This paper presents pyNewmarkDisp, an open-source application software developed in Python for calculating the spatial distribution of permanent displacements ($u_{\mathrm{p}}$) in earthquake-triggered shallow landslides through the Newmark sliding block method. $u_{\mathrm{p}}$are often estimated by simplified methods using empirical correlations instead of the direct Newmark method in which ground motion accelerations exceeding a critical value are integrated twice. These simplified procedures have traditionally been justified because of the time consumption, labor intensiveness, and impracticality of the direct Newmark method at large spatial scales. However, pyNewmarkDisp has a well-optimized code taking advantage of NumPy array operations and just-in-time Numba compilation that allows performing the direct Newmark method at large-scale spatial domains, even on general-purpose computers. pyNewmarkDisp is exemplified through a hypothetical study case, and its results are compared to three empirical correlations, showing significant differences and confirming that simplified methods can lead to inaccurate results.

本文介绍了pyNewmarkDisp，一个用 Python 开发的开源应用软件，用于计算永久位移的空间分布（${你}_{\mathrm{p}}$）通过纽马克滑块法处理地震引发的浅​​层滑坡。${你}_{\mathrm{p}}$通常使用经验相关性的简化方法来估计，而不是直接的纽马克方法，其中超过临界值的地震动加速度被积分两次。这些简化的程序传统上是合理的，因为直接纽马克方法在大空间尺度上耗时、劳动密集且不切实际。然而，pyNewmarkDisp具有充分优化的代码，利用 NumPy 数组操作和即时 Numba 编译，允许在大规模空间域甚至通用计算机上执行直接 Newmark 方法。pyNewmarkDisp通过一个假设的研究案例进行了举例说明，其结果与三个经验相关性进行了比较，显示出显着差异，并证实简化的方法可能会导致不准确的结果。