Landslide Displacement Prediction via Attentive Graph Neural Network
Landslides are among the most common geological hazards that result in considerable human and economic losses globally. Researchers have put great efforts into addressing the landslide prediction problem for decades. Previous methods either focus on analyzing the landslide inventory maps obtained fr...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-04-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/8/1919 |
| _version_ | 1797434229237219328 |
|---|---|
| author | Ping Kuang Rongfan Li Ying Huang Jin Wu Xucheng Luo Fan Zhou |
| author_facet | Ping Kuang Rongfan Li Ying Huang Jin Wu Xucheng Luo Fan Zhou |
| author_sort | Ping Kuang |
| collection | DOAJ |
| description | Landslides are among the most common geological hazards that result in considerable human and economic losses globally. Researchers have put great efforts into addressing the landslide prediction problem for decades. Previous methods either focus on analyzing the landslide inventory maps obtained from aerial photography and satellite images or propose machine learning models—trained on historical land deformation data—to predict future displacement and sedimentation. However, existing approaches generally fail to capture complex spatial deformations and their inter-dependencies in different areas. This work presents a novel landslide prediction model based on graph neural networks, which utilizes graph convolutions to aggregate spatial correlations among different monitored locations. Besides, we introduce a novel locally historical transformer network to capture dynamic spatio-temporal relations and predict the surface deformation. We conduct extensive experiments on real-world data and demonstrate that our model significantly outperforms state-of-the-art approaches in terms of prediction accuracy and model interpretations. |
| first_indexed | 2024-03-09T10:30:13Z |
| format | Article |
| id | doaj.art-7b375de419e040c0ad18e6b981160dab |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T10:30:13Z |
| publishDate | 2022-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-7b375de419e040c0ad18e6b981160dab2023-12-01T21:22:12ZengMDPI AGRemote Sensing2072-42922022-04-01148191910.3390/rs14081919Landslide Displacement Prediction via Attentive Graph Neural NetworkPing Kuang0Rongfan Li1Ying Huang2Jin Wu3Xucheng Luo4Fan Zhou5School of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu 610054, ChinaSchool of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu 610054, ChinaCHN Energy Dadu River Big Data Service Co., Chengdu 610054, ChinaSchool of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu 610054, ChinaSchool of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu 610054, ChinaSchool of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu 610054, ChinaLandslides are among the most common geological hazards that result in considerable human and economic losses globally. Researchers have put great efforts into addressing the landslide prediction problem for decades. Previous methods either focus on analyzing the landslide inventory maps obtained from aerial photography and satellite images or propose machine learning models—trained on historical land deformation data—to predict future displacement and sedimentation. However, existing approaches generally fail to capture complex spatial deformations and their inter-dependencies in different areas. This work presents a novel landslide prediction model based on graph neural networks, which utilizes graph convolutions to aggregate spatial correlations among different monitored locations. Besides, we introduce a novel locally historical transformer network to capture dynamic spatio-temporal relations and predict the surface deformation. We conduct extensive experiments on real-world data and demonstrate that our model significantly outperforms state-of-the-art approaches in terms of prediction accuracy and model interpretations.https://www.mdpi.com/2072-4292/14/8/1919landslide predictiongeological data analysisgraph neural networksself-attentionspatio-temporal masking |
| spellingShingle | Ping Kuang Rongfan Li Ying Huang Jin Wu Xucheng Luo Fan Zhou Landslide Displacement Prediction via Attentive Graph Neural Network Remote Sensing landslide prediction geological data analysis graph neural networks self-attention spatio-temporal masking |
| title | Landslide Displacement Prediction via Attentive Graph Neural Network |
| title_full | Landslide Displacement Prediction via Attentive Graph Neural Network |
| title_fullStr | Landslide Displacement Prediction via Attentive Graph Neural Network |
| title_full_unstemmed | Landslide Displacement Prediction via Attentive Graph Neural Network |
| title_short | Landslide Displacement Prediction via Attentive Graph Neural Network |
| title_sort | landslide displacement prediction via attentive graph neural network |
| topic | landslide prediction geological data analysis graph neural networks self-attention spatio-temporal masking |
| url | https://www.mdpi.com/2072-4292/14/8/1919 |
| work_keys_str_mv | AT pingkuang landslidedisplacementpredictionviaattentivegraphneuralnetwork AT rongfanli landslidedisplacementpredictionviaattentivegraphneuralnetwork AT yinghuang landslidedisplacementpredictionviaattentivegraphneuralnetwork AT jinwu landslidedisplacementpredictionviaattentivegraphneuralnetwork AT xuchengluo landslidedisplacementpredictionviaattentivegraphneuralnetwork AT fanzhou landslidedisplacementpredictionviaattentivegraphneuralnetwork |