Using short-interval landslide inventories to build short-term and overall spatial prediction models for earthquake-triggered landslides based on machine learning for the 2018 Lombok earthquake sequence

Abstract

During an earthquake sequence, there are often multiple recurring landslides. Understanding the spatial distribution of the landslides triggered by the rst earthquake can help us predict the landslide susceptibility for subsequent shakes over a short term. This study used two landslide inventories from the Lombok earthquake sequence in Indonesia in 2018 to construct a short-term secondary disaster prediction model and an overall spatial prediction model using four machine learning algorithms. The average accuracy of the positive samples predicted by the prediction model was 7.1% lower than that of the short-term model. The highest accuracy of the overall prediction model was 14.9% higher, on average, and the area under the ROC curve (AUC) score was 8.1% higher, on average, but the corresponding probability thresholds were lower. The reason for this difference is that, in the short-term prediction model, since most of the landslides in the rst landslide inventory were prone to fail two or more times due to the effect of multiple earthquakes, the prediction results have a high positive rate. This feature of the short-term prediction model makes it suitable for landslide rescue guidance in a sequence of earthquakes. In contrast, the overall prediction model can better represent the spatial distribution of the earthquake-triggered landslides in the area.

Environmental effects following a seismic sequence: the 2019 Cotabato - Davao del Sur (Philippines) earthquakes

“…Overall, we claim that the multi-temporal inventory developed in this study may be useful in the analysis of susceptibility to EEEs for subsequent earthquakes over a short-term (e.g., Burrows et al, 2020;Xue et al, 2022) and the mutual role of seismic shaking and rainfall (e.g., Fan et al, 2020;Tanyas et al, 2021).…”

Artificial Intelligence-Based Landslide Studies in Indonesia: A Systematic Review in Recent Years

“…The other popular AI method that is commonly used for building LSM is Logistic Regression (LR). At least four papers use this multivariate statistical approach, where several researchers combine LR with another AI method, like ANN, DT, or Frequency Ratio (FR), to generate LSM [6], [8], [11], [14]. LSM produced by LR has an accuracy range of 68.7-85% [8], [11].…”

“…LSM produced by LR has an accuracy range of 68.7-85% [8], [11]. Other less popular AI methods that used for generating LSM in Indonesia are ANN, FR, and DT [6], [8], [12]. Several modifications of ANN, such as Certain factor ANN and Deep Neural Network, were developed to do the same analysis [6], [12].…”

“…High-quality landslide map, with accuracy between 85->95%, were produced using many kinds of AI approach like Logistic Regression (LR), Support Vector Machine (SVM), Decision Tree (DT), and Artificial Neural Network (ANN). In many cases, their performance is also combated and compared to find the most powerful one [6], [8], [11]. This "competition" also, unconsciously, helps to develop the production of a landslide model for mitigation matter.…”

“…Landslide research, such as susceptibility mapping, always uses many parameters to build a high-quality map. Landslide susceptibility mappings using the AI approach in Indonesia also, in several cases, include feature selection or modification in their procedures [6], [11], [12]. It indeed can develop the performance of the model or map.…”