Scholarlink

<p>Accurate climate forecasts are critical for managing water resources in Sri Lanka, where climate change and extreme weather impact socioeconomic sectors. This study evaluates three bias-correction (BC) methods – Linear Scaling (LS), Quantile Mapping (QM), and a novel Extremes-Adjusted Quantile Mapping (QMEX) – for CMIP6 GCM precipitation and temperature data, assessing their effects on streamflow simulations in the Malwathu Oya River Basin (MORB). Eight GCMs were ranked using a multi-metric evaluation framework based on compromise programming and Lp-norm aggregation using APHRODITE precipitation and Princeton University temperature datasets. LS was applied to temperature, while precipitation was corrected with LS, QM, and QMEX across three climatic zones. EC-Earth3, the top-ranked GCM, was bias-corrected and projected until 2050 under SSP2-4.5 and SSP5-8.5 scenarios. QMEX outperformed other methods (KGE &gt; 0.7 vs. KGE &lt; 0.5), effectively preserving extreme precipitation events, while LS substantially improved minimum and maximum temperature biases. Projections indicate increased total precipitation, heightened flood risks, and more intense wet seasons across all zones. Temperatures are expected to rise, particularly during the dry season, with smaller changes in the rainy season. The bias-corrected climate projections were then used to assess streamflow responses in the MORB using the HEC-HMS hydrological model. Responses to climate change are strongly influenced by extreme events, with some years under SSP5-8.5 showing higher streamflow than baseline. These results demonstrate that effective bias correction is essential for improving climate-driven hydrological modeling, enhancing the reliability of future projections, and supporting adaptive water resource planning in vulnerable river basins.</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

Novel bias correction framework for CMIP6 climate projections over Sri Lanka and streamflow impacts in the Malwathu River Basin

  • 期刊论文

  • 发表日期：

    2026年7月28日

查看全文
  • Sareeha Vasanthakumar,
  • Mohanasundaram Shanmugam,
  • Sangam Shrestha,
  • Mukand S Babel,
  • Ho Huu Loc,
  • Sushil Kumar Himanshu

摘要

Accurate climate forecasts are critical for managing water resources in Sri Lanka, where climate change and extreme weather impact socioeconomic sectors. This study evaluates three bias-correction (BC) methods – Linear Scaling (LS), Quantile Mapping (QM), and a novel Extremes-Adjusted Quantile Mapping (QMEX) – for CMIP6 GCM precipitation and temperature data, assessing their effects on streamflow simulations in the Malwathu Oya River Basin (MORB). Eight GCMs were ranked using a multi-metric evaluation framework based on compromise programming and Lp-norm aggregation using APHRODITE precipitation and Princeton University temperature datasets. LS was applied to temperature, while precipitation was corrected with LS, QM, and QMEX across three climatic zones. EC-Earth3, the top-ranked GCM, was bias-corrected and projected until 2050 under SSP2-4.5 and SSP5-8.5 scenarios. QMEX outperformed other methods (KGE > 0.7 vs. KGE < 0.5), effectively preserving extreme precipitation events, while LS substantially improved minimum and maximum temperature biases. Projections indicate increased total precipitation, heightened flood risks, and more intense wet seasons across all zones. Temperatures are expected to rise, particularly during the dry season, with smaller changes in the rainy season. The bias-corrected climate projections were then used to assess streamflow responses in the MORB using the HEC-HMS hydrological model. Responses to climate change are strongly influenced by extreme events, with some years under SSP5-8.5 showing higher streamflow than baseline. These results demonstrate that effective bias correction is essential for improving climate-driven hydrological modeling, enhancing the reliability of future projections, and supporting adaptive water resource planning in vulnerable river basins.