Scholarlink

<p>The present study investigates the role of channel bottlenecking and anthropogenic interventions in shaping flood dynamics and influencing floodplain wetland sustainability in the moribund deltaic Ganges floodplain, India. The primary objective is to evaluate how channel bottlenecking, embankment construction, damming, source closure of avulsed channels, and urban-induced channel constriction affect flood magnitude, frequency, duration, and wetland transformation. To achieve this, extensive field-based measurements of channel morphology and flood characteristics were conducted, complemented by advanced machine learning (ML) techniques for flood susceptibility mapping. The sensitivity of spatial flood susceptibility to bottlenecking and embankment parameters was quantified, while the effects of damming were examined using downstream water level data. Results show that the convergence of multiple rivers (Mayurakshi, Kuya, Mor, Banki, and Dwarka) within a small low-lying area has created a highly flood-prone setting. Among the applied ML algorithms, the Random Forest (RF) model has demonstrated the highest predictive performance for flood susceptibility mapping. Channel bottlenecking in the confluence segment has substantially increased flood magnitude, expanding the area of extreme flood susceptibility from 7.22 to 84.35 km<sup>2</sup>. Conversely, embankment installation has enhanced the river’s carrying capacity from 53 to 88% reducing the extreme flood-susceptible area to 3.39 km<sup>2</sup>. Damming has decreased downstream water levels by 6–38%, while source closure of avulsed channel has reduced flood frequency, duration, and depth. In contrast, urbanization-induced channel constriction has exacerbated flood risk. However, certain flood mitigation measures have disrupted hydrological connectivity between rivers and floodplain wetlands, accelerating wetland degradation and transformation. By integrating field-driven channel morphological parameters with hydrological modifications such as bottlenecking, embanking, and damming into the ML-based modelling framework, this study provides a novel and comprehensive assessment of flood dynamics and wetland sustainability in a floodplain environment. The findings particularly, measured sensitivity of bottlenecking and embanking river offer valuable insights for revisiting current flood management policies.</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

Impact of channel bottlenecking and anthropogeomorphic interventions on flood and wetland conditions in the lower gangetic floodplain

  • 期刊论文

  • 发表日期：

    2025年10月17日

查看全文
  • Ripan Ghosh,
  • Sutapa Mukhopadhyay,
  • Sandipta Debanshi,
  • Surajit Let,
  • Manabendra Let,
  • Priyanka Das,
  • Swapan Talukdar,
  • Susanta Mahato,
  • Swades Pal

摘要

The present study investigates the role of channel bottlenecking and anthropogenic interventions in shaping flood dynamics and influencing floodplain wetland sustainability in the moribund deltaic Ganges floodplain, India. The primary objective is to evaluate how channel bottlenecking, embankment construction, damming, source closure of avulsed channels, and urban-induced channel constriction affect flood magnitude, frequency, duration, and wetland transformation. To achieve this, extensive field-based measurements of channel morphology and flood characteristics were conducted, complemented by advanced machine learning (ML) techniques for flood susceptibility mapping. The sensitivity of spatial flood susceptibility to bottlenecking and embankment parameters was quantified, while the effects of damming were examined using downstream water level data. Results show that the convergence of multiple rivers (Mayurakshi, Kuya, Mor, Banki, and Dwarka) within a small low-lying area has created a highly flood-prone setting. Among the applied ML algorithms, the Random Forest (RF) model has demonstrated the highest predictive performance for flood susceptibility mapping. Channel bottlenecking in the confluence segment has substantially increased flood magnitude, expanding the area of extreme flood susceptibility from 7.22 to 84.35 km2. Conversely, embankment installation has enhanced the river’s carrying capacity from 53 to 88% reducing the extreme flood-susceptible area to 3.39 km2. Damming has decreased downstream water levels by 6–38%, while source closure of avulsed channel has reduced flood frequency, duration, and depth. In contrast, urbanization-induced channel constriction has exacerbated flood risk. However, certain flood mitigation measures have disrupted hydrological connectivity between rivers and floodplain wetlands, accelerating wetland degradation and transformation. By integrating field-driven channel morphological parameters with hydrological modifications such as bottlenecking, embanking, and damming into the ML-based modelling framework, this study provides a novel and comprehensive assessment of flood dynamics and wetland sustainability in a floodplain environment. The findings particularly, measured sensitivity of bottlenecking and embanking river offer valuable insights for revisiting current flood management policies.