Topological Feature-Based Method Example

This method uses 3D shape of landslide by incorporating elevation data of landslides via SRTM 30 meters DEM.

Import topological features based module functions from Landsifier library

import topo_based_model
from topo_based_model import read_shapefiles
from topo_based_model import make_3d_polygons
from topo_based_model import get_ml_features
from topo_based_model import classify_inventory_tda
from topo_based_model import plot_topological_results
import numpy as np

In this example, we are using six landslide inventories spread over Japan archipaelogo. Out of six-inventories, three inventories are earthquake-triggered and rest three inventories are rainfall-induced inventories.

  • Earthquake-triggered inventories (Hokkaido, Iwata and Niigata region)

  • Rainfall-induced inventories (Kumamoto, Fukuoka and Saka region)

Import Shapefiles of Earthquake and Rainfall triggered landslide Inventories

  • Earthquake triggered inventories

earth_hokkaido_shapefile = read_shapefiles("E:/Germany Visit/Landslide Datasets/Japan Inventory/Earthquake_hokkaido_polygons.shp")
earth_iwata_shapefile = read_shapefiles("E:/Germany Visit/Landslide Datasets/Japan Inventory/Earthquake_iwata_polygons.shp")
earth_niigata_shapefile =read_shapefiles("E:/Germany Visit/Landslide Datasets/Japan Inventory/Earthquake_niigata_polygons.shp")

  • Rainfall triggered inventories

rain_kumamoto_shapefile = read_shapefiles("E:/Germany Visit/Landslide Datasets/Japan Inventory/Rainfall_kumamoto_polygons.shp")
rain_fukuoka_shapefile = read_shapefiles("E:/Germany Visit/Landslide Datasets/Japan Inventory/Rainfall_fukuoka_polygons.shp")
rain_saka_shapefile = read_shapefiles("E:/Germany Visit/Landslide Datasets/Japan Inventory/Rainfall_saka_polygons.shp")

DEM files

dem_location="E:/dem_japan/"
inventory_name_list=['hokkaido.tif','iwata.tif','niigata.tif','kumamoto.tif','fukuoka.tif','saka.tif']

Convert 2D landslide polygons to 3D landslide point cloud

  • Earthquake triggered inventories

pointcloud_earth_hokkaido=make_3d_polygons(earth_hokkaido_shapefile,dem_location,inventory_name_list[0],1)
pointcloud_earth_iwata=make_3d_polygons(earth_iwata_shapefile,dem_location,inventory_name_list[1],1)
pointcloud_earth_niigata=make_3d_polygons(earth_niigata_shapefile,dem_location,inventory_name_list[2],1)

  • Rainfall triggered inventories

pointcloud_rain_kumamoto=make_3d_polygons(rain_kumamoto_shapefile,dem_location,inventory_name_list[3],1)
pointcloud_rain_fukuoka=make_3d_polygons(rain_fukuoka_shapefile,dem_location,inventory_name_list[4],1)
pointcloud_rain_saka=make_3d_polygons(rain_saka_shapefile,dem_location,inventory_name_list[5],1)

Get ML features from 3d point cloud data

  • Earthquake triggered inventories

features_earth_hokkaido=get_tda_features(pointcloud_earth_hokkaido)
features_earth_iwata=get_tda_features(pointcloud_earth_iwata)
features_earth_niigata=get_tda_features(pointcloud_earth_niigata)

  • Rainfall triggered inventories

features_rain_kumamoto=get_tda_features(pointcloud_rain_kumamoto)
features_rain_fukuoka=get_tda_features(pointcloud_rain_fukuoka)
features_rain_saka=get_tda_features(pointcloud_rain_saka)

Taking one of the landslide inventory as the Testing inventory

  • Case 1: Hokkaido (Earthquake -triggerd) is a testing inventory

Training the algorithm on rest five inventories and test it on Hokkaido inventory.

earthquake_inventory_features=np.vstack((features_earth_iwata,features_earth_niigata))
rainfall_inventory_features=np.vstack((features_rain_kumamoto,features_rain_fukuoka,features_rain_saka))
test_inventory_features=features_earth_hokkaido
predict_probability=classify_inventory_tda(earthquake_inventory_features,rainfall_inventory_features,test_inventory_features)

Visualization of the classification results

plot_topological_results(predict_probability_hokkaido)
_images/hokkaido_top.png

  • Case 2: Iwata (Earthquake -triggerd) is a testing inventory

Training the algorithm on rest five inventories and test it on Iwata inventory.

earthquake_inventory_features=np.vstack((features_earth_hokkaido,features_earth_niigata))
rainfall_inventory_features=np.vstack((features_rain_kumamoto,features_rain_fukuoka,features_rain_saka))
test_inventory_features=  features_earth_iwata
predict_probability_iwata=classify_inventory_tda(earthquake_inventory_features,rainfall_inventory_features,test_inventory_features)

Visualization of the classification results

plot_topological_results(predict_probability_iwata)
_images/iwata_topo.png

  • Case 3: Niigata (Earthquake -triggerd) is a testing inventory

Training the algorithm on rest five inventories and test it on Iwata inventory.

earthquake_inventory_features=np.vstack((features_earth_hokkaido,features_earth_iwata))
rainfall_inventory_features=np.vstack((features_rain_kumamoto,features_rain_fukuoka,features_rain_saka))
test_inventory_features=features_earth_niigata
predict_probability_niigata=classify_inventory_tda(earthquake_inventory_features,rainfall_inventory_features,test_inventory_features)

Visualization of the classification results

plot_topological_results(predict_probability_niigata)
_images/niigata_topo.png

  • Case 4: Kumamoto (Rainfall -triggerd) is a testing inventory

Training the algorithm on rest five inventories and test it on Kumamoto inventory.

earthquake_inventory_features=np.vstack((features_earth_hokkaido,features_earth_iwata,features_earth_niigata))
rainfall_inventory_features=np.vstack((features_rain_fukuoka,features_rain_saka))
test_inventory_features=features_rain_kumamoto
predict_probability_kumamoto=classify_inventory_tda(earthquake_inventory_features,rainfall_inventory_features,test_inventory_features)

Visualization of the classification results

plot_topological_results(predict_probability_kumamoto)
_images/kumamoto_topo.png

  • Case 5: Fukuoka (Rainfall -triggerd) is a testing inventory

Training the algorithm on rest five inventories and test it on Kumamoto inventory.

earthquake_inventory_features=np.vstack((features_earth_hokkaido,features_earth_iwata,features_earth_niigata))
rainfall_inventory_features=np.vstack((features_rain_kumamoto,features_rain_saka))
test_inventory_features=features_rain_fukuoka
predict_probability_fukuoka=classify_inventory_tda(earthquake_inventory_features,rainfall_inventory_features,test_inventory_features)

Visualization of the classification results

plot_geometric_results(predict_probability_fukuoka)
_images/fukuoka_topo.png

  • Case 6: Saka (Rainfall -triggerd) is a testing inventory

Training the algorithm on rest five inventories and test it on Saka inventory.

earthquake_inventory_features=np.vstack((features_earth_hokkaido,features_earth_iwata,features_earth_niigata))
rainfall_inventory_features=np.vstack((features_rain_kumamoto,features_rain_fukuoka))
test_inventory_features=features_rain_saka
predict_probability_saka=classify_inventory_tda(earthquake_inventory_features,rainfall_inventory_features,test_inventory_features)

Visualization of the classification results

plot_topological_results(predict_probability_saka)
_images/iwata_topo.png