CoChat

The DOI matches its Crossref record; 91 of the 91 references it cites were found in a registry or an index; none of the 53 sources its claims cite is checked for support yet.

Construction of a daily streamflow dataset for Peru using a similarity-based regionalization approach and a hybrid hydrological modeling framework

Open access
Llauca, Harold, Leon, Karen, Lavado-Casimiro, Waldo · 2023 · Journal of Hydrology: Regional Studies

Abstract

A total of 11,913 sub-catchments in Peru and transboundary catchments with neighboring countries in South America. This paper aims to develop a national hydrological model using physiographic and climatic characteristics to identify donor and receptor sub-catchments (sub-zones). Therefore, we use the hydrometeorological PISCO dataset (0.1º x 0.1º) to drive a sub-catchment conceptual rainfall-runoff (ARNO/VIC) model and a river-routing (RAPID) model in thousands of river reaches. We identify 43 hydrological zones (with 122 sub-zones) to run the hybrid hydrological modeling framework (ARNO/VIC+RAPID) with previously calibrated and validated parameters with 43 fluviometric stations for 1981–2020. Simulated flow series show a higher performance at daily scale (KGE ≥ 0.75, NSEsqrt ≥ 0.65, MARE ≤ 1, and −25% ≤ PBIAS ≤ 25%) for catchments located at the Pacific coast and the Andes-Amazon transition, and good representation (R≥0.75) of seasonal and interannual variability. Increasing hydrological hazards such as floods highlight the importance of a systematic hydrological analysis and modeling at national level in gauged and ungauged catchments in Peru. This study introduces a new hydrological dataset of simulated daily flow series. The results are helpful for short-term flood risk scenario simulations and long-term water resource planning as the outcomes can provide valuable information for hydrologists and water resource managers in Peruvian regions with limited or no access to in-situ networks.

Overview

CoChat Summary

TL;DR

The paper develops a national daily streamflow dataset for Peru by combining similarity-based regionalization with an ARNO/VIC rainfall-runoff model and RAPID river routing, forced by the PISCO dataset. It simulates 1981–2020 daily flows across 11,913 river reaches and reports good performance in most gauged catchments, especially the Pacific coast and Andes-Amazon transition, with weaker performance in the Amazon lowlands.

Key findings

  • The study identifies 43 hydrological zones and 122 sub-zones across 11,913 sub-catchments (mean size 120 km²); 73% of sub-catchments are donors/gauged and 27% are receptors/ungauged, with the Pacific slope having the largest receptor share (38%), followed by Titicaca (29%) and Atlantic (16%).
  • For calibration and validation, the model performs well (KGE ≥ 0.75, NSEsqrt ≥ 0.65, MARE ≤ 1, and −25% ≤ PBIAS ≤ 25%) for stations in the Pacific, Titicaca, and Andes-Amazon transition, while Amazon lowland stations show low performance (≤ 0.5 in KGE and NSEsqrt).
  • For the total period, high KGE and NSEsqrt values occur at 71% and 68% of all stations, respectively; 95% of stations with lower relative error values also confirm good low-flow representation, and many gauges show R ≥ 0.75 at daily, monthly, and annual scales, though RMSE is higher in the Amazon lowlands (e.g., TAM = 14710.2 mm/year, BEL = 974.91 mm/year, TAB = 737.95 mm/year).
  • Leave-one-out cross-validation in 15 Pacific catchments shows good regionalization skill in central and south Pacific gauges, but northern catchments remain challenging for high-middle flows; most catchments with negative calibration bias have positive bias after cross-validation.

Methods

  • Study domain and data: 11,913 sub-catchments in Peru and transboundary catchments, using the GEOGloWS ECMWF river network; 43 fluviometric stations with observed daily flows for 1981–2020, split into 60% calibration and 40% validation.
  • Regionalization: seven climatic and physiographic predictors (P, TA, PET, AI, SLO, CLAY, fTC); hierarchical clustering with a maximum of three clusters per gauged catchment; dissimilarity computed with the Beck et al. equation normalized by IQR; donor selection based on lowest dissimilarity within 300 km.
  • Hydrological modeling: PISCO P and PET drive the ARNO/VIC model in FUSE; RAPID Muskingum routing generates daily streamflow; SCE-UA optimizes parameters using KGE as the objective function; performance is assessed with KGE, NSEsqrt, MARE, and PBIAS, plus nine hydrological signatures and LOOCV for 15 Pacific catchments.

Limitations

  • The paper states that only the single best donor within a 300 km radius was selected due to computational limitations and the small ungauged area (27%), which may increase regionalization uncertainties; future work should increase the number of donor candidates.
  • The paper states that unsatisfactory results in northeastern Peru/Amazon lowlands may reflect PISCO P uncertainties, sparse rainfall observations and lack of rainfall estimates in equatorial regions, daily PET climatologies increasing forcing uncertainties, limited fluviometric stations in the Amazon slope, and channel geometry uncertainties affecting Atlantic routing; it also notes assumptions about naturalized flows despite water regulation, mono-objective calibration emphasizing high-middle flows, and shorter records influencing calibration.
  • Evident from the text: LOOCV was applied only to 15 catchments in the Pacific drainage, not to all drainages; snow cover fraction and drainage area were not used as regionalization predictors; and PET was used only as daily climatology rather than a time-varying series.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1016/j.ejrh.2023.101381

Cites

91 of 91 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Image 1
CC-BY · From the PDF
Fig. 1 (a) Study domain boundaries that consider all catchments in the Pacific, Atlantic, and Titicaca slopes in Peru, and transboundary catchments with neighbors ' countries. Fluviometric stations are shown on the map as black dots. Zoom-in panels show (b) detail of sub-catchments employed for rainfall-runoff modeling, and (c) the river network used for river routing modeling.
Fig. 1 (a) Study domain boundaries that consider all catchments in the Pacific, Atlantic, and Titicaca slopes in Peru, and transboundary catchments with neighbors ' countries. Fluviometric stations are shown on the map as black dots. Zoom-in panels show (b) detail of sub-catchments employed for rainfall-runoff modeling, and (c) the river network used for river routing modeling.CC-BY · From the PDF
Fig. 2 Framework for the sub-catchment national simulation of daily flow series applying a similarity-based regionalization to identify hydro­ logical zones and sub-zones, and a hybrid modeling framework for parameter calibration and streamflow simulation across the whole Peruvian river network. P: precipitation; PET: potential evapotranspiration; TA: air temperature; AI: aridity index; SLO: surface slope; CLAY: soil clay content; fTC: fraction of tree cover; RU: runoff; Qsim: simulated streamflow; Qobs: observed streamflow; θ : model parameters.
Fig. 2 Framework for the sub-catchment national simulation of daily flow series applying a similarity-based regionalization to identify hydro­ logical zones and sub-zones, and a hybrid modeling framework for parameter calibration and streamflow simulation across the whole Peruvian river network. P: precipitation; PET: potential evapotranspiration; TA: air temperature; AI: aridity index; SLO: surface slope; CLAY: soil clay content; fTC: fraction of tree cover; RU: runoff; Qsim: simulated streamflow; Qobs: observed streamflow; θ : model parameters.CC-BY · From the PDF
Image 4
CC-BY · From the PDF
Fig. 4 On the left side: climatic (P: precipitation; TA: air temperature; PET: potential evapotranspiration; AI: aridity index) and physiographic (SLO: surface slope; CLAY: soil clay content; fTC: fraction of tree cover) predictor maps. On the right side: colored areas represent gauged catchments and sub-catchments (donors of model parameters), and white areas correspond to ungauged sub-catchments (receptors of model parameters). Points in black correspond to fluviometric stations. The zoom-in panel shows details of hydrological zones (e.g. SDO) and their respective sub-zones (e.g., SDO-1, SDO-2 and SDO-3) for three catchments in the Central Pacific slope.
Fig. 4 On the left side: climatic (P: precipitation; TA: air temperature; PET: potential evapotranspiration; AI: aridity index) and physiographic (SLO: surface slope; CLAY: soil clay content; fTC: fraction of tree cover) predictor maps. On the right side: colored areas represent gauged catchments and sub-catchments (donors of model parameters), and white areas correspond to ungauged sub-catchments (receptors of model parameters). Points in black correspond to fluviometric stations. The zoom-in panel shows details of hydrological zones (e.g. SDO) and their respective sub-zones (e.g., SDO-1, SDO-2 and SDO-3) for three catchments in the Central Pacific slope.CC-BY · From the PDF
Fig. 5 (a-d) Spatial distribution of metrics for the national hydrological model performance assessment during calibration, validation, and total period (1981 -2020). Each metric is focused on a specific aspect of hydrographs (KGE: middle-high flows; NSEsqrt: general flows; MARE and PBIAS: model error and bias). Colored points represent gauges ' performance.
Fig. 5 (a-d) Spatial distribution of metrics for the national hydrological model performance assessment during calibration, validation, and total period (1981 -2020). Each metric is focused on a specific aspect of hydrographs (KGE: middle-high flows; NSEsqrt: general flows; MARE and PBIAS: model error and bias). Colored points represent gauges ' performance.CC-BY · From the PDF
Image 7
CC-BY · From the PDF
Fig. 7 (a-f) Comparison between observed (blue) and simulated (gray) flow series at daily [mm/day], monthly [mm/month] and yearly [mm/ year] time scales for six fluviometric stations with the highest record length during 1981 -2020.
Fig. 7 (a-f) Comparison between observed (blue) and simulated (gray) flow series at daily [mm/day], monthly [mm/month] and yearly [mm/ year] time scales for six fluviometric stations with the highest record length during 1981 -2020.CC-BY · From the PDF
Fig. 8 (a-i) Standard error bars between nine predicted and observed hydrological signatures in gauged catchments. Standard error was computed in four intervals of observed flow record coverage (0 -25%, 25 -50%, 50 -75%, and 75 -100%) for 1981 -2020. Points in magenta represent the median for each interval.
Fig. 8 (a-i) Standard error bars between nine predicted and observed hydrological signatures in gauged catchments. Standard error was computed in four intervals of observed flow record coverage (0 -25%, 25 -50%, 50 -75%, and 75 -100%) for 1981 -2020. Points in magenta represent the median for each interval.CC-BY · From the PDF
Fig. 9 Qualitative ratings of daily flow simulation on gauged catchments in the study domain based on (a) high and middle flows and (b) general flows performances in the national hydrological model.
Fig. 9 Qualitative ratings of daily flow simulation on gauged catchments in the study domain based on (a) high and middle flows and (b) general flows performances in the national hydrological model.CC-BY · From the PDF

Verification

Claims the paper makes

0 Valid · 1 Not valid · 39 Not checked · 116 Own results and interpretations

85 more claims

Create a free account to see all 156 claims and to verify the 42 references still open.

See all claims — free account
  • It may state two separate claims as one. 37
  • Probably not a substantial claim. 41

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

3 of 3

Citations

Cites

Cited by

See all 29

Metadata

Authors
Llauca, H., Leon, K., & Lavado-Casimiro, W.
Affiliations
Servicio Nacional de Meteorología e Hidrología del Perú (PE)
Corresponding author
Llauca, Harold
Venue
Journal of Hydrology: Regional Studies
Publisher
Elsevier BV
Published
2023-06
Volume
47
Pages
101381
ISSN
2214-5818 (ISSN-L)
OpenAlex
W4365140647
Cited by
29
References
90
Citations by year
2026: 13 · 2025: 8 · 2024: 4 · 2023: 4
FWCI
2.03
Type
journal article
Language
en
Open access
Licence
Locations
  • Journal of Hydrology Regional Studies (publishedVersion, cc-by)
  • LA Referencia (Red Federada de Repositorios Institucionales de Publicaciones Científicas) (acceptedVersion, cc-by-nc-nd)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Greater South Information System
Topics
  • Physical Sciences › Environmental Science › Water Science and Technology › Hydrology and Watershed Management Studies
  • Physical Sciences › Environmental Science › Management, Monitoring, Policy and Law › Environmental Monitoring and Data Management
  • Physical Sciences › Environmental Science › Environmental Engineering › Soil Moisture and Remote Sensing
Keywords
daily streamflow, hydrological regionalization, Peru, ungauged basins, transboundary catchment, flood risk, water resources planning

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-11-19, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-19 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.