We have addressed the problem of measuring the compressional wave velocity at a very shallow depth in unconsolidated dune sand. Because the overburden stress is very small at shallow depths, the respective velocity is small and the seismic signal is weak. This is why such data are scarce, in the lab and in the field. Our approach is to stage a high-resolution seismic experiment with a dense geophone line with spacing varying between 10 and 25 cm, allowing us to produce a velocity-depth relation in the upper 1 m interval. These results are combined with another survey in which the geophone spacing is 2 m and the dominant frequency is an order of magnitude lower than in the first survey. The latter results give us the velocity profile in the deeper interval between 1 and 7 m, down to the base of the dune. The velocity rapidly increases from about 48 m/s in the first few centimeters to 231 m/s at 1 m depth and then gradually increases to 425 m/s at 7 m depth. This is the first time when such a low velocity has been recorded at extremely shallow depths in sand in situ. The velocity profile thus generated is statistically fitted with a simple analytical equation. Our velocity values are higher than those published previously for beach sand. We find that using replacement or tomogram velocities instead of an accurately measured velocity profile may result in 23%–44% error in the static correction.

中文翻译：

---

沙丘中很浅深度的P波速度分布

我们已经解决了在未固结的沙丘中非常浅的深度处测量压缩波速度的问题。由于在浅层深度上覆应力很小，因此相应的速度很小，地震信号也很弱。这就是为什么在实验室和现场都缺乏此类数据的原因。我们的方法是使用间距在10到25 cm之间的密集检波器线进行高分辨率地震实验，从而使我们能够在1 m的较高区间产生速度-深度关系。这些结果与另一次调查相结合，在该调查中，检波器间距为2 m，并且主频比第一次调查低一个数量级。后者的结果为我们提供了从1到7 m的更深间隔直到沙丘底部的速度分布图。速度从前几厘米的约48 m / s迅速增加到1 m深度的231 m / s，然后逐渐增加到7 m深度的425 m / s。这是第一次在沙地极浅的深度处记录到如此低的速度。这样生成的速度曲线在统计上适合一个简单的解析方程。我们的速度值高于以前发布的海滩沙子的速度值。我们发现，使用置换速度或断层图速度而不是精确测量的速度曲线可能会导致静态校正中23％–44％的误差。如此生成的速度曲线在统计上拟合有一个简单的解析方程。我们的速度值高于先前发布的海滩沙尘速度值。我们发现，使用置换速度或层析成像速度而不是精确测量的速度曲线可能会导致静态校正误差23％–44％。如此生成的速度曲线在统计上拟合有一个简单的解析方程。我们的速度值高于以前发布的海滩沙子的速度值。我们发现，使用置换速度或断层图速度而不是精确测量的速度曲线可能会导致静态校正中23％–44％的误差。