RiskCity Exercise 5: Generating an elements at risk database

1
RiskCityExercise 5 Generating an elements at
risk database

• Cees van Westen (ed)

2
Elements at risk / Assets

• What may be impacted by a hazard event?

3
Two options

• When you dont have any available data
• We assume that you have at least a high
  resolution image from Google Earth
• When you have available data
• Building footprint map
• Lidar DSM
• Census data

Depending on your interest in the topic you may
select to either do Exercise 3.1 (creating a
database by starting from scratch), or Exercise
3.2 (creating a database with available footprint
information). You can also decide to do both
exercises, although that might perhaps take a bit
too much time
4
If you dont have data

• You have to
• Generate mapping units
• Create the attribute data for
• Urban land use
• Number of buildings
• Population

5
Flowchart do it yourself option
Input data
Screen digitize boundaries
Sample buildings by landuse type
Polygonize
Interpret land use type
Calculate based on land use type building
Calculate based on land use type
6
Downloading imagery from Google Earth

• Many area in the world are covered by high
  resolution imagery.
• Better first consult than download
• For detailed download you need Google Earth Pro
  (cost 400 US )
• You can download 4000 4800 resolution
• Here we dont have Google Earth Pro on all
  computers. Only one in room 4 105
• We have downloaded it already for you
• At home you might like to try the trial version
  of the Goolge Earth Pro, which allows to download
  high resolution images. Go to http//earth.google
  .com/intl/en/product_comparison.html

7
Digitizing maps
Sensor
8
Digitizing mapping units
Screen digitizing from high resolution image, on
the basis of a digital road map
Checking segments, and generation of polygons
with unique identifiers
9
Digitizing mapping units
High res image
Digitize a new point
Select lines and rename / delete them
Digitize segments
Create a node / remove a node
Added segment
Select points and move them
10
Check segments

• Before making polygons you have to make sure all
  lines are connected
• Error types
• Dead end in segment (1)
• Intersection without node (2, 3)
• Double line (4)
• Self overlap (5)

Digitize segments
Check segments
Added segment
11
Determining land use
Generation of land use legend, with relevant
classes for vulnerability assessment, and keeping
in mind population difference
Interpreting predominant landuse from the high
resolution image
12
Landuse classification

• Urban landuse mapping

13
Fill in missing parts
14
Estimating number of buildings

• Methods
• Count all buildings in the map.
• Sample buildings for landuse types

• Steps
• Calculate building size
• building_sizeiff(buildings_sampled0,0, area/
  buildings_sampled)
• Average building size per land use type
• nr_buildingsiff(isundef(buildings_sampled),area/
  avg_building_size, buildings_sampled)

15
Estimating population distribution

• Link the number of people per building to land
  use type

• Daytime_populationnr_buildings
  person_building daytime

16
If you have available data
17
Number of buildings

• Cross Building map with mapping units.
• how much of the mapping unit is not built-up
• how many individual buildings there are per
  mapping unit
• the average building size for each urban land use.

Areavacantiff(isundef(building_map),area,0) Area
_buildingiff(isundef(building_map),?,area) Build
ingiff(isundef(building_map),0,1)
18
Aggregate results to mapping units

• Calculate per mapping unit
• Total_area total area per mapping unit
• Total_vacant_area total vacant area per mapping
  unit
• Avg_Size average building size per mapping unit
• Nr_buildings number of buildings per mapping
  unit
• Percvacant Total_vacant_area /Total_area

19
Building height floorspace
DEM from topomap
DEM from Lidar
minus
Division by avg. building height
Masking out areas without buildings
Landuse map
20
Altitude of objects
Command Line
21
Calculate number of floors

• Altitude_difLidarDEM-TopoDem
• floor_nriff(Altitude_dif lt3,0, Altitude_dif /3)
• Floorsiff(isundef(building_map),0,floor_nr)

22
Calculate height of buildings

• First we cross the Building_map with the map
  Floors, which gives us all the combinations of
  floors per building type.
• Then we calculate per building the maximum number
  of floors, and the total floor space for each
  building.
• The resulting values are then read in the Cross
  table that links the mapping units with the
  building IDs (Mapping_units_building).
• And finally the total floorspace information is
  aggregated into the table Mapping_units_attributes

23
Calculate floorspace

• FloorspaceNr_floorsArea_building
• Open the cross table Mapping_units_building. And
  join with the table Building_map. Read in the
  columns Nr_floors and Floorspace
• Aggregate to Table Mapping_units_attributes
• Nr_floors_avg average number of floors per
  building in mapping unit
• Floorspace floorspace per mapping unit

24
Population estimate

• We have information on the population per ward.
• We know the floorspace per mapping unit
• We can therefore distribute the total population
  per ward over the mapping unit, also keeping in
  mind the land use types.
• This exercise is not written out something for
  the final project