Scholarlink

<p>Hydrological model selection in data-scarce tropical basins often prioritizes software accessibility or aggregate statistical performance over empirical suitability, leading to misapplication and suboptimal water governance. This study introduces a paradigm-shifting, transferable framework arguing that alignment with specific water management objectives, not universal metrics, should dictate hydrological model choice in data-scarce tropical basins. We develop an integrated methodology combining Markov Chain-Cellular Automata LULC projection with Monte Carlo uncertainty analysis, statistical residual-based model comparison, and a three-step decision guide. Applied to Nigeria’s Chanchaga River Basin (171.6% built-up expansion, 2000-2020), urbanization drove a 25.1% increase in peak discharge and 19.5% reduction in baseflow). Validation of the LULC projection yielded a Kappa index of 0.84 and an overall accuracy of 89.2%. HEC-HMS demonstrated superior peak-flow capture (high-flow NSE = 0.78), whereas SWAT better represented baseflow (low-flow NSE = 0.61). A paired t-test confirmed HEC-HMS’s statistically significant superiority in reducing simulation error (t = 3.42, <i>p</i> &lt; 0.001; Cohen’s d = 0.68). A cross-calibration experiment confirmed that performance differences are structural rather than calibration-driven. Projections indicate a 37.1% total increase in peak discharge by 2027 under a business-as-usual scenario, and 18.5% (low-growth) and 9.2% (green infrastructure) under alternative scenarios. The resultant objective-driven decision framework recommends HEC-HMS for flood mitigation and SWAT for water-resource planning, providing a transferable methodology with explicit application guidelines.</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

An objective-driven framework for hydrological model selection in data-scarce tropical regions: integrating LULC projection and uncertainty analysis

  • 期刊论文

  • 发表日期：

    2026年7月15日

查看全文
  • Ekundayo A. Adesina,
  • Hassan M. Suraj

摘要

Hydrological model selection in data-scarce tropical basins often prioritizes software accessibility or aggregate statistical performance over empirical suitability, leading to misapplication and suboptimal water governance. This study introduces a paradigm-shifting, transferable framework arguing that alignment with specific water management objectives, not universal metrics, should dictate hydrological model choice in data-scarce tropical basins. We develop an integrated methodology combining Markov Chain-Cellular Automata LULC projection with Monte Carlo uncertainty analysis, statistical residual-based model comparison, and a three-step decision guide. Applied to Nigeria’s Chanchaga River Basin (171.6% built-up expansion, 2000-2020), urbanization drove a 25.1% increase in peak discharge and 19.5% reduction in baseflow). Validation of the LULC projection yielded a Kappa index of 0.84 and an overall accuracy of 89.2%. HEC-HMS demonstrated superior peak-flow capture (high-flow NSE = 0.78), whereas SWAT better represented baseflow (low-flow NSE = 0.61). A paired t-test confirmed HEC-HMS’s statistically significant superiority in reducing simulation error (t = 3.42, p < 0.001; Cohen’s d = 0.68). A cross-calibration experiment confirmed that performance differences are structural rather than calibration-driven. Projections indicate a 37.1% total increase in peak discharge by 2027 under a business-as-usual scenario, and 18.5% (low-growth) and 9.2% (green infrastructure) under alternative scenarios. The resultant objective-driven decision framework recommends HEC-HMS for flood mitigation and SWAT for water-resource planning, providing a transferable methodology with explicit application guidelines.