Landslide susceptibility modeling by interpretable neural network
Landslide susceptibility modeling by interpretable neural network
复制标题
DOI：
10.1038/s43247-023-00806-5
复制
发表时间：
2022-01
期刊：
影响因子：
--
通讯作者：
Khaled Youssef;K. Shao;S. Moon;Louis-S. Bouchard
中科院分区：
文献类型：
--
作者：
Khaled Youssef;K. Shao;S. Moon;Louis-S. Bouchard
众所周知，滑坡很难预测，因为许多在空间和时间上变化的因素都有助于斜坡的稳定性。人工神经网络(ANN)已被证明可以提高预测精度，但在很大程度上是无法解释的。在这里，我们介绍了一个附加的人工神经网络优化框架来评估滑坡的易感性，以及数据集划分和结果解释技术。我们将具有完全可解释性、高精度、高泛化能力和低模型复杂性的方法称为可叠加神经网络(SNN)优化。我们通过训练三个不同喜马拉雅地区的滑坡清单模型来验证我们的方法。我们的SNN的性能优于基于物理和统计的模型，并实现了与最先进的深度神经网络相似的性能。SNN模型发现，坡度和降雨量以及坡度的乘积是滑坡易感性的重要主要因素，这突显了强烈的坡度-气候耦合以及小气候对滑坡发生的重要性。
Landslides are notoriously difficult to predict because numerous spatially and temporally varying factors contribute to slope stability. Artificial neural networks (ANN) have been shown to improve prediction accuracy but are largely uninterpretable. Here we introduce an additive ANN optimization framework to assess landslide susceptibility, as well as dataset division and outcome interpretation techniques. We refer to our approach, which features full interpretability, high accuracy, high generalizability and low model complexity, as superposable neural network (SNN) optimization. We validate our approach by training models on landslide inventories from three different easternmost Himalaya regions. Our SNN outperformed physically-based and statistical models and achieved similar performance to state-of-the-art deep neural networks. The SNN models found the product of slope and precipitation and hillslope aspect to be important primary contributors to high landslide susceptibility, which highlights the importance of strong slope-climate couplings, along with microclimates, on landslide occurrences.