Bibliographic Details
| Title: |
Early Warning for Stepwise Landslides Based on Traffic Light System: A Case Study in China |
| Authors: |
Shuangshuang Wu, Zhigang Tao, Li Zhang, Song Chen |
| Source: |
Remote Sensing, Vol 16, Iss 23, p 4391 (2024) |
| Publisher Information: |
MDPI AG, 2024. |
| Publication Year: |
2024 |
| Collection: |
LCC:Science |
| Subject Terms: |
traffic light system (TLS), jerk, threshold, stepwise, early warning, closed-loop, Science |
| Description: |
The phenomenon of stepwise landslides, characterized by displacement exhibiting a step-like pattern, is often influenced by reservoir operations and seasonal rainfall. Traditional early warning models face challenges in accurately predicting the sudden initiation and cessation of displacement, primarily because conventional indicators such as rate or acceleration are ineffective in these scenarios. This underscores the urgent need for innovative early warning models and indicators. Viewing step-like displacement through the lens of three phases—stop, start, and acceleration—aligns with the green-yellow-red warning paradigm of the Traffic Light System (TLS). This study introduces a novel early warning model based on the TLS, incorporating jerk, the derivative of displacement acceleration, as a critical indicator. Empirical data and theoretical analysis validate jerk’s significance, demonstrating its clear pattern before and after step-like deformations and its temporal alignment with the deformation’s conclusion. A comprehensive threshold network encompassing rate, acceleration, and jerk is established for the TLS. The model’s application to the Shuiwenzhan landslide case illustrates its capability to signal in a timely manner the onset and acceleration of step-like deformations with yellow and red lights, respectively. It also uniquely determines the deformation’s end through jerk differential analysis, which is a feat seldom achieved by previous models. Furthermore, leveraging the C5.0 machine learning algorithm, a comparison between the predictive capabilities of the TLS model and a pure rate threshold model reveals that the TLS model achieves a 93% accuracy rate, outperforming the latter by 7 percentage points. Additionally, in response to the shortcomings of existing warning and emergency response strategies for this landslide, a closed-loop management framework is proposed, grounded in the TLS. This framework encompasses four critical stages: hazard monitoring, warning issuance, emergency response, and post-event analysis. We also suggest support measures to ensure implementation of the framework. |
| Document Type: |
article |
| File Description: |
electronic resource |
| Language: |
English |
| ISSN: |
2072-4292 |
| Relation: |
https://www.mdpi.com/2072-4292/16/23/4391; https://doaj.org/toc/2072-4292 |
| DOI: |
10.3390/rs16234391 |
| Access URL: |
https://doaj.org/article/6fd15211f5b446c58a1151729420fb09 |
| Accession Number: |
edsdoj.6fd15211f5b446c58a1151729420fb09 |
| Database: |
Directory of Open Access Journals |