub>m</sub>, maximum temperature, minimum temperature) while floods were correlated with precipitation and extreme temperature events (TX90p, TN10p). These results can provide valuable guidance for climatic disaster reduction in critical zones.</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

Characterizing extreme climate events at different time scales and their contributions to agricultural drought and flooding areas

  • 期刊论文

  • 发表日期：

    2026年2月25日

查看全文
  • Rong Tang,
  • Ying Xia,
  • Junda Du,
  • Long Qian,
  • Hui Wang,
  • Yangzan Ou

摘要

Climate change is intensifying regional hydroclimatic extremes, yet systematic understanding of their multi-scale impacts on agricultural water disasters remains limited, particularly regarding critical crop growth stages in key production zones. This study employed Modified Mann–Kendall (MMK) trend analysis, correlation analysis, and random forest (RF) models to investigate the climate extremes in Hunan Province of China over 1961–2020 at various time scales (seasonal, hydrological, and crop-growing seasonal scales) and their contributions to agricultural water disasters. Results indicate that precipitation decreased in spring (March–May) and autumn seasons (September–November) (significant in 3.7% stations) but increased at 96.6% of stations in summer (June–August) and winter (December-February). Temperature extremes shifted toward warm events, with Warm days (TX90p)/ summer days (SU) increasing and cold nights (TN10p)/ frost days (FD) decreasing at more than 92.6% stations. Hydrologically, heavy rainfall days (R50) surged in 77.8% stations across flood-seasons (April-September) but decreased in 92.6% stations across non-flood (October–March) seasons. The temperature indicators (especially extreme heat events) generally increased. Crop-growing-seasonal analyses showed R50 increased during rice (April-August) and rapeseed (September-May) seasons (77.8% and 88.0%, respectively). The lake districts experienced greater trends of heavy rainfall and warming than non-lake districts. The correlations between climate indicators and the agricultural water disaster-affected areas were most significant during the rice-growing season, especially for precipitation (r = − 0.58 for drought and r = 0.62 for flood), R50 (− 0.55 drought and 0.62 flood) and TX90p (0.39 drought and − 0.37 flood). RF modeling indicated that Tm (mean temperature) and CDD (consecutive dry days) were dominant cross-temporal climatic factors, and precipitation best explained the variations in autumn/rapeseed-season flood. Mechanistically, droughts were temperature-dependent (Tm, maximum temperature, minimum temperature) while floods were correlated with precipitation and extreme temperature events (TX90p, TN10p). These results can provide valuable guidance for climatic disaster reduction in critical zones.