Wheres Waldo Incorporating People into Hazard and Risk Models and Applications

1
Wheres Waldo? Incorporating People into Hazard
and Risk Models and Applications

• Dr. Robert S. Chen
• Deputy Director and Senior Research Scientist,
  CIESIN, The Earth Institute at Columbia
  University
• Manager, NASA Socioeconomic Data and Applications
  Center (SEDAC)
• Secretary General, Committee for Science and
  Technology (CODATA), International Council for
  Science
• e-mail bchen_at_ciesin.columbia.edu

2
Integration of Data and Models

• Grand Challenge 2 Understand the natural
  processes that produce hazards
• New data must be collected and incorporated into
  advanced and validated models that support an
  improved understanding of underlying natural
  system processes and enhance assessment of the
  impacts

3
Relevant Population Data Exist

• 2000 census income by block group applied to
  block populations and then gridded at 250 m
  resolution

4
Historical Hazard Data Are Available

• Integrated analysis of six major natural hazards
  floods, drought, cyclones, earthquakes,
  volcanoes, and landslides in relationship to
  exposed population and economic activity and
  associated vulnerability 2.5 x 2.5 grid cells

5
Risks Can Be Compared

• Mortality weights determined from twenty years of
  EMDAT data, applied to different World Bank
  income classes

• Economic loss weights determined from twenty
  years of EMDAT data, applied to different World
  Bank income classes
• Losses also available relative to GDP density

6
Integration Requires Varied Data

• Estimate the population residing
• Within 1 and 2 km buffers of the coastline
• At an elevation of 10 meters or less

• Population data from censuses
• Irregular-shaped units
• Who slept here or usual residence
• Elevation data from earth observations (SRTM)
• Uniform gridded dataset
• Not available at 30m resolution globally!!!

7
User Interfaces Will VaryIntergraph GeoMedia

• Example from Open Geospatial Consortium New York
  City testbed
• Intergraph client showing NYC imagery with
  overlaid features
• Fire districts (red lines) drawn from
  WFS-compliant SEDAC server

8
NIEHS GIS Portal

• Post-disaster effort led by the National
  Institute for Environmental Health Sciences
  (NIEHS) to coordinate spatial environmental
  information after Hurricane Katrina
• Ad Hoc response still under development
• Open source OGC-compliant mapping tools

9
ReliefWeb/GIST

• Geographic Information Support Team (GIST)
  includes participation from UN, national, and
  non-governmental agencies
• Coordinates spatial data access, integration, and
  dissemination after the tsunami and other
  humanitarian crises

10
NASA World Wind

• Download-able PC open source application with 3-D
  visualization capabilities, on-the-fly access to
  distributed data
• View of Louisiana with SEDAC pop density for 2000
  grid from WMS, USGS urban area ortho below

11
TerraViva! SEDAC

• Self-contained PC application supporting multiple
  linked windows, multiple datasets, multiple
  projections, flexible queries
• Developed by Isciences now featuring SEDAC
  geospatial datasets

12
Google Earth

• Google Earth launched with SEDACs Global Rural
  Urban Mapping Project (GRUMP) data for the
  northeast U.S.

13
Conclusions

• Decision Makers need to know who is at risk from
  hazards
• Many relevant sources of socioeconomic data
  exist, including information on population
  exposure and vulnerability
• Data on multiple hazards are also available, and
  can be incorporated into multi-hazard risk models
  that take into account potential impacts
• Unmet needs include
• Dynamic population models that provide current
  and projected assessments of population location,
  demographic characteristics, current status,
  response
• Comparable data, information, and models across
  hazard phases, from prevention and mitigation to
  emergency response to recovery and reconstruction
• Full access to hazard-related data by both the
  research and practitioner communities, e.g.
• GOOD SAMARITAN PRINCIPLE for data use
  inhumanitarian situations?!
• Access to publicly funded data like global
  30-meter SRTM!
• See editorial last week in Science by Iwata and
  Chen

14
Lessons for GEOSS Enterprise Architecture

• GEOSS needs to accommodate wide variety of data
  types and sources, including both observed and
  model-generated data, historical and current
  data, and natural and social science data
• GEOSS needs standards that cut across both
  observational data and model communities
• GEOSS needs to meet user expectations for
  powerful but user friendly interfaces to high
  quality, integrated data and models
• GEOSS needs to pay attention early on to data
  rights, usage, confidentiality, and access,
  balancing open access needs against intellectual
  property, privacy, and national security concerns
• Last, but not least, GEOSS needs to pay attention
  to the people part of the equation!