RA2CE feature: Isolated locations

This notebook contains an example of the isolated locations analysis of the RA2CE model. This analysis can be used to assess whether locations can still reach the main part of a network during a disruption. RA2CE specifies the reason of isolation: whether the location is isolated because of link disruption close to the location or because of link disruptions in other locations further away. For example, a residential home could be disrupted in using the road network because of high water on the road right in front of the house or because they cannot leave their neighbourhood because of road flooding.

First of all, we will import the packages we need to execute the notebook:

import geopandas as gpd
import pandas as pd
import matplotlib.pyplot as plt
from pathlib import Path
from ra2ce.ra2ce_handler import Ra2ceHandler #import the ra2cehandler to run ra2ce analyses
import rasterio
import shutil

First, we will set the path to the RA2CE example folder and we will check if the network and analysis ini files indeed exist. We remove any files that might already be inside the output folders.

root_dir = Path("data","isolated_locations")  # set the root directory for the data

network_ini = root_dir / "network.ini"
assert network_ini.is_file()

analyses_ini = root_dir / "analyses.ini"
assert analyses_ini.is_file()

output_path = root_dir / "output"
output_graph_path = root_dir / "static" / "output_graph"

for p in [output_path, output_graph_path]:
    if p.exists():
        shutil.rmtree(p)

Configuring the network

Firstly we initialize RA2CE and configure the network with the network.ini and analyses.ini files. See below the contents of the network.ini file and pay special attention to the configurations in bold. These are filled in especially for this isolated locations analysis.

Network.ini content > [project] name = isolated_locations [network] directed = False source = OSM download primary_file = None diversion_file = None file_id = None polygon = map.geojson network_type = drive road_types = motorway,motorway_link,trunk,trunk_link,primary,primary_link,secondary,secondary_link,tertiary,tertiary_link save_gpkg = True [origins_destinations] origins = None destinations = None origins_names = None destinations_names = None id_name_origin_destination = None origin_count = None origin_out_fraction = None category = None [isolation] locations = POI.gpkg [hazard] hazard_map = max_flood_depth.tif hazard_id = None hazard_field_name = waterdepth aggregate_wl = max hazard_crs = EPSG:32736 [cleanup] # Can only be used if a network is created from a shapefile snapping_threshold = None segmentation_length = None merge_lines = False merge_on_id = False cut_at_intersections = False

As you can see in the network.ini file above, we submit a hazard map in the configuration file: max_flood_depth.tif. We will inspect this hazard map below.

hazard_folder = root_dir / "static" / "hazard" # find the hazard folder where you locate your hazard map
hazard_map = hazard_folder / "fake_flood.tif" # set the name of the hazard map

# Open the TIF file using rasterio
with rasterio.open(hazard_map) as src:
    # Read the TIF file as a numpy array
    tif_array = src.read(1)  # Change the band index (1) if necessary

plt.figure(figsize=(10, 10))
plt.imshow(tif_array, cmap='Blues', vmin=0, vmax=1)  # Change the colormap if desired
plt.colorbar(label='Pixel Values')
plt.title('hazard map')
plt.show()

Including, amongst others, this hazard map we initialize RA2CE and configure the network. This takes around 2 minutes, depending on internet and computer speed.

handler = Ra2ceHandler(network=network_ini, analysis=analyses_ini)
handler.configure()

The resulting network (edges and nodes) including hazard overlay is shown in the map below, bounded by the region of interest (map.geojson) that was submitted in the network.ini file.

# REGION
region = root_dir / "static" / "network" / "map.geojson"
region = gpd.read_file(region)
m = region.explore(style_kwds={"color": "purple", "fill": False, "weight": 5})

# EDGES
edges_with_hazard = root_dir / "static" / "output_graph" / "base_graph_hazard_edges.gpkg"
edges = gpd.read_file(edges_with_hazard)
edges.explore(m=m, column="EV1_ma", cmap="brg_r", style_kwds={"weight": 3})

# LOCATIONS TO CHECK ISOLATION
locations_path = root_dir / "static" / "network" / "POI.gpkg"
locations = gpd.read_file(locations_path)
locations.explore(m=m, column="building")

Running the analysis

Next, we run the isolated location analysis with ra2ce, which is already configured with the following analyses.ini file. See below the contents of this file.

Analyses.ini content > [project] name = isolated_locations [analysis1] name = multilink isolated locations analysis = multi_link_isolated_locations aggregate_wl = max threshold = 1 weighing = length category_field_name = category save_gpkg = True save_csv = True

handler.run_analysis()

Inspect results

## ROADS
m = edges.explore(column="EV1_ma", cmap="brg_r", style_kwds={"weight": 3})

## ISOLATED LOCATIONS
locations_results_path = root_dir / "output" / "multi_link_isolated_locations" / "multi_link_isolated_locations.gpkg"
locations_results = gpd.read_file(locations_results_path)
locations_results.explore(m=m, column="i_type_EV1", cmap="gist_rainbow")
m

We also check the summarized results. As you can see in the table below, most locations are isolated due to flooding. Some are isolated due to road disruptions in another part of the network.

tabulated_results_path = root_dir / "output" / "multi_link_isolated_locations" / "multi_link_isolated_locations_results.csv"
tabulated_results = pd.read_csv(tabulated_results_path)
tabulated_results.sort_values("i_type_EV1")