中文摘要：

黄河深切于青藏高原东北缘的构造山地中，留下了山地和黄河形成演化的记录，但其高阶地年龄因超出常规测年技术的范围而存在困难。文章尝试用原地成因宇宙核素^10Be和^26Al以及垂直剖面取样技术，通过剖面X^2最小统计同时求出了年龄、侵蚀速率、继承浓度以及阶地沉积物密度等参数，所获得的黑山峡口247m最高基座阶上口地的参考年龄和剥蚀速率分别约为2．4Ma和15．5＋8/-4m／Ma，这是该河段高阶地测年的一次新探索。根据这一结果及相关的构造地貌特征，认为香山一天景山断裂带西段2．4Ma前以逆冲为主，以后逐渐过渡到左旋走滑，该段黄河诞生于逆冲作用减弱时期。

英文摘要：

The northeastern margin of Tibetan Plateau is characterized by very active thrust and strike-slip fauhs, which contribute mainly to the uplifting of mountain ranges and subsidence of basins along the margin. The Yellow River flows through the faults and cuts into the mountain ranges. Some terraces were uplifted several hundred meters higher above the modern river, while others were offset several kilometers off the modem river. Obviously, these terraces recorded the activities of the fauhs and the evolutionary history of the mountains, basins and the river itself. To study these records, dating these terraces is very important. However, it is difficuh to date those high level terraces by conventional techniques because of the dating range limits. Terrestrial in situ cosmogenic nuclides （TCN） of ^10Be and ^26Al and vertical sample profile techniques, which take nucleon spallation, negative muon capture, and fast muon reactions account at the same time, can be used to date river terrace sediments over million- year timescales even though terraces are not well preserved. We use these techniques to date alluvium preserved on the highest terrace of 247m above the Yellow River at the area of the Heishanxia Gorge mouth in the northeast margin of Tibetan Plateau. Three amalgamated samples （consists of 48, 74, 70 quartzite clasts respectively） ranging in depth from 0, 2m to 4m are collected and measured in PRIME Lab, Purdue University. By profile fitting of X^2 minimization solving for age, erosion rate, inheritance as well as bulk density, we have obtained a preliminary resuh of age of about 2.4Ma and a 15.5 ＋8/-4m/Ma erosional rate for the terrace. The erosion rate is similar to those in other arid mountainous areas. The age is much older than the calcium carbonate coats dating on the same terrace （1.57Ma）. It is probably because that the coats dating was from samples collected from erosionally removed terrace, rather than the oldest ones. Although probably due to insufficient sampling restricted by prese