SLIDE 1
GNS Science
GNS Science
Contents
- Development of an Earthquake-Induced Landslide (EIL) forecast
tool for NZ
- Model Structure and Training
Using Artificial Intelligence to forecast the location of - - PowerPoint PPT Presentation
Using Artificial Intelligence to forecast the location of - - PowerPoint PPT Presentation
Using Artificial Intelligence to forecast the location of earthquake- and post-earthquake-induced landslides r.huso@gns.cri.nz GNS Science Contents Development of an Earthquake-Induced Landslide (EIL) forecast tool for NZ Model
SLIDE 2
SLIDE 3
GNS Science
AI versus Log Regression
AI model Log Regression model
0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1
True positive rate False positive rate
Log Re gres sion - K aikoura trained AI - Murchis
- n a
nd Ina ng ahua trained
- AI model trained on Inangahua and Murchison EQ landslide
datasets, then used to forecast Kaikoura EQ landslides
- Log Regression model trained on Kaikoura EQ and landslides
- nly
- AI model statistically preforms well in forecasting Kaikoura EQ
landslides
ROC curve showing model performance
SLIDE 4
GNS Science
- Variables investigated (12)
- Variables with statistical significance:
– Slope angle – Distance to surface fault rupture – Elevation – Geology – PGA or PGV – LSR: Local Slope Relief
Faults Landslides
SLIDE 5
GNS Science
Data Difficulties
- Data discontinuities and one-hot encoding
– Natural ordinal relationship between the categories?
- ‘cold’, warm’, and ‘hot’
– Not suitable for one-hot encoding
- Quaternary gravel, debris, sand
- Neogene siltstone, sandstone
- Cretaceous conglomerate, igneous rocks, limestone, mudstone, siltstone
- Early Cretaceous igneous rocks
- Paleogene igneous rocks, limestone, limestone
SLIDE 6
GNS Science
Data Difficulties (2)
- Data scales
– Distance (~100,000 m) – Elevation (~2000 m) – LSR (~200) – PGA (~100) – Slope (~90) – GeolCode (1, 2, 3, 4, 5)
SLIDE 7
GNS Science
Data discontinuities
GroupInputData Config.json ScaleInputData CaptureDataLimits HyperCubeSmoothing
SLIDE 8
GNS Science
Structure Difficulties
- Model structure
– Number of hidden layers – Number of nodes per layer – Amount of “Dropout”
SLIDE 9
GNS Science
Structure Difficulties
HyperparameterOpt Config.json HyperparameterOpt OptimalEpochs
SLIDE 10
GNS Science
Model Training and Testing
- Train the model
- Apply to Out Of Sample data (if any)
- Rank the input features
- Show charts
SLIDE 11
GNS Science
Model Training and Testing
Config.json TrainModel TrainModel ApplyModel FeatureRanking ShowCharts
SLIDE 12
GNS Science
Model Charts – Confusion Matrix
SLIDE 13
GNS Science
Model Charts – Receiver Operating Characteristic
SLIDE 14
GNS Science
Model Charts – Receiver Operating Characteristic
SLIDE 15
GNS Science
Summary
- Rapid forecasts of landslide probability and impacts in near-real
time (5-7 minutes) after an event would help to focus such response efforts
- Several discrete steps are required to produce a useful model that
can be applied when an earthquake event happens
SLIDE 16
GNS Science