Assessing Mountainous River Basin Nonpoint Source Pollution Using Event-Aware SWAT+ Calibration with HO–SUFI-2: Qingjiang River Basin, China
Year: 2025
Type: preprint
Abstract: Storm-pulse non-point source pollution in the mountainous Qingjiang River Basin is analyzed to capture event-driven nutrient dynamics while ensuring multi-year robustness for nitrogen and phosphorus management. We develop a daily-forced, monthly-evaluated SWAT+ hydro–water-quality model (2008–2018) calibrated via a two-stage scheme—Hippopotamus Optimization for global search followed by SUFI-2 under 95PPU to contract posteriors. An event-aware objective blends monthly NSE terms for discharge (Q), TN, TP, NH4+-N, and NO3−-N with soft daily-window penalties on peak magnitude, time-to-peak, and recession slope. Calibration yielded flow NSE = 0.79, R2 = 0.80, PBIAS = −5.3%, P-factor = 87%, R-factor = 1.09; nutrient NSEs were 0.76 (NH4+-N), 0.78 (NO3−-N), and 0.73 (TP). Out-of-sample testing (2020–2024) without re-calibration sustained performance—flow NSE = 0.75, R2 = 0.77, PBIAS = −6.1%, P-factor = 85%, R-factor = 1.15; nutrient NSEs 0.73/0.75/0.71—with station-level transferability (Gaobazhou Q/TN/TP NSE = 0.67/0.52/0.45; Geheyan 0.62/0.46/0.42). Seasonally, storms governed concentration extremes; spatially, indicators intensified from forested headwaters to the mid-lower agricultural–urban corridor, where cropland dominated sources (urban ≈12–15%, grassland <10%). The proposed event-aware SWAT+ objective, coupled with HO–SUFI-2, improves efficiency, preserves storm timing, and narrows predictive uncertainty. Findings support actionable measures—detaining 10–20% of event flow for 6–12 h, installing 15–30 m riparian buffers, and pausing fertilization 48–72 h before forecast storms—offering a transferable pathway for robust cross-period assessment and policy design in similar mountainous basins.
Source: SSRN Electronic Journal
DOI: 10.2139/ssrn.5531693
Authors: Ruining Wang, Xianqi Zhang, Hongyang Zhang, Yanbin Yang, Zelin Tao
Language: English
Subfield: Water Science and Technology
Field: Environmental Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4415028778