澜沧江自然条件下输沙质量通量与体积径流量的关系

为了研究澜沧江水沙关系和提高利用体积径流量预测泥沙质量通量的准确性，以电站建设前1982—2000年澜沧江上游旧州站和中游戛旧站沙的质量浓度和体积径流量实测资料为研究对象，根据自然状态下泥沙质量通量和体积径流量变化幅度的差异，建立输沙质量通量导数与体积径流量的关系式. 公式系数为流域系统属性的综合反映，表示平均体积径流量下输沙质量通量的变化率，公式指数表示体积径流量变化对输沙质量通量变化率的影响. 根据上中游站的指数差异可知，泥沙质量通量除依赖上游来沙外还依赖沿程冲刷和支流入汇的补给. 通过对输沙质量通量导数公式进行积分得出输沙质量通量与体积径流量的理论关系. 结果表明，输沙质量通量峰值约落后体积径流量峰值1 d，据此优化反向传播神经网络（BP-NN），可以较好地改良优化前模型预测中的峰值偏移现象，提高预测精度.

Relation between sediment mass flux and volume runoff under natural condition of Lancang River

The relationship between sediment mass flux derivative and volume runoff was developed according to the difference in rangeability of sediment mass flux and volume runoff under natural condition. Based on the measured data of sediment mass concentration and volume runoff of the upriver Jiuzhou station and the middleriver Gajiu station of Lancang River from 1982 to 2000 before the construction of the power station, this work studied the relation between water and sediment in Lancang River and improved the accuracy of using the volume runoff to predict the sediment mass flux. The coefficient of formula comprehensively reflectes the property of the basin system and represents the change rate of sediment mass flux at mean volume runoff. The index of formula reflectes the influence of volume runoff change on the change rate of sediment mass flux. The difference of indexes between upriver and middleriver stations shows that the sediment mass flux not only depends on the upstream sediment, but also on the alongshore erosion and the recharge of tributaries. The theoretical relationship between sediment mass flux and volume runoff is obtained by the integration of derivative formula of sediment mass flux. Results show that the peak of volume runoff appeares one day before the peak of sediment mass flux, based on which the back-propagation neural network (BP-NN) can be optimized. The optimized method can be used to better improve the peak shift phenomenon predicted by the original model and improve the prediction accuracy.