Sediment accumulation rates and thermal trackers suggest a substantial and global increase in erosion rates over the past few million years. That increase is commonly associated with the impact of the Northern Hemisphere glaciation, but methodological biases have led researchers to debate this hypothesis. Here, we test whether Himalayan erosion rates increased by measuring beryllium-10 (¹⁰Be) in the sediment of the Bengal Bay seabed. Sediment originated from rocks that produced ¹⁰Be under the impact of cosmic rays during erosion near surface. Thus, the ¹⁰Be concentrations indicate erosion rates. The ¹⁰Be concentration of the Bengal Bay sediment depends on the contributions of the Ganga and Brahmaputra rivers. Their sediments have distinct ¹⁰Be concentrations because of distinct elevations and erosion in their drainage basins. Variable contributions could thus complicate erosion-rate calculation. We traced these contributions by a provenance study using the strontium (Sr) and neodymium (Nd) isotopic sediment compositions. Within uncertainties of ±30%, our reconstructed past erosion rates show no long-term increase for the past six million years. This stability suggests that climatic changes during the late Cenozoic have an undetectable impact on the erosion patterns in the Himalayas, at least on the ten thousand to million year timescales accounted for by our dataset.

沉积物累积速率和热追踪仪表明，过去几百万年以来，全球侵蚀率大幅度提高。这种增加通常与北半球冰河的影响有关，但是方法上的偏见导致研究人员对这一假设进行了辩论。在这里，我们通过测量孟加拉湾海底沉积物中的铍10（¹⁰ Be）来测试喜马拉雅河侵蚀速率是否增加。沉积物起源于在近地表侵蚀期间在宇宙射线的作用下产生¹⁰ Be的岩石。因此，¹⁰ Be浓度表示腐蚀速率。在¹⁰孟加拉湾沉积物的浓度取决于恒河和雅鲁藏布江的贡献。由于其流域内不同的海拔和侵蚀，它们的沉积物具有不同的¹⁰ Be浓度。因此，变量的贡献可能会使侵蚀率的计算复杂化。我们通过使用锶（Sr）和钕（Nd）同位素沉积物成分的物产研究来追踪这些贡献。在±30％的不确定性范围内，我们重建的过去侵蚀率显示过去六百万年没有长期增长。这种稳定性表明，新生代末期的气候变化对喜马拉雅山的侵蚀模式没有可察觉的影响，至少对我们的数据集所说明的一万到一百万年的时间尺度没有影响。