Source code for safe.engine.impact_functions_for_testing.BNPB_earthquake_guidelines
"""Impact function based on Padang 2009 post earthquake survey
This impact function estimates percentual damage to buildings as a
function of ground shaking measured in MMI.
Buildings are currently assumed to be represented in OpenStreetMap with
attributes collected as during the July 2011 Indonesian mapping competition.
This impact function maps the OSM buildings into 2 classes:
Unreinforced masonry (URM) and reinforced masonry (RM) according to
the guidelines.
"""
import numpy
from safe.impact_functions.core import FunctionProvider
from safe.impact_functions.core import get_hazard_layer, get_exposure_layer
from safe.storage.vector import Vector
from safe.common.utilities import ugettext as tr
from safe.impact_functions.mappings import osm2bnpb
from safe.impact_functions.mappings import unspecific2bnpb
from safe.impact_functions.mappings import sigab2bnpb
from safe.engine.interpolation import assign_hazard_values_to_exposure_data
# Damage 'curves' for the two vulnerability classes
damage_parameters = {'URM': [6, 7],
'RM': [6, 8]}
[docs]class EarthquakeGuidelinesFunction(FunctionProvider):
"""Risk plugin for BNPB guidelines for earthquake damage to buildings
:param requires category=='hazard' and \
subcategory.startswith('earthquake') and \
layertype=='raster' and \
unit=='MMI'
:param requires category=='exposure' and \
subcategory=='structure' and \
layertype=='vector'
"""
vclass_tag = 'VCLASS'
target_field = 'DMGLEVEL'
[docs] def run(self, layers):
"""Risk plugin for earthquake school damage
"""
# Extract data
H = get_hazard_layer(layers) # Ground shaking
E = get_exposure_layer(layers) # Building locations
keywords = E.get_keywords()
if 'datatype' in keywords:
datatype = keywords['datatype']
if datatype.lower() == 'osm':
# Map from OSM attributes to the guideline classes (URM and RM)
E = osm2bnpb(E, target_attribute=self.vclass_tag)
elif datatype.lower() == 'sigab':
# Map from SIGAB attributes to the guideline classes
# (URM and RM)
E = sigab2bnpb(E)
else:
E = unspecific2bnpb(E, target_attribute=self.vclass_tag)
else:
E = unspecific2bnpb(E, target_attribute=self.vclass_tag)
# Interpolate hazard level to building locations
H = assign_hazard_values_to_exposure_data(H, E,
attribute_name='MMI')
# Extract relevant numerical data
coordinates = E.get_geometry()
shaking = H.get_data()
N = len(shaking)
# List attributes to carry forward to result layer
attributes = E.get_attribute_names()
# Calculate building damage
count3 = 0
count2 = 0
count1 = 0
count_unknown = 0
building_damage = []
for i in range(N):
mmi = float(shaking[i]['MMI'])
building_class = E.get_data(self.vclass_tag, i)
lo, hi = damage_parameters[building_class]
if numpy.isnan(mmi):
# If we don't know the shake level assign Not-a-Number
damage = numpy.nan
count_unknown += 1
elif mmi < lo:
damage = 1 # Low
count1 += 1
elif lo <= mmi < hi:
damage = 2 # Medium
count2 += 1
elif mmi >= hi:
damage = 3 # High
count3 += 1
else:
msg = 'Undefined shakelevel %s' % str(mmi)
raise Exception(msg)
# Collect shake level and calculated damage
result_dict = {self.target_field: damage,
'MMI': mmi}
# Carry all orginal attributes forward
for key in attributes:
result_dict[key] = E.get_data(key, i)
# Record result for this feature
building_damage.append(result_dict)
# Create report
impact_summary = ('<table border="0" width="320px">'
' <tr><th><b>%s</b></th><th><b>%s</b></th></th>'
' <tr></tr>'
' <tr><td>%s:</td><td>%i</td></tr>'
' <tr><td>%s (10-25%%):</td><td>%i</td></tr>'
' <tr><td>%s (25-50%%):</td><td>%i</td></tr>'
' <tr><td>%s (50-100%%):</td><td>%i</td></tr>'
% (tr('Buildings'), tr('Total'),
tr('All'), N,
tr('Low damage'), count1,
tr('Medium damage'), count2,
tr('High damage'), count3))
impact_summary += (' <tr><td>%s (NaN):</td><td>%i</td></tr>'
% ('Unknown', count_unknown))
impact_summary += '</table>'
# Create style
style_classes = [dict(label=tr('Low damage'), min=0.5, max=1.5,
colour='#fecc5c', transparency=0),
dict(label=tr('Medium damage'), min=1.5, max=2.5,
colour='#fd8d3c', transparency=0),
dict(label=tr('High damage'), min=2.5, max=3.5,
colour='#f31a1c', transparency=0)]
style_info = dict(target_field=self.target_field,
style_classes=style_classes)
# Create vector layer and return
V = Vector(data=building_damage,
projection=E.get_projection(),
geometry=coordinates,
name='Estimated damage level',
keywords={'impact_summary': impact_summary},
style_info=style_info)
return V