Scholarlink

<p>Extreme precipitation is a major driver of hydrological risk in southern Brazil, yet its reliable characterization is constrained by strong spatial heterogeneity and coastal–orographic controls. This study addresses this limitation by evaluating the ability of satellite-derived and reanalysis precipitation datasets to represent extreme rainfall over Paraná State, Brazil, during 1983–2024. Daily precipitation from the Brazilian Daily Weather Gridded Data (BR-DWGD) is adopted as the observational reference, and five gridded products are assessed, including CHIRPS and PERSIANN-CDR, as well as ERA5, ERA5-Land, and AgERA5. Extreme precipitation is quantified using Expert Team on Climate Change Detection and Indices (ETCCDI) metrics that capture complementary dimensions of the rainfall regime, including accumulated totals (PRCPTOT), short-duration maxima (RX1Day and RX5Day), heavy-precipitation totals (R95p), dry-spell persistence (CDD), and wet-day intensity (SDII). Indices are computed at annual and seasonal scales (DJF, MAM, JJA, and SON) and evaluated using Willmott’s index of agreement and percent bias, with diagnostics summarized at both state and river-basin scales to assess physiographic modulation of skill. Results indicate that dataset performance is conditioned by index type, temporal aggregation, and seasonal forcing, yielding a marked coastal to interior gradient. CHIRPS exhibits the most consistent performance for accumulated indicators, particularly PRCPTOT and RX5Day, with high agreement and limited bias across interior basins. In contrast, daily maxima and percentile-based extremes show reduced skill across all products, with the strongest degradation in Serra do Mar coastal and orographic sectors, where spatial displacement and attenuation of peak intensities persist. Reanalysis products perform better under winter synoptic forcing, but they retain underestimation of short-duration extremes, whereas summer convective conditions amplify uncertainties across datasets. Overall, no single dataset performs consistently across indices, seasons, and physiographic domains. CHIRPS is most suitable for monitoring precipitation totals and multi-day extremes, whereas daily and percentile-based extremes require cautious, basin-specific interpretation to support hydroclimatic assessment and climate-risk applications in Paraná State.</p> Graphical Abstract <p></p> <p>This graphical abstract provides a concise and visually organized summary of the study “Performance of Satellite and Reanalysis Products for ETCCDI Rainfall Extremes in Paraná State, Brazil”. The figure synthesizes the main methodological steps and key findings, visually linking gridded precipitation datasets, ETCCDI indices, and spatial contrasts in product skill across a climatically and topographically complex study region. The objective, placed at the top, highlights the evaluation of satellite-derived and reanalysis precipitation products in representing ETCCDI rainfall extremes over Paraná State, Brazil, using BR-DWGD daily precipitation as the observational reference for 1983–2024. In the Methodology section, data sources are represented by an observational dataset icon for BR-DWGD and satellite/reanalysis icons for the gridded products (CHIRPS, PERSIANN-CDR, ERA5, ERA5-Land, and AgERA5), supported by performance metrics (Willmott’s d and PBIAS) applied at annual and seasonal scales with basin-level spatial assessment. Analytical methods are summarized through the ETCCDI precipitation indices (PRCPTOT, RX1Day, RX5Day, R95p, CDD, SDII), emphasizing the contrast between accumulated and short-duration extremes. The Key Results section illustrates that CHIRPS provides the highest and most consistent skill for accumulated rainfall (PRCPTOT) and multi-day extremes, particularly over interior basins, whereas short-duration and percentile-based extremes (RX1Day and R95p) exhibit reduced skill across all products, with strongest degradation in coastal–orographic sectors influenced by the Serra do Mar. Reanalysis datasets show comparatively improved performance under winter synoptic forcing but systematically underestimate short-duration extremes, reinforcing scale-dependent limitations. The Conclusion conveys that no single dataset performs uniformly across indices and seasons; CHIRPS is most suitable for magnitude-focused applications involving totals and multi-day extremes, while daily extremes require basin-specific interpretation and cautious product selection in topographically complex regions, supporting robust hydroclimatic assessments and climate-risk applications across Paraná State.</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

Performance of Satellite and Reanalysis Products for ETCCDI Rainfall Extremes in Paraná State, Brazil

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月21日

查看全文
  • Paulo Miguel de Bodas Terassi,
  • Jakeline Baratto

摘要

Extreme precipitation is a major driver of hydrological risk in southern Brazil, yet its reliable characterization is constrained by strong spatial heterogeneity and coastal–orographic controls. This study addresses this limitation by evaluating the ability of satellite-derived and reanalysis precipitation datasets to represent extreme rainfall over Paraná State, Brazil, during 1983–2024. Daily precipitation from the Brazilian Daily Weather Gridded Data (BR-DWGD) is adopted as the observational reference, and five gridded products are assessed, including CHIRPS and PERSIANN-CDR, as well as ERA5, ERA5-Land, and AgERA5. Extreme precipitation is quantified using Expert Team on Climate Change Detection and Indices (ETCCDI) metrics that capture complementary dimensions of the rainfall regime, including accumulated totals (PRCPTOT), short-duration maxima (RX1Day and RX5Day), heavy-precipitation totals (R95p), dry-spell persistence (CDD), and wet-day intensity (SDII). Indices are computed at annual and seasonal scales (DJF, MAM, JJA, and SON) and evaluated using Willmott’s index of agreement and percent bias, with diagnostics summarized at both state and river-basin scales to assess physiographic modulation of skill. Results indicate that dataset performance is conditioned by index type, temporal aggregation, and seasonal forcing, yielding a marked coastal to interior gradient. CHIRPS exhibits the most consistent performance for accumulated indicators, particularly PRCPTOT and RX5Day, with high agreement and limited bias across interior basins. In contrast, daily maxima and percentile-based extremes show reduced skill across all products, with the strongest degradation in Serra do Mar coastal and orographic sectors, where spatial displacement and attenuation of peak intensities persist. Reanalysis products perform better under winter synoptic forcing, but they retain underestimation of short-duration extremes, whereas summer convective conditions amplify uncertainties across datasets. Overall, no single dataset performs consistently across indices, seasons, and physiographic domains. CHIRPS is most suitable for monitoring precipitation totals and multi-day extremes, whereas daily and percentile-based extremes require cautious, basin-specific interpretation to support hydroclimatic assessment and climate-risk applications in Paraná State.

Graphical Abstract

This graphical abstract provides a concise and visually organized summary of the study “Performance of Satellite and Reanalysis Products for ETCCDI Rainfall Extremes in Paraná State, Brazil”. The figure synthesizes the main methodological steps and key findings, visually linking gridded precipitation datasets, ETCCDI indices, and spatial contrasts in product skill across a climatically and topographically complex study region. The objective, placed at the top, highlights the evaluation of satellite-derived and reanalysis precipitation products in representing ETCCDI rainfall extremes over Paraná State, Brazil, using BR-DWGD daily precipitation as the observational reference for 1983–2024. In the Methodology section, data sources are represented by an observational dataset icon for BR-DWGD and satellite/reanalysis icons for the gridded products (CHIRPS, PERSIANN-CDR, ERA5, ERA5-Land, and AgERA5), supported by performance metrics (Willmott’s d and PBIAS) applied at annual and seasonal scales with basin-level spatial assessment. Analytical methods are summarized through the ETCCDI precipitation indices (PRCPTOT, RX1Day, RX5Day, R95p, CDD, SDII), emphasizing the contrast between accumulated and short-duration extremes. The Key Results section illustrates that CHIRPS provides the highest and most consistent skill for accumulated rainfall (PRCPTOT) and multi-day extremes, particularly over interior basins, whereas short-duration and percentile-based extremes (RX1Day and R95p) exhibit reduced skill across all products, with strongest degradation in coastal–orographic sectors influenced by the Serra do Mar. Reanalysis datasets show comparatively improved performance under winter synoptic forcing but systematically underestimate short-duration extremes, reinforcing scale-dependent limitations. The Conclusion conveys that no single dataset performs uniformly across indices and seasons; CHIRPS is most suitable for magnitude-focused applications involving totals and multi-day extremes, while daily extremes require basin-specific interpretation and cautious product selection in topographically complex regions, supporting robust hydroclimatic assessments and climate-risk applications across Paraná State.