Scholarlink

<p>Precipitation is a key variable for hydrological research but observed data collection faces challenges due to sparse coverage and inaccessibility, particularly in remote regions. Gridded precipitation datasets, such as APHRODITE (Asian Precipitation Highly Resolved Observational Data Integration Towards Evaluation of Water Resources, hereafter AP), provide an effective alternative due to their high spatial and temporal resolution, wide coverage, and reliable performance. However, AP often underrepresents critical rainfall features, especially extreme precipitation events, which are vital for accurate hydrological modeling. This study evaluates three bias correction techniques applied to the AP precipitation dataset across Sri Lanka and examines their impact on simulating streamflow in the Malwathu Oya River Basin. The methods include linear scaling (APLS), quantile-based cumulative distribution function matching (APQM), and an extreme precipitation adjustment (APEX) designed to preserve mean, variance, and extremes. Results indicate that APEX outperforms other methods in representing precipitation accurately. Root mean square error (RMSE) values for APEX ranged from 2.09 to 3.06&#xa0;mm/day across Wet, Intermediate, and Dry zones, compared with 4.2–6.8&#xa0;mm/day for AP, 4.2–8.3&#xa0;mm/day for APLS, and 4.03–5.65&#xa0;mm/day for APQM. The correlation of extreme precipitation indices was strongest for APEX (0.55–0.98), while AP, APLS, and APQM ranged from 0.26 to 0.94, 0.31–0.95, and 0.39–0.93, respectively. Streamflow simulations using APEX-corrected precipitation also demonstrated improved performance, closely reflecting observed flows. These findings highlight that APEX-corrected AP precipitation is superior in preserving key statistical properties, particularly extremes, and offers a reliable alternative precipitation source for river basins in Sri Lanka with limited rain gauge networks, enhancing the accuracy of hydrological modeling and water resource management.</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

Evaluating extreme precipitation and streamflow characteristics with novel bias adjusted APHRODITE precipitation gridded datasets: A case in Malwathu Oya river basin in Sri Lanka

  • 期刊论文

  • 发表日期：

    2026年1月10日

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

摘要

Precipitation is a key variable for hydrological research but observed data collection faces challenges due to sparse coverage and inaccessibility, particularly in remote regions. Gridded precipitation datasets, such as APHRODITE (Asian Precipitation Highly Resolved Observational Data Integration Towards Evaluation of Water Resources, hereafter AP), provide an effective alternative due to their high spatial and temporal resolution, wide coverage, and reliable performance. However, AP often underrepresents critical rainfall features, especially extreme precipitation events, which are vital for accurate hydrological modeling. This study evaluates three bias correction techniques applied to the AP precipitation dataset across Sri Lanka and examines their impact on simulating streamflow in the Malwathu Oya River Basin. The methods include linear scaling (APLS), quantile-based cumulative distribution function matching (APQM), and an extreme precipitation adjustment (APEX) designed to preserve mean, variance, and extremes. Results indicate that APEX outperforms other methods in representing precipitation accurately. Root mean square error (RMSE) values for APEX ranged from 2.09 to 3.06 mm/day across Wet, Intermediate, and Dry zones, compared with 4.2–6.8 mm/day for AP, 4.2–8.3 mm/day for APLS, and 4.03–5.65 mm/day for APQM. The correlation of extreme precipitation indices was strongest for APEX (0.55–0.98), while AP, APLS, and APQM ranged from 0.26 to 0.94, 0.31–0.95, and 0.39–0.93, respectively. Streamflow simulations using APEX-corrected precipitation also demonstrated improved performance, closely reflecting observed flows. These findings highlight that APEX-corrected AP precipitation is superior in preserving key statistical properties, particularly extremes, and offers a reliable alternative precipitation source for river basins in Sri Lanka with limited rain gauge networks, enhancing the accuracy of hydrological modeling and water resource management.