高寒流域融雪径流模拟改进及关键驱动因子分析

Improvement of snowmelt runoff simulation and analysis of key driving factors in alpine river basins

摘要

【目的】针对分布式水文模型SPHY在寒区融雪径流模拟中物理机制不足(忽略太阳辐射驱动、采用静态度日因子)的问题,优化增强其融雪模块,提升寒区融雪径流模拟精度。【方法】主要进行三个方面改进:(1)构建辐射耦合的融雪潜在能量项;(2)提出基于雪深—太阳辐射关系的季节性动态度日因子;(3)基于随机森林等量化不同因子在融雪过程中的重要性。以青藏高原东南部尼洋河流域为研究区,基于2010—2015年气象、水文及遥感数据,验证改进后模型的模拟性能,并通过趋势分析、相关性分析及随机森林模型,量化不同气象因子对融雪径流的贡献率。【结果】模型模拟结果:率定期与验证期内,改进后的模型融雪期平均纳什效率系数(NSE)提升8%,均方根误差(RMSE)降低17%,整体百分比偏差(PBIAS)由-14.77%优化至-0.05%。驱动因子分析显示,气温是融雪径流的主导因子(贡献率41.79%),风速贡献率达25.58%;能量因子(气温+辐射)总贡献率达61.08%。【结论】耦合太阳辐射与季节性动态度日因子的改进方案,有效增强了SPHY模型在寒区的径流模拟能力。研究成果可为其他高寒流域径流模拟提供一定的参考,为气候变化下高寒流域径流预报及水资源管理等提供一定的科学支撑。

Abstract

[Objective] To address the issue of insufficient physical mechanisms in the distributed hydrological model SPHY for snowmelt runoff simulation in cold regions(specifically the neglect of solar radiation and the use of static degree-day factors), the snowmelt module is optimized and enhanced to improve the accuracy of snowmelt runoff simulation in cold regions. [Methods]Three improvements were made:(1) constructing a radiation-coupled potential snowmelt energy term;(2) proposing a seasonal dynamic degree-day factor based on the snow depth-solar radiation relationship; and(3) quantifying the importance of different factors during the snowmelt process based on random forest. Taking the Niyang River Basin in the southeastern Qinghai-Xizang Plateau as the study area, the simulation performance of the improved model was verified based on meteorological, hydrological, and remote sensing data from 2010 to 2015. The contribution rates of different meteorological factors to snowmelt runoff were quantified through trend analysis, correlation analysis, and random forest models. [Results] The model simulation result showed that during the calibration and validation periods, the average Nash-Sutcliffe efficiency(NSE) of the improved model during the snowmelt period increased by 8%, the root mean square error(RMSE) reduced by 17%, and the overall Percent bias(PBIAS) was optimized from-14. 77% to-0. 05%. Driving factor analysis showed that air temperature was the dominant factor for snowmelt runoff(contribution rate of 41. 79%), and the wind speed contribution rate reached 25. 58%. The total contribution rate of energy factors(air temperature and radiation) reached 61. 08%. [Conclusion] The improved scheme coupling solar radiation with the seasonal dynamic degree-day factor effectively enhances the runoff simulation capability of the SPHY model in cold regions. The findings provide references for runoff simulation in other alpine river basins, and offer scientific support for runoff forecasting and water resource management under climate change.

关键词

SPHY模型 / 融雪径流 / 动态度日因子 / 辐射耦合 / 尼洋河流域 / 高寒流域 / 驱动因子分析 / 随机森林

Key words

SPHY model / snowmelt runoff / dynamic degree-day factor / radiation coupling / Niyang River Basin / alpine watershed / driving factor analysis / Random Forest(RF)

基金资助

国家自然科学基金面上项目(52479021)