An ensemble neural network approach for space–time landslide predictive modelling

International Journal of Applied Earth Observation and Geoinformation, Год журнала: 2024, Номер 132, С. 104037 - 104037

Опубликована: Июль 29, 2024

There is an urgent need for accurate and effective Landslide Early Warning Systems (LEWS). Most LEWS are currently based on temporally-aggregated measures of rainfall derived from either in-situ measurements or satellite-based estimates. Relying a summary metric precipitation may not capture the complexity signal its dynamics in space time triggering landslides. Here, we present proof-of-concept constructing integrated spatio-temporal modelling framework. Our proposed methodology builds upon recent approach that uses daily series instead traditional scalar aggregation. Specifically, partition study area into slope units use Gated Recurrent Unit (GRU) to process satellite-derived combine output features with second neural network (NN) tasked capturing effect terrain characteristics. To assess if our enhances accuracy, applied it Vietnam benchmarked against counterpart where replaced corresponding representative cumulated precipitation. The duration was set at 14 days as proved produce best performance. results show protocol leads better performance hindcasting landslides when making continuous over time. While tested here, can be extended obtained weather forecasts, potentially leading actual landslide forecasts.

Язык: Английский

---

Spatiotemporal changes of landslide susceptibility in response to rainfall and its future prediction — A case study of Sichuan Province, China

Ecological Informatics, Год журнала: 2024, Номер unknown, С. 102862 - 102862

Опубликована: Окт. 1, 2024

Язык: Английский

---

Spatiotemporal dynamics of landslide susceptibility under future climate change and land use scenarios

Environmental Research Letters, Год журнала: 2024, Номер 19(12), С. 124016 - 124016

Опубликована: Окт. 23, 2024

Abstract Mountainous landslides are expected to worsen due environmental changes, yet few studies have quantified their future risks. To address this gap, we conducted a comprehensive analysis of the eastern Hindukush region Pakistan. A geospatial database was developed, and logistic regression employed evaluate baseline landslide susceptibility for 2020. Using latest coupled model intercomparison project 6 models under three shared socioeconomic pathways (SSPs) cellular automata-Markov model, projected rainfall land use/land cover patterns 2040, 2070, 2100, respectively. Our results reveal significant changes in use patterns, particularly long-term (2070 2100). Future then predicted based on these projections. By high-risk areas increase substantially all SSP scenarios, with largest increases observed SSP5-8.5 (56.52%), SSP2-4.5 (53.55%), SSP1-2.6 (22.45%). will rise by 43.08% (SSP1-2.6), 40.88% (SSP2-4.5), 12.60% (SSP5-8.5). However, minimal compared baseline, 9.45% 1.69% 7.63% These findings provide crucial insights into relationship between risks support development climate risk mitigation, planning, disaster management strategies mountainous regions.

Язык: Английский

---

Multi‐hazard assessment using machine learning and remote sensing in the North Central region of Vietnam

Transactions in GIS, Год журнала: 2023, Номер 27(5), С. 1614 - 1640

Опубликована: Июль 23, 2023

Abstract Natural hazards constitute a diverse category and are unevenly distributed in time space. This hinders predictive efforts, leading to significant impacts on human life economies. Multi‐hazard prediction is vital for any natural hazard risk management plan. The main objective of this study was the development multi‐hazard susceptibility mapping framework, by combining two hazards—flooding landslides—in North Central region Vietnam. accomplished using support vector machines, random forest, AdaBoost. input data consisted 4591 flood points, 1315 landslide 13 conditioning factors, split into training (70%), testing (30%) datasets. accuracy models' predictions evaluated statistical indices root mean square error, area under curve (AUC), absolute coefficient determination. All proposed models were good at predicting susceptibility, with AUC values over 0.95. Among them, value machine model 0.98 0.99 flood, respectively. For forest model, these 0.98, AdaBoost, they 0.99. maps built maps. results showed that approximately 60% affected landslides, 30% 8% both hazards. These illustrate how one regions Vietnam most severely hazards, particularly flooding, landslides. adapt evaluate different scales, although expert intervention also required, optimize algorithms. can provide valuable point reference decision makers sustainable land‐use planning infrastructure faced multiple prevent reduce more effectively frequency floods landslides their damage property.

Язык: Английский

---

Assessing the impact of climate change on landslide recurrence intervals in Nakhon Si Thammarat Province, Thailand, using CMIP6 climate models

Progress in Disaster Science, Год журнала: 2024, Номер 22, С. 100330 - 100330

Опубликована: Апрель 1, 2024

Climate variability and climate change may influence the frequency recurrence interval of landslides. Precipitation, as a main triggering factor landslides, be influenced by change. Changes in precipitation directly affect landslide intervals. Considering change, partial duration series method critical rainfall threshold are combined with simulated Phase 6 Coupled Model Intercomparison Project (CMIP6) general circulation models (GCMs) to predict changes future intervals Nakhon Si Thammarat Province, Thailand. The analytical results predicted over next 20 years (2023 2042). SSP1–2.6 SSP2–4.5 adopted socioeconomic development scenarios. According predictions, showed that return period occurrence will shorter than historical period; moreover, fluctuate greatly. DWR meteorological station shows most fluctuation for SSP1–2.6. In addition, Station experiences significant decrease approximately 35% under For SSP2–4.5, period. observed decline intervals, reduction 40%. Hence, landslides increase future. A comparison between revealed yielded lower periods.

Язык: Английский

---