金沙江下游区间来沙驱动因子分析及产沙预测模型

Factors controlling sediment yield and prediction of ungauged areas in the Lower Jinsha River basin

随着长江上游梯级开发的快速推进,金沙江下游干支流来沙占比失衡现象日益突出,区间水沙实测资料欠缺的问题严重影响了对干流库群泥沙淤积的预测。为系统性了解并掌握金沙江下游区间内泥沙分布全貌,明晰区间来沙时空异性特征以及归因,基于已有支流水文站泥沙序列与因子集,采用秩相关分析研究了区域内地形、气候、土地覆被、空间尺度各类因子对流域产沙的潜在影响和各因子之间的变化独立性。通过参数降维与双重回归分析相结合的方法识别关键因子组合并构建输沙模数预测模型。研究结果表明:8°以上坡度占比、气温和集水面积的驱动因子组合能够较为完整地解释研究区域产沙机制,并在此基础上建立了能解释约92%输沙模数变化性的预测模型。根据模型计算得到广大无实测资料区域的输沙模数分布范围为87~1189 t/(km^(2)·a),近50 a来减小幅度约50~300 t/(km^(2)·a)。在此基础上识别白鹤滩、溪洛渡高产沙及上下游低产沙区间,并分析了该空间不均衡特征的削弱趋势。

Recent installment of cascade reservoirs in the Lower Jinsha River basin urges quantitative studies of sediment inflows from ungauged tributaries and sediment pattern in this basin.Based on existing observations from 18 gauging stations and a multi-factor dataset,the Spearman rank correlation analysis was used to examine the responses of sediment yield to potential controlling factors at the catchment scale.Dimensionality reduction and regression methods were integrated to identify different sets of factor variables for predicting specific sediment yield(SSY).The results show that proportion of slopes over 8°,temperature and catchment area together can best describe the regional mechanism of sediment yield,and the obtained model explained 92%of SSY variability.The estimated SSY of ungauged areas varied from 87 t/(km^(2)·a)to 1189 t/(km^(2)·a)and has reduced by 50—300 t/(km^(2)·a)over the past 50 years.This research detected both high-yield and low-yield tributaries in the Baihetan and Xiluodu reservoirs,as well as the declining trend of SSY spatial heterogeneity.