Jing et al. (2026) Spatiotemporal Variability of Runoff Coefficients and Rainfall–Runoff Responses in a Mountainous Basin Using Integrated Hydrological Analogy and SCS–CN Models

Identification

• Journal: Water Resources Management
• Year: 2026
• Date: 2026-02-25
• Authors: Zhang Jing, Ping Yang, Chen Zi-jing, Peng Shi-tao, Xing Bing
• DOI: 10.1007/s11269-026-04555-w

Research Groups

• Colleges of Hehai, Chong Qing Jiao Tong University, Chong Qing, China
• Laboratory of Environmental Protection in Water Transport Engineering, Tianjin Research Institute of Water Transport Engineering, Tianjin, China

Short Summary

This study developed a three-stage coupling framework integrating hydrological analogy, SCS-CN calibration, and reciprocal modeling to quantify spatiotemporal runoff coefficient variability in a mountainous basin, demonstrating enhanced predictive reliability and mechanistic interpretability for runoff estimation under data-scarce conditions.

Objective

• To quantify spatiotemporal variability of runoff coefficients (RCs) and to diagnose rainfall–runoff mechanisms across land-use types in a mountainous basin using an integrated hydrological analogy and SCS–CN models framework.

Study Configuration

• Spatial Scale: Fenggao River basin (Southwest China), a first-order tributary of the Leixi River, with a catchment area of 73.81 km². The basin was delineated into 8 sub-watersheds. Rainfall data was downscaled to a 30 m × 30 m grid, and land-use data was at 30 m × 30 m resolution.
• Temporal Scale: Monthly rainfall data for 2023. Landsat 8–9 imagery acquired in July 2023. Runoff estimates and runoff coefficients were calculated on monthly and annual bases for 2023.

Methodology and Data

• Models used:
  • Hydrological analogy method (area-proportion method)
  • Soil Conservation Service Curve Number (SCS-CN) method
  • Reciprocal model (1/RC = a + b/Pre)
• Data sources:
  • Monthly rainfall data for 2023 from the National Tibetan Plateau Scientific Data Center (0.0083333° spatial resolution, approximately 1 km), downscaled to 30 m using Inverse Distance Weighting (IDW).
  • Landsat 8–9 imagery (July 2023, 30 m spatial resolution) for land-use classification (seven categories: transportation land, shrubland, forest land, grassland, water bodies, cropland, and built-up land).
  • Long-term runoff records from the Yutan Hydrological Station (drainage area of 865 km²) for hydrological analogy-based runoff estimation.
  • Field survey plots, high-resolution Google Earth historical imagery, and UAV orthophotos for land-use classification training samples.

Main Results

• The calibrated SCS-CN model showed strong agreement with runoff estimates derived from the hydrological analogy method (correlation coefficient r = 0.9732, coefficient of determination R² = 0.823, Root Mean Square Error RMSE = 0.165, Mean Absolute Error MAE = 0.119).
• Runoff coefficients exhibited pronounced seasonal variability, ranging from 7.12% in January to 48.64% in July, with an annual mean basin coefficient of 35.86%.
• Land-use heterogeneity exerted significant control on runoff generation:
  • Water bodies (mean RC: 84.37%) and transportation land (mean RC: 67.69%) exhibited high and stable RCs.
  • Cropland (mean RC: 34.40%) and built-up land (mean RC: 32.50%) showed transitional responses.
  • Forest (mean RC: 22.39%), grassland (mean RC: 25.53%), and shrubland (mean RC: 16.63%) functioned as hydrological buffers.
• The reciprocal model successfully captured nonlinear rainfall–runoff dynamics across all land-use categories (R² > 0.85), providing physically interpretable parameters:
  • Water bodies (1/a ≈ 102) and transportation land (1/a ≈ 102) had the highest potential maximum RCs and lowest rainfall sensitivity (b values of 0.1064 and 0.1069, respectively).
  • Shrubland (1/a ≈ 79) and forest land (1/a ≈ 103) displayed the highest rainfall sensitivity (b values of 3.7447 and 2.7478, respectively), indicating strong buffering under dry conditions and rapid response under intense rainfall.
• The basin-wide average reciprocal model parameters were a = 0.0137 and b = 1.7817, corresponding to a maximum potential RC of 1/a ≈ 72.9, with a high R² of 0.972.

Contributions

• Advanced a transferable, structured three-stage coupling framework that integrates empirical regional runoff scaling, process-based SCS-CN calibration, and nonlinear reciprocal modeling for runoff estimation in data-scarce mountainous basins.
• Enhanced both the predictive reliability and mechanistic interpretability of runoff processes by explicitly linking land-use structure to rainfall sensitivity through physically interpretable parameters of the reciprocal model.
• Provided a practical diagnostic and predictive framework for watershed management, flood forecasting, and scenario analysis under the combined pressures of accelerating urbanization and climate variability.

Funding

• National Key Research and Development Plan Project (2024YFC3712304)
• Key projects of Chong Qing Municipal Education Commission (KJZD-K202400707)

Citation

@article{Jing2026Spatiotemporal,
  author = {Jing, Zhang and Yang, Ping and Zi-jing, Chen and Shi-tao, Peng and Bing, Xing},
  title = {Spatiotemporal Variability of Runoff Coefficients and Rainfall–Runoff Responses in a Mountainous Basin Using Integrated Hydrological Analogy and SCS–CN Models},
  journal = {Water Resources Management},
  year = {2026},
  doi = {10.1007/s11269-026-04555-w},
  url = {https://doi.org/10.1007/s11269-026-04555-w}
}